1 tbd Sun Microsystems, Inc.
4 2.6.16.60-0.39.3 (SLES 10),
5 2.6.18-128.1.14.el5 (RHEL 5),
6 2.6.22.14 vanilla (kernel.org).
7 * Client support for unpatched kernels:
8 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
9 2.6.16 - 2.6.21 vanilla (kernel.org)
10 * Recommended e2fsprogs version: 1.41.6.sun1
11 * Note that reiserfs quotas are disabled on SLES 10 in this kernel.
12 * RHEL 4 and RHEL 5/SLES 10 clients behaves differently on 'cd' to a
13 removed cwd "./" (refer to Bugzilla 14399).
14 * File join has been disabled in this release, refer to Bugzilla 16929.
18 Description: Improved cleanup procedure for conf-sanity test_46a.
22 Description: Do not start bulk transfer, if client tried to reconnect.
23 Details : Do not start bulk transfer, if client tried to reconnect,
24 but failed due to active rpcs. Check for exp_abort_active_req
29 Description: Prevent inconsistences between linux and lustre mount structures.
30 Details : Wait indefinitely in server_wait_finished() until mnt_count drops.
31 Make the sleep interruptible.
33 Severity : enhancement
35 Description: provide server to client comms path
36 Details : server to client communications path via new LDLM_SET_INFO rpc
40 Description: truncate starts GFP_FS allocation under transaction causing deadlock
41 Details : ldiskfs_truncate calls grab_cache_page which may start page
42 allocation under an open transaction. This may lead to
43 calling prune_icache with consequent lustre reentrance.
47 Description: send correctly lsm on open replay
48 Details : MDS is trust to LSM size on replay open, but client can set wrong size
53 Description: enable client interface failover
54 Details : When a child reconnects from another NID, properly update export
55 nid hash position and ldlm reverse import.
57 Severity : enhancement
59 Description: Communicate OST degraded/readonly state via statfs to MDS
60 Details : Flags in the statfs returned from OSTs indicate whether the
61 OST is in a degraded RAID state, or if the filesystem has
62 turned read-only after a filesystem error is detected.
67 Description: don't panic if EPROTO was hit when reading symlink
68 Details : correctly handling request reference in error cases.
70 Severity : enhancement
72 Description: Add LustreNetLink, a kernel-userspace communcation path.
74 Severity : enhancement
76 Description: Update kernel to SLES10 SP2 2.6.16.60-0.39.3.
81 Description: access to llog context before init.
82 Details : move handling CATALOGS file at osc layer and forbid access to llog
86 Frequency : rare, only when client is writing to OST on local node
88 Description: Avoid deadlock for local client writes
89 Details : Use new OBD_BRW_MEMALLOC flag to notify OST about writes in the
90 memory freeing context. This allows OST threads to set the
91 PF_MEMALLOC flag on task structures in order to allocate memory
92 from reserved pools and complete IO.
93 Use GFP_HIGHUSER for OST allocations for non-local client writes,
94 so that the OST threads generate memory pressure and allow
95 inactive pages to be reclaimed.
97 Severity : enhancement
99 Description: Update kernel to RHEL5.3 2.6.18-128.1.14.el5.
104 Description: lock ordering violation between &cli->cl_sem and _lprocfs_lock
105 Details : move ldlm namespace creation in setup phase to avoid grab
106 _lprocfs_lock with cli_sem held.
110 Description: Temporarily disable grant shrink.
111 Details : Disable the feature for debugging.
114 Frequency : only when formatting test filesystems
116 Description: Unable to run several mkfs.lustre on loop devices at the same time
117 Details : mkfs.lustre returns error 256 on the concurrent loop devices
118 formatting. The solution is to proper handle the error.
120 Severity : enhancement
122 Description: Update kernel to RHEL5.3 2.6.18-128.1.6.el5.
124 Severity : enhancement
126 Description: Update kernel to SLES10 SP2 2.6.16.60-0.37.
131 Description: resolve race between obd_disconnect and class_disconnect_exports
132 Details : if obd_disconnect will be called to already disconnected export he
133 forget release one reference and osc module can't unloaded.
135 Severity : enhancement
137 Description: Allow tuning service thread via /proc
138 Details : For each service a new
139 /proc/fs/lustre/{service}/*/thread_{min,max,started} entry is
140 created that can be used to set min/max thread counts, and get the
141 current number of running threads.
146 Descriptoin: don't return error if have partially created objects for file.
147 Details : lov_update_create_set uses set->set_success as index for created
148 objects, so if some requests will be failed, they will have hole
149 at end of array and qos_shrink_lsm for allocate correct lsm.
151 Severity : enhancement
153 Description: Update OFED support to 1.4.1
158 Description: Reduce small size read RPC
159 Details : Set read-ahead limit for every file and only do read-ahead when
160 available read-ahead pages are bigger than 1M to avoid small size
163 Severity : enhancement
165 Description: move AT tunable parameters for more consistent usage
166 Details : add AT tunables under /proc/sys/lustre, add to conf_param parsing
168 Severity : enhancement
170 Description: add lazystatfs mount option to allow statfs(2) to skip down OSTs
171 Details : allow skip disconnected ost for send statfs request and hide error
177 Description: fix racy locking of mballoc block bitmaps causing BUG
178 Details : The locking of the mballoc buddy bitmap and the in-memory
179 block bitmap was using two different spin locks in some
180 cases. This made it possible to incorrectly access the
181 mballoc bitmap while another process was modifying it,
182 causing a sanity assertion to fail. While no on-disk corruption
183 was reported, there was some risk of this happening.
186 Frequency : rare, on llog test 6
188 Descriptoin: don't allow connect to already connected import
189 Details : allowing connect to already connected import is hide connecting problem.
194 Descriptoin: allow kill process which wait statahead result
195 Details : for some reasons 'ls' can stick in waiting result from statahead,
196 in this case need way for kill this process.
198 Severity : enhancement
200 Description: Add state history info file, enhance import info file
201 Details : Track import connection state changes in a new osc/mdc proc file;
202 add overview-type data to the osc/mdc import proc file.
204 Severity : enhancement
206 Description: MDS create should not wait for statfs RPC while holding DLM lock.
208 Severity : enhancement
210 Description: Update to RHEL5U3 kernel-2.6.18-128.1.1.el5.
215 Descriptoin: OST grant too much space to client even there are not enough space.
216 Details : Client will shrink its grant cache to OST if there are no write
217 activity over 6 mins (GRANT_SHRINK_INTERVAL), and OST will retrieve
218 this grant cache if there are already not enough avaible space
219 (left_space < total_clients * 32M).
222 Frequency : start MDS on uncleanly shutdowned MDS device
224 Descriptoin: ll_sync thread stay in waiting mds<>ost recovery finished
225 Details : stay in waiting mds<>ost recovery finished produce random bugs
226 due race between two ll_sync thread for one lov target. send
227 ACTIVATE event only if connect realy finished and import have
231 Frequency : rare, connect and disconnect target at same time
233 Descriptoin: ASSERTION(atomic_read(&imp->imp_inflight) == 0
234 Details : don't call obd_disconnect under lov_lock. this long time
235 operation and can block ptlrpcd which answer to connect request.
240 Descriptoin: don't lose wakeup for imp_recovery_waitq
241 Details : recover_import_no_retry or invalidate_import and import_close can
242 both sleep on imp_recovery_waitq, but we was send only one wakeup
246 Frequency : always with long access acl
248 Descriptoin: mds can't pack reply with long acl.
249 Details : mds don't control size of acl but they limited by reint/getattr
252 Severity : enhancement
254 Description: Update to SLES10 kernel-2.6.16.60-0.33.
256 Severity : enhancement
258 Description: Update to RHEL5 kernel-2.6.18-92.1.22.el5.
261 Frequency : start MDS on uncleanly shutdowned MDS device
263 Descriptoin: aborting recovery hang on MDS
264 Details : don't throttle destroy RPCs for the MDT.
267 Frequency : on remount
269 Description: external journal device not working after the remount
270 Details : clear dev_rdonly flag for external journal devices in
276 Description: shutdown vs evict race
277 Details : client_disconnect_export vs connect request race.
278 if client will evicted at this time - we start invalidate
279 thread without referece to import and import can be freed
283 Frequency : rare, need acl's on inode.
285 Description: client can't handle ost additional correctly
286 Details : if ost was added after client connected to mds client can have
287 hit lnet_try_match_md ... to big messages to wide striped files.
288 in this case need teach client to handle config events about add
289 lov target and update client max ea size at that event.
291 Severity : enhancement
293 Description: Changelogs
294 Details : Changelogs are a lightweight mechanism to track filesystem
295 metadata and namespace changes. The changelog is recorded
296 permanently on the MDTs, and is periodically "consumed" / purged
297 when records are no longer needed.
299 Severity : enhancement
301 Description: compact fld format with extents
302 Details : Store range of seq rather than every seq in FLD. Seq
303 controller update FLD rather than clients. In Case of CMD, mdt0
304 has FLD, all other metadata server act as non persistent proxy
305 for FLD queries and cache fld entries in fld cache.
310 Description: don't skip ost target if they assigned to file
311 Details : Drop slow OSCs if we can, but not for requested start idx.
312 This means "if OSC is slow and it is not the requested
313 start OST, then it can be skipped, otherwise skip it only
314 if it is inactive/recovering/out-of-space.
318 Description: more cleanup in mds_lov
319 Details : not send LOV EA under replay, we can't know about they size at this
320 time. Don't allow client connect to mds before any ost connected,
321 for avoid problems with LOV EA size and returning EIO to client.
323 Severity : enhancement
325 Description: Interoperability at server side (Disk interoperability)
327 Severity : enhancement
329 Description: Update to RHEL5 kernel-2.6.18-92.1.17.el5.
331 Severity : enhancement
333 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.31.
335 Severity : enhancement
337 Description: New client IO stack (CLIO).
339 Severity : enhancement
341 Description: Commit on sharing. Eliminate inter-client dependencies between
342 uncommitted transactions by doing transaction commits.
343 Thereby clients may recovery independently.
346 Frequency : Create a symlink file with a very long name
348 Description: ldlm_cancel_pack()) ASSERTION(max >= dlm->lock_count + count)
349 Details : If there is no extra space in the request for early cancels,
350 ldlm_req_handles_avail() returns 0 instead of a negative value.
352 Severity : enhancement
354 Description: Add /proc entry for import status
355 Details : The mdc, osc, and mgc import directories now have
356 an import directory that contains useful import data for debugging
359 Severity : enhancement
361 Description: Re-disable certain /proc logging
362 Details : Enable and disable client's offset_stats, extents_stats and
363 extents_stats_per_process stats logging on the fly.
366 Frequency : Only on FC kernels 2.6.22+
368 Description: oops in statahead
369 Details : Do not drop reference count for the dentry from VFS when lookup,
370 VFS will do that by itself.
372 Severity : enhancement
374 Description: Generic /proc file permissions
375 Details : Set /Proc file permissions in a more generic way to enable non-
376 root users operate on some /proc files.
380 Description: Hitting mdc_commit_close() ASSERTION
381 Details : Properly handle request reference release in
382 ll_release_openhandle().
386 Description: quota recovery deadlock during mds failover
387 Details : This patch includes att18982, att18236, att18237 in bz14840.
389 1. fix osts hang when mds does failover with quotaon
390 2. prevent watchdog storm when osts threads wait for the
395 Frequency : only patchless client
396 Description: add workaround for race between add/remove dentry from hash
398 Severity : enhancement
400 Description: Allow OST glimpses to return PW locks
404 Description: LBUG when llog conf file is full
405 Details : When llog bitmap is full, ENOSPC should be returned for plain log.
409 Description: Prevent import from entering FULL state when server in recovery
413 Description: service mount cannot take device name with ":"
414 Details : Only when device name contains ":/" will mount treat it as
420 Description: replace ptlrpcd with the statahead thread to interpret the async
421 statahead RPC callback
425 Frequency : on recovery
426 Description: I/O failures after umount during fail back
427 Details : if client reconnected to restarted server we need join to recovery
428 instead of find server handler is changed and process self
429 eviction with cancel all locks.
431 Severity : enhancement
433 Description: Update to RHEL5 kernel-2.6.18-92.1.10.el5.
435 Severity : enhancement
437 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.27.
439 Severity : enhancement
441 Description: Upcall on Lustre log has been dumped
442 Details : Allow for a user mode script to be called once a Lustre log has
443 been dumped. It passes the filename of the dumped log to the
444 script, the location of the script can be specified via
445 /proc/sys/lnet/debug_log_upcall.
450 Description: avoid idr_remove called for id which is not allocated.
451 Details : Move assigment s_dev for clustered nfs to end of initialization,
452 to avoid problem with error handling.
457 Description: avoid Already found the key in hash [CONN_UNUSED_HASH] messages
458 Details : When connection is reused this not moved from CONN_UNUSED_HASH
459 into CONN_USED_HASH and this prodice warning when put connection
460 again in unused hash.
463 Severity : enhancement
465 Description: File striping can now be set to use an arbitrary pool of OSTs.
467 Severity : enhancement
469 Description: Export bytes_read/bytes_write count on OSC/OST.
473 Description: Early reply size mismatch, MGC loses connection
474 Details : Apply the MGS_CONNECT_SUPPORTED mask at reconnect time so
475 the connect flags are properly negotiated.
480 Description: quotas are not honored with O_DIRECT
481 Details : all writes with the flag O_DIRECT will use grants which leads to
482 this problem. Now using OBD_BRW_SYNC to guard this.
486 Description: add quota statistics
487 Details : 1. sort out quota proc entries and proc code.
488 2. add quota statistics
490 Severity : enhancement
492 Description: enable quota support for HEAD.
496 Description: Properly propagate oinfo flags from lov to osc for statfs
497 Details : restore missing copy oi_flags to lov requests.
499 Severity : enhancement
501 Description: Add man pages for llobdstat(8), llstat(8), plot-llstat(8),
502 : l_getgroups(8), lst(8), routerstat(8)
503 Details : included man pages for llobdstat(8), llstat(8),
504 : plot-llstat(8), l_getgroups(8), lst(8), routerstat(8)
506 Severity : enhancement
508 Description: configure's --enable-quota should check the
509 : kernel .config for CONFIG_QUOTA
510 Details : configure is terminated if --enable-quota is passed but
511 : no quota support is in kernel
515 Description: Remove portals compatibility
516 Details : Remove portals compatibility, not interoperable with releases
521 Description: Resolve device initialization race
522 Details : Prevent proc handler from accessing devices added to the
523 obd_devs array but yet be intialized.
525 Severity : enhancement
527 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.23.
529 Severity : enhancement
531 Description: Update to RHEL5 kernel-2.6.18-92.1.6.el5.
536 Description: Using wrong pointer in osc_brw_prep_request
537 Details : Access to array[-1] can produce panic if kernel compiled with
538 CONFIG_PAGE_ALLOC enabled
542 Description: Client runs out of low memory
543 Details : Consider only lowmem when counting initial number of llap pages
547 Description: *optional* service tags registration
548 Details : if the "service tags" package is installed on a Lustre node
549 When the filesystem is mounted, a local-node service tag will
550 be created. See http://inventory.sun.com/ for more information
551 about the Service Tags asset management system.
555 Description: Kernel BUG tries to release flock
556 Details : Lustre does not destroy flock lock before last reference goes
557 away. So always drop flock locks when client is evicted and
558 perform unlock regardless of successfulness of speaking to MDS.
562 Description: add refcount protection for osc callbacks, avoid panic on shutdown
566 Description: sanity test 65a fails if stripecount of -1 is set
567 Details : handle -1 striping on filesystem in ll_dirstripe_verify
572 Description: ASSERTION(CheckWriteback(page,cmd)) failed
573 Details : badly clear PG_Writeback bit in ll_ap_completion can produce false
576 Severity : enhancement
578 Description: Update to RHEL5 kernel-2.6.18-53.1.21.el5.
582 Description: do not process already freed flock
583 Details : flock can possibly be freed by another thread before it reaches
584 to ldlm_flock_completion_ast.
588 Description: LBUG during stress test
589 Details : Need properly lock accesses the flock deadlock detection list.
593 Description: oops in page fault handler
594 Details : kernel page fault handler can return two special 'pages' in error
595 case, don't try dereference NOPAGE_SIGBUS and NOPAGE_OMM.
599 Description: timeout with invalidate import.
600 Details : ptlrpcd_check call obd_zombie_impexp_cull and wait request which
601 should be handled by ptlrpcd. This produce long age waiting and
602 -ETIMEOUT ptlrpc_invalidate_import and as result LASSERT.
604 Severity : enhancement
606 Description: Update to RHEL5 kernel-2.6.18-53.1.19.el5.
610 Description: enable MGS and MDT services start separately
611 Details : add a 'nomgs' option in mount.lustre to enable start a MDT with
612 a co-located MGS without starting the MGS, which is a complement
613 to 'nosvc' mount option.
617 Frequency : after recovery
618 Description: precreate to many object's after del orphan.
619 Details : del orphan st in oscc last_id == next_id and this triger growing
620 count of precreated objects. Set flag LOW to skip increase count
621 of precreated objects.
625 Frequency : rare, on clear nid stats
626 Description: ASSERTION(client_stat->nid_exp_ref_count == 0)
627 Details : when clean nid stats sometimes try destroy live entry,
628 and this produce panic in free.
632 Description: Stack overflow during MDS log replay
633 ease stack pressure by using a thread dealing llog_process.
637 Description: wait until IO finished before start new when do lock cancel.
638 Details : VM protocol want old IO finished before start new, in this case
639 need wait until PG_writeback is cleared until check dirty flag and
640 call writepages in lock cancel callback.
642 Severity : enhancement
644 Description: using special macro for print time and cleanup in includes.
648 Description: mds_mfd_close() ASSERTION(rc == 0)
649 Details : In mds_mfd_close(), we need protect inode's writecount change
650 within its orphan write semaphore to prevent possible races.
654 Description: Obsolete CURRENT_SECONDS and use cfs_time_current_sec() instead.
658 Frequency : rare, on shutdown ost
659 Description: don't hit live lock with umount ost.
660 Details : shrink_dcache_parent can be in long loop with destroy dentries,
661 use shrink_dcache_sb instead.
665 Description: don't panic with use echo client
666 Details : echo client pass NULL as client nid pointer and this produce null
671 Description: fix build on ppc32
672 Details : compile code with -m64 flag produce wrong object file for ppc32.
676 Description: add message levels for liblustreapi
680 Description: fix for occasional failure case of -ENOSPC in recovery-small tests
681 Details : Move the 'good_osts' check before the 'total_bavail' check. This
682 will result in an -EAGAIN and in the exit call path we call
683 alloc_rr() which will with increasing aggressiveness attempt to
684 aquire precreated objects on the minimum number of required OSCs.
688 Description: Use old size assignment to avoid deadlock
689 Details : Reverts the changes in bugs 2369 and bug 14138 that introduced
690 the scheduling while holding a spinlock. We do not need locking
691 for size in ll_update_inode() because size is only updated from
692 the MDS for directories or files without objects, so there is no
693 other place to do the update, and concurrent access to such inodes
694 are protected by the inode lock.
698 Description: resolve "_IOWR redefined" build error on SLES10
702 Description: dump the memory debugging after all modules are unloaded to
703 suppress false negative in conf_sanity test 39
705 Severity : enhancement
707 Description: build kernel-ib packages for OFED 1.3 in our release cycle
712 Description: fix SLES kernel versioning
713 Details : the kernel version for our SLES 10 kernel did not include a "-"
714 before the "smp" at the end. while this was not a problem in
715 general, it did mean that software trying to use the kernel
716 version to try to detect a vendor specific kernel would fail.
717 this was most evident by the OFED build scripts.
721 Description: Don't update lov_desc members until making sure they are valid
722 Details : When updating lov_desc members via proc fs, need fix their
723 validities before doing the real update.
727 Description: don't put request into delay list while invalidate in flight.
728 Details : ptlrpc_delay_request sometimes put in delay list while invalidate
729 import in flight. this produce timeout for invalidate and sometimes
730 can cause stale data.
734 Frequency : on ppc only
735 Description: not convert ost objects for directory because it's not exist.
736 Details : ll_dir_getstripe assume dirrectory has ost objects but this wrong.
740 Description: Add FMODE_EXEC file flag for SLES10 SP1 kernel.
742 Severity : enhancement
744 Description: Update to support 2.6.22.14 vanilla kernel.
748 Frequency : rare, on recovery
749 Description: read procfs can produce deadlock in some situation
750 Details : Holding lprocfs lock which send rpc can produce block for destroy
751 obd objects and this also block reconnect with -EALREADY. This
752 isn't fix all lprocfs bugs - but make it rare.
754 Severity : enhancement
756 Description: Update kernel to RHEL5 2.6.18-53.1.14.el5.
759 Frequency : frequent on X2 node
761 Description: mdc_set_open_replay_data LBUG
762 Details : Set replay data for requests that are eligible for replay.
766 Description: lustre_mgs: operation 101 on unconnected MGS
767 Details : When MGC is disconnected from MGS long enough, MGS will evict the
768 MGC, and late on MGC cannot successfully connect to MGS and a lot
769 of the error messages complaining that MGS is not connected.
773 Frequency : on network error
774 Description: panic with double free request if network error
775 Details : mdc_finish_enqueue is finish request if any network error ocuring,
776 but it's true only for synchronus enqueue, for async enqueue
777 (via ptlrpcd) this incorrect and ptlrpcd want finish request
780 Severity : enhancement
782 Description: client-side metadata stat-ahead during readdir(directory readahead)
783 Details : perform client-side metadata stat-ahead when the client detects
784 readdir and sequential stat of dir entries therein
787 Frequency : on start mds
789 Description: Implement get_info(last_id) in obdfilter.
792 Frequency : occasional
794 Description: Correctly check stale fid, not start epoch if ost not support SOM
795 Details : open with flag O_CREATE need set old fid in op_fid3 because
796 op_fid2 was overwritten with new generated fid, but mds can answer
797 with one of these two fids and both is not stale. Setattr
798 incorrectly started an epoch and assume will be called
799 done_writing, but without SOM done_writing ever being called.
802 Frequency : rare, depends on device drivers and load
804 Description: MDS or OSS nodes crash due to stack overflow
805 Details : Code changes in 1.8.0 increased the stack usage of some functions.
806 In some cases, in conjunction with device drivers that use a lot
807 of stack the MDS (or possibly OSS) service threads could overflow
808 the stack. One change which was identified to consume additional
809 stack has been reworked to avoid the extra stack usage.
812 Frequency : occasional
814 Description: Do not fail import if osc_interpret_create gets -EAGAIN
815 Details : If osc_interpret_create got -EAGAIN it immediately exits and
816 wakeup oscc_waitq. After wakeup oscc_wait_for_objects call
817 oscc_has_objects and see OSC has no objests and call
818 oscc_internal_create to resend create request.
820 Severity : enhancement
822 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.5.
824 Severity : enhancement
826 Description: Update to RHEL5 latest kernel-2.6.18-53.1.13.el5.
829 Frequency : very rare
831 Description: Fix replay if there is an un-replied request and open
832 Details : In some cases, older replay request will revert the
833 mcd->mcd_last_xid on MDS which is used to record the client's
836 Severity : enhancement
838 Description: Update to RHEL5 latest kernel-2.6.18-53.1.6.el5.
840 Severity : enhancement
842 Description: Add rhel5 support to HEAD.
844 Serverity : enhancement
846 Description: Update RHEL4 kernel to 2.6.9-67.0.4.
851 Description: Don't allow skipping OSTs if index has been specified.
852 Details : Don't allow skipping OSTs if index has been specified, make
853 locking in internal create lots better.
857 Description: LBUG in ptlrpc_check_set() bad phase ebc0de00
858 Details : access to bitfield in structure is always rounded to long
859 and this produce problem with not atomic change any bit.
863 Description: Lustre make rpms failed.
864 Details : Remove ldiskfs spec file to avoids rpmbuild be confused when
865 builds Lustre rpms from tarball.
868 Frequency : rare on shutdown ost
870 Description: If llog cancel was not send before clean_exports phase, this can
871 produce deadlock in llog code.
872 Details : If llog thread has last reference to obd and call class_import_put
873 this produce deadlock because llog_cleanup_commit_master wait when
874 last llog_commit_thread exited, but this never success because was
875 called from llog_commit_thread.
879 Description: allow userland application know is lost one of stripes.
880 Details : fill lvb_blocks with error code on ost and return it to
881 application if error flag found.
885 Description: NULL lov_tgts causing MDS oops
886 Details : more safe checks for NULL lov_tgts for avoid oops.
888 Severity : enhancement
890 Description: Update to RHEL4 latest kernel-2.6.9-67.0.1.EL.
894 Descriptoin: make lov_create() will not stuck in obd_statfs_rqset()
895 Details : If an OST is down the MDS will hang indefinitely in
896 obd_statfs_rqset() waiting for the statfs data. While for
897 MDS QOS usage of statfs, it should not stuck in waiting.
899 Severity : enhancement
901 Description: remote_acl support
902 Details : Support ACL-based permission check for remote user.
903 Support setfacl/getfacl for remote user with the utils
904 "lfs {l,r}{s,g}etfacl" which follow the same parameter format as
905 the system "{s,g}etfacl" utils.
907 Severity : enhancement
909 Description: Update to RHEL4 U6 kernel-2.6.9-67.EL.
911 Severity : enhancement
913 Description: Update to RHEL5 latest kernel-2.6.18-53.1.4.el5.
917 Description: make mgs_setparam() handle fsname containing dash
918 Details : fsname containing a dash does not work with lctl conf_param
920 Severity : enhancement
922 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.3.
924 Severity : enhancement
926 Description: Update to RHEL5 Update-1 kernel 2.6.18-53.el5.
930 Frequency : rare, at shutdown
931 Description: access already free / zero obd_namespace.
932 Details : if client_disconnect_export was called without force flag set,
933 and exist connect request in flight, this can produce access to
934 NULL pointer (or already free pointer) when connect_interpret
935 store ocd flags in obd_namespace.
939 Frequency : only at startup
940 Description: not alloc memory with spinlock held.
941 Details : allocation memory with GFP_KERNEL can produce sleep deadlock,
942 if any spinlock held.
944 Severity : enhancement
946 Description: make lustre randomly fail allocating memory
947 Details : Make lustre randomly failed allocating memory for testing purpose.
949 Severity : enhancement
951 Description: lost problems with lov objid file
952 Details : Fixes some scability and access to not inited memory problems
953 in work with lov objdid file.
958 Description: lfs find does not continue on file error
959 Details : Continue other files processing when a file/dir is absent.
963 Description: Inconsistent usage of lustre_pack_reply()
964 Details : Standardize the usage of lustre_pack_reply() such that it
965 always generate a CERROR on failure.
968 Frequency : occasional
970 Description: MDS hang or stay in waiting lock
971 Details : If client receive lock with CBPENDING flag ldlm need send lock
972 cancel as separate rpc, to avoid situation when cancel request
973 can't processed due all i/o threads stay in wait lock.
977 Description: Update to RHEL5 kernel 2.6.18-8.1.15.el5.
981 Description: Update to SLES10 SP1 kernel 2.6.16.53-0.16
985 Description: Update to SLES9 kernel-2.6.5-7.287.3.
989 Description: Update to RHEL4 latest kernel.
991 Severity : enhancement
993 Description: Build SLES10 patchless client fails
994 Details : The configure was broken by run ./configure with
995 --with-linux-obj=.... argument for patchless client. When the
996 configure use --with-linux-obj, the LINUXINCLUDE= -Iinclude
997 can't search header adequately. Use absolute path such as
998 -I($LINUX)/include instead.
1002 Description: interrupt oig_wait produce painc on resend.
1003 Details : brw_redo_request can be used for resend requests from ptlrpcd and
1004 private set, and this produce situation when rq_ptlrpcd_data not
1005 copyed to new allocated request and triggered LBUG on assert
1006 req->rq_ptlrpcd_data != NULL. But this member used only for wakeup
1007 ptlrpcd set if request is changed and can be safety changed to use
1012 Description: LASSERT_{REQ,REP}SWAB macros are buggy
1013 Details : If SWAB_PARANOIA is disabled, the LASSERT_REQSWAB and
1014 LASSERT_REPSWAB macros become no-ops, which is incorrect. Drop
1015 these macros and replace them with their difinitions instead.
1019 Description: Update kernel patches for SLES10 2.6.16.53-0.8.
1020 Details : Update which_patch & target file for SLES10 latest kernel.
1023 Description: Remove client patches from SLES 10 kernel.
1024 Details : This causes SLES 10 clients to behave as patchless clients
1025 even on a Lustre-patched (server) kernel.
1027 Severity : enhancement
1029 Description: self-adjustable client's lru lists
1030 Details : use adaptive algorithm for managing client cached locks lru
1031 lists according to current server load, other client's work
1032 pattern, memory activities, etc. Both, server and client
1033 side namespaces provide number of proc tunables for controlling
1036 Severity : enhancement
1038 Description: light-weight GSS support
1039 Details : Support krb5n and krb5a mode, which keep Kerberos 5 authentication
1040 and reduce performance overhead.
1042 Severity : enhancement
1044 Description: Linux keyring support
1045 Details : Support using service of Linux keyring for Lustre GSS internal
1046 context refresh/cache mechanism.
1050 Description: Fix errors in lfs documentation
1051 Details : Fixes man pages
1055 Description: don't use GFP_* in generic Lustre code.
1056 Details : Use cfs_alloc_* functions and CFS_* flags for code portability.
1060 Description: obdclass is limited by single OBD_ALLOC(idarray)
1061 Details : replace OBD_ALLOC/OBD_FREE with OBD_VMALLOC/OBD_VFREE
1065 Description: warnings with build patchless client with vanila 2.6.19 and up
1066 Details : change old ctl_table style and replace ctl_table/ctl_table_header
1067 with cfs_sysctl_table_t/cfs_sysctl_table_header_t
1071 Frequency : Only for SLES
1072 Description: sanity_quota fail test_1
1073 Details : There are multiple occurences of $TSTUSR in SLES's /etc/group
1074 file, which makes TSTID[2] inunique.
1078 Frequency : Only for SLES9
1079 Description: Kernel patches for SLES9 2.6.5-7.286 kernel
1080 Details : Update target/ChangeLog/which_patch .
1084 Frequency : Only for test_10 in sanity-quota.sh
1085 Description: a bug in quota test code
1086 Details : checking a test flag in a wrong place
1090 Frequency : Only for test_13 in sanity-quota.sh
1091 Description: a bug in quota test script
1092 Details : shouldn't check overfull allocated quota
1096 Description: jbd statistics
1097 Details : Port older jbd statistics patch for sles10
1099 Severity : enhancement
1101 Description: Add printing inode info into message about error in writepage.
1106 Description: reply_lock_interpret crash due to race with it and lock cancel.
1107 Details : Do not replay locks that are being cancelled. Do not reference
1108 locks by their address during replay, just by their handle.
1112 Frequency : When flocks are used.
1113 Description: assertion failure in ldlm_cli_enquque_fini for non NULL lock.
1114 Details : Flock locks might destroy just granted lock if it could be merged
1115 with another existing flock, this is done in completion handler,
1116 so teach ldlm_cli_enquque_fini that this is a valid case for
1122 Description: Oops in read and write path when failing to allocate lock.
1123 Details : Check if lock allocation failed and return error back.
1127 Description: lstripe command fails for valid OST index
1128 Details : The stripe offset is compared to lov->desc.ld_tgt_count
1129 instead of lov->desc.ld_active_tgt_count.
1133 Description: sanity.sh failed test 103
1134 Details : RHEL mis-interpret setfacl "-X" param, so we won't test setfacl
1139 Description: df doesn't work properly if diskfs blocksize != 4K
1140 Details : Choose biggest blocksize of OST's as the LOV's blocksize.
1144 Description: lfs find on -1 stripe looping in lsm_lmm_verify_common()
1145 Details : Avoid lov_verify_lmm_common() on directory with -1 stripe count.
1147 Severity : enhancement
1149 Description: Adaptive timeouts
1150 Details : RPC timeouts adapt to changing server load and network
1151 conditions to reduce resend attempts and improve recovery time.
1155 Description: llapi_file_create() does not allow some changes
1156 Details : add llapi_file_open() that allows specifying the mode and
1157 open flags, and also returns an open file handle.
1161 Description: merge and cleanup kernel patches.
1162 Details : Remove mnt_lustre_list in vfs_intent-2.6-rhel4.patch.
1166 Description: Add journal checksum support.(Kernel part)
1167 Details : The journal checksum feature adds two new flags i.e
1168 JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT and
1169 JBD2_FEATURE_COMPAT_CHECKSUM. JBD2_FEATURE_CHECKSUM flag
1170 indicates that the commit block contains the checksum for
1171 the blocks described by the descriptor blocks. Now commit
1172 record can be sent to disk without waiting for descriptor
1173 blocks to be written to disk. This behavior is controlled
1174 using JBD2_FEATURE_ASYNC_COMMIT flag.
1178 Description: OSS needs mutliple precreate threads
1179 Details : Add ability to start more than one create thread per OSS.
1183 Description: Sanity.sh test_65k failed: setstripe should have succeeded
1184 Details : execute lfs setstripe on client
1188 Description: mds_obd_create error creating tmp object
1189 Details : When the user sets quota on root, llog will be affected and can't
1190 create files and write files.
1193 Frequency : Always on ia64 patchless client, and possibly others.
1195 Description: Add EXPORT_SYMBOL check for node_to_cpumask symbol.
1196 Details : This allows the patchless client to be loaded on architectures
1197 without this export.
1201 Description: RedHat Update kernel for RHEL5
1202 Details : Add the kernel config file for RHEL5.
1206 Description: RedHat Update kernel for RHEL5
1207 Details : Modify the kernel config file more closer RHEL5.
1211 Description: Build failure against Centos5 (RHEL5)
1212 Details : Define PAGE_SIZE when it isn't present.
1216 Description: test_7 on sanity-quota.sh can't be used on separate machines
1217 Details : add facet to handle it
1221 Description: "ll_intent_file_open()) lock enqueue: err: -13" with nfs
1222 Details : with NFS, the anon dentry's parent was set to itself in
1223 d_alloc_anon(), so in MDS, we use rec->ur_fid1 to find the
1224 corresponding dentry other than use rec->ur_name.
1226 Severity : enhancement
1228 Description: lfs setstripe enhancement
1229 Details : Make lfs setstripe understand 'k', 'm' and 'g' for stripe size.
1233 Description: enable data checksumming by default
1234 Details : enable checksum by default, allow --disable-checksum
1235 configure option and "-o nochecksum" mount option. Checksums
1236 can also be disabled at runtime via $LPROC/osc/*/checksum_pages.
1240 Description: lustre support for RHEL5
1241 Details : Add support for RHEL5.
1245 Description: Client eviction due to failover config
1246 Details : after a connection loss, the lustre client should attempt to
1247 reconnect to the last active server first before trying the
1248 other potential connections.
1252 Description: when mds and osts use different quota unit(32bit and 64bit),
1253 quota will be released repeatly.
1254 Details : void sending multiple quota reqs to mds, which will keep the
1255 status between the reqs.
1259 Description: osts not allocated evenly to files
1260 Details : change the condition to increase offset_idx
1263 Frequency : Always for filesystems larger than 2TB on 32-bit systems.
1265 Description: Data corruption for OSTs that are formatted larger than 2TB
1267 Details : When generating the bio request for lustre file writes the
1268 sector number would overflow a temporary variable before being
1269 used for the IO. The data reads correctly from Lustre (which
1270 will overflow in a similar manner) but other file data or
1271 filesystem metadata may be corrupted in some cases.
1275 Description: Tune the kernel for good SCSI performance.
1276 Details : Set the value of /sys/block/{dev}/queue/max_sectors_kb
1277 to the value of /sys/block/{dev}/queue/max_hw_sectors_kb
1282 Description: rewrite ext2-derived code in obdclass/uuid.c
1283 Details : rewrite inherited code (uuid parsing code from ext2 utils)
1284 from scratch preserving functionality.
1288 Description: "lfs find -obd UUID" prints directories
1289 Details : "lfs find -obd UUID" will return all directory names instead
1290 of just file names. It is incorrect because the directories
1291 do not reside on the OSTs.
1295 Description: use i_size_read and i_size_write in 2.6 port
1296 Details : replace inode->i_size access with i_size_read/write()
1299 Frequency : when removing large files
1301 Description: scheduling issue during removal of large Lustre files
1302 Details : Don't take the BKL in fsfilt_ext3_setattr() for 2.6 kernels.
1303 It causes scheduling issues when removing large files (17TB in the
1307 Frequency : only with liblustre clients on XT3
1309 Description: evictions taking too long
1310 Details : allow llrd to evict clients directly on OSTs
1313 Frequency : only on ppc
1315 Description: /proc/fs/lustre/devices broken on ppc
1316 Details : The patch as applied to 1.6.2 doesn't look correct for all arches.
1317 We should make sure the type of 'index' is loff_t and then cast
1318 explicitly as needed below. Do not assign an explicitly cast
1322 Frequency : only for rhel5
1324 Description: Kernel patches update for RHEL5 2.6.18-8.1.10.el5.
1325 Details : Modify the target file & which_kernel.
1327 Severity : enhancement
1329 Description: omit set fsid for export NFS
1330 Details : fix set/restore device id for avoid EMFILE error and mark lustre
1331 fs as FS_REQUIRES_DEV for avoid problems with generate fsid.
1335 Frequency : Always, for kernels after 2.6.16
1336 Description: Fix warning idr_remove called for id=.. which is not allocated.
1337 Details : Last kernels save old s_dev before kill super and not allow
1338 to restore from callback - restore it before call kill_anon_super.
1342 Description: buffer overruns could theoretically occur
1343 Details : llapi_semantic_traverse() modifies the "path" argument by
1344 appending values to the end of the origin string, and a
1345 overrun may occur. Adding buffer overrun check in liblustreapi.
1349 Description: Fix error on 'ls .' at the top of the Lustre mount.
1350 Details : Don't revalidate dentry if it is a root dentry.
1354 Description: Kernel patches update for RHEL4 2.6.9-55.0.6.
1355 Details : Modify vm-tunables-rhel4.patch.
1359 Description: Kernel config for 2.6.18-vanilla.
1360 Details : Modify targets/2.6-vanilla.target.in.
1361 Add config file kernel-2.6.18-2.6-vanilla-i686.config.
1362 Add config file kernel-2.6.18-2.6-vanilla-i686-smp.config.
1363 Add config file kernel-2.6.18-2.6-vanilla-x86_64.config.
1364 Add config file kernel-2.6.18-2.6-vanilla-x86_64-smp.config.
1368 Description: improve handling recoverable errors
1369 Details : if request processig with error which can be recoverable on server
1370 request should be resend, otherwise page released from cache and
1375 Description: Kernel patches update for RHEL5 2.6.18-8.1.14.el5.
1376 Details : Modify target file & which_patch.
1377 A flaw was found in the IA32 system call emulation provided
1378 on AMD64 and Intel 64 platforms. An improperly validated 64-bit
1379 value could be stored in the %RAX register, which could trigger an
1380 out-of-bounds system call table access. An untrusted local user
1381 could exploit this flaw to run code in the kernel
1382 (ie a root privilege escalation). (CVE-2007-4573).
1386 Description: O_DIRECT bypasses client statistics.
1387 Details : When running with O_DIRECT I/O, neither the client rpc_stats nor
1388 read_ahead_stats were updated. copied the stats section from
1389 osc_send_oap_rpc() into async_internal().
1393 Description: Add jbd statistics patch for RHEL5 and 2.6.18-vanilla
1397 Description: change order of libsysio includes
1398 Details : '#include sysio.h' should always come before '#include xtio.h'
1402 Description: handle "serious error: objid * already exists" more gracefully
1403 Details : If LAST_ID value on disk is smaller than the objects existing in
1404 the O/0/d* directories, it indicates disk corruption and causes an
1405 LBUG(). If the object is 0-length, then we should use the existing
1406 object. This will help to avoid a full fsck in most cases.
1408 Severity : enhancement
1410 Description: adapt the lustre_config script to support the upgrade case
1411 Details : Add "-u" option for lustre_config script to support upgrading 1.4
1412 server targets to 1.6 in parallel.
1416 Description: To avoid grant space > avaible space when the disk is almost
1417 full. Without this patch you might see the error "grant XXXX >
1418 available" or some LBUG about grant, when the disk is almost
1420 Details : In filter_check_grant, for non_grant cache write, we should
1421 check the left space by if (*left > ungranted + bytes), instead
1422 of (*left > ungranted), because only we are sure the left space
1423 is enough for another "bytes", then the ungrant space should be
1424 increase. In client, we should update cl_avail_grant only there
1425 is OBD_MD_FLGRANT in the reply.
1429 Description: Update RHEL 4 kernel to fix local root privilege escalation.
1430 Details : Update to the latest RHEL 4 kernel to fix the vulnerability
1431 described in CVE-2007-4573. This problem could allow untrusted
1432 local users to gain root access.
1435 Frequency : when using O_DIRECT and quotas
1437 Description: Incorrect file ownership on O_DIRECT output files
1438 Details : block usage reported by 'lfs quota' does not take into account
1439 files that have been written with O_DIRECT.
1444 Description: touch file failed when fs is not full
1445 Details : OST in recovery should not be discarded by MDS in alloc_qos(),
1446 otherwise we can get ENOSP while fs is not full.
1450 Description: parallel lock callbacks
1451 Details : Instead of sending blocking and completion callbacks as separated
1452 requests, adding them to a set and sending in parallel.
1455 Frequency : only for Cray XT3
1456 Bugzilla : 12829/13455
1457 Description: Changing primary group doesn't change the group lustre assigns to
1459 Details : When CRAY_XT3 is defined, the fsgid supplied by the client is
1460 overridden with the primary group provided by the group upcall,
1461 whereas the supplied fsgid can be trusted if it is in the list of
1462 supplementary groups returned by the group upcall.
1464 Severity : enhancement
1466 Description: Allow masking D_WARNING, D_ERROR messages from console
1467 Details : Console messages can now be disabled via lnet.printk.
1471 Description: User code with malformed file open parameter crashes client node
1472 Details : Before packing join_file req, all the related reference should be
1473 checked carefully in case some malformed flags cause fake
1474 join_file req on client.
1478 Description: LDLM_ENQUEUE races with LDLM_CP_CALLBACK
1479 Details : ldlm_completion_ast() assumes that a lock is granted when the req
1480 mode is equal to the granted mode. However, it should also check
1481 that LDLM_FL_CP_REQD is not set.
1485 Description: Heavy nfs access might result in deadlocks
1486 Details : After ELC code landed, it is now improper to enqueue any mds
1487 locks under och_sem, because enqueue might want to decide to
1488 cancel open locks for same inode we are holding och_sem for.
1492 Description: Client eviction while running blogbench
1493 Details : A lot of unlink operations with concurrent I/O can lead to a
1494 deadlock causing evictions. To address the problem, the number of
1495 oustanding OST_DESTROY requests is now throttled to
1496 max_rpcs_in_flight per OSC and LDLM_FL_DISCARD_DATA blocking
1497 callbacks are processed in priority.
1501 Description: enable ACLs on MDS by default
1502 Details : ACLs must be enabled on MDS by default.
1505 Frequency : PPC/PPC64 only
1507 Description: conflicts between asm-ppc64/types.h and lustre_types.h
1508 Details : fix duplicated definitions between asm-ppc64/types.h and
1509 lustre_types.h on PPC.
1512 Frequency : PPC/PPC64 only
1514 Description: asm-ppc/segment.h does not exist
1515 Details : fix compile issue on PPC.
1519 Description: better handle error messages in extents code
1522 Frequency : RHEL4 only
1524 Description: mkfs is very slow on IA64/RHEL4
1525 Details : A performance regression has been discovered in the MPT Fusion
1526 driver between versions 3.02.73rh and 3.02.99.00rh. As a
1527 consequence, we have downgraded the MPT Fusion driver in the RHEL4
1528 kernel from 3.02.99.00 to 3.02.73 until this problem is fixed.
1530 Severity : enhancement
1532 Description: SNMP support enhancement
1533 Details : Adding total number of sampled request for an MDS node in snmp
1536 Severity : enhancement
1538 Description: Optimize ldlm waiting list processing for PR extent locks
1539 Details : When processing waiting list for read extent lock and meeting read
1540 lock that is same or wider to it that is not contended, skip
1541 processing rest of the list and immediatelly return current
1542 status of conflictness, since we are guaranteed there are no
1543 conflicting locks in the rest of the list.
1547 Description: Time out and refuse to reconnect
1548 Details : When the failover node is the primary node, it is possible
1549 to have two identical connections in imp_conn_list. We must
1550 compare not conn's pointers but NIDs, otherwise we can defeat
1551 connection throttling.
1555 Description: port llog fixes from b1_6 into HEAD
1556 Details : Port llog reference couting and some llog cleanups from b1_6
1557 (bug 10800) into HEAD, for protect from panic and access to already
1558 free llog structures.
1562 Description: Detect stride IO mode in read-ahead
1563 Details : When a client does stride read, read-ahead should detect that and
1564 read-ahead pages according to the detected stride pattern.
1568 Description: data checksumming impacts single node performance
1569 Details : add support for several checksum algorithm. Currently, only CRC32
1570 and Adler-32 are supported. The checksum type can be changed on
1571 the fly via /proc/fs/lustre/osc/*/checksum_type.
1575 Description: use adler32 for page checksums
1576 Details : when available, use the Adler-32 algorithm instead of CRC32 for
1581 Description: build for x2 fails
1582 Details : fix compile issue on Cray systems.
1586 Description: Properly match for duplicate locks
1587 Details : Due to different lock order from skiplists code, we need to
1588 traverse entire list for now
1591 Frequency : only on PPC/SLES10
1593 Description: "BITS_PER_LONG is not 32 or 64" in linux/idr.h
1594 Details : On SLES10/PPC, fs.h includes idr.h which requires BITS_PER_LONG to
1595 be defined. Add a hack in mkfs_lustre.c to work around this compile
1600 Description: LASSERT on MDS when client holding flock lock dies
1601 Details : ldlm pool logic depends on number of granted locks equal to
1602 number of released locks which is not true for flock locks, so
1603 just exclude such locks from consideration.
1607 Description: MDS deadlock with many ll_sync_lov threads and I/O stalled
1608 Details : Use fsfilt_sync() for both the whole filesystem sync and
1609 individual file sync to eliminate dangerous inode locking
1610 with I_LOCK that can lead to a deadlock.
1614 Description: performance in 1.6.3
1615 Details : Force q->max_phys_segments to MAX_PHYS_SEGMENTS on SLES10 to be
1616 sure that 1MB requests are not fragmented by the block layer.
1618 Severity : enhancement
1620 Description: organize the server-side client stats on per-nid basis
1621 Details : Change the structure of stats under obdfilter and mds to
1629 The "uuid"s file would list the uuids of _active_ exports.
1630 And the clear entry is to clear all stats and stale nids.
1632 Severity : enhancement
1634 Description: eliminate client locks in face of contention
1635 Details : file contention detection and lockless i/o implementation
1636 for contended files.
1640 Description: Reinitialize optind to 0 so that interactive lfs works in all cases
1643 Frequency : very rare, if additional xattrs are used on kernels >= 2.6.12
1645 Description: MDS may lose file striping (and hence file data) in some cases
1646 Details : If there are additional extended attributes stored on the MDS,
1647 in particular ACLs, SELinux, or user attributes (if user_xattr
1648 is specified for the client mount options) then there is a risk
1649 of attribute loss. Additionally, the Lustre file striping
1650 needs to be larger than default (e.g. striped over all OSTs),
1651 and an additional attribute must be stored initially in the
1652 inode and then increase in size enough to be moved to the
1653 external attribute block (e.g. ACL growing in size) for file
1658 Description: skiplist implementation simplification
1659 Details : skiplists are used to group compatible locks on granted list
1660 that was implemented as tracking first and last lock of each
1661 lock group the patch changes that to using doubly linked lists
1665 Description: MDS LBUG: ASSERTION(!IS_ERR(dchild))
1666 Details : Change LASSERTs to client eviction (i.e. abort client's recovery)
1667 because LASSERT on both the data supplied by a client, and the
1668 data on disk is dangerous and incorrect.
1670 Severity : enhancement
1672 Description: Slow truncate/writes to huge files at high offsets.
1673 Details : Directly associate cached pages to lock that protect those pages,
1674 this allows us to quickly find what pages to write and remove
1675 once lock callback is received.
1679 Description: more ldlm soft lockups
1680 Details : In ldlm_resource_add_lock(), call to ldlm_resource_dump()
1681 starve other threads from the resource lock for a long time in
1682 case of long waiting queue, so change the debug level from
1683 D_OTHER to the less frequently used D_INFO.
1685 Severity : enhancement
1687 Description: add -gid, -group, -uid, -user options to lfs find
1691 Description: Hung threads in invalidate_inode_pages2_range
1692 Details : The direct IO path doesn't call check_rpcs to submit a new RPC
1693 once one is completed. As a result, some RPCs are stuck in the
1694 queue and are never sent.
1698 Description: filter threads hungs on waiting journal commit
1699 Details : Cleanup filter group llog code, then only filter group llog will
1700 be only created in the MDS/OST syncing process.
1704 Description: Procfs and llog threads access destoryed import sometimes.
1705 Details : Sync the import destoryed process with procfs and llog threads by
1706 the import refcount and semaphore.
1708 Severity : enhancement
1710 Description: openlock cache of b1_6 port to HEAD
1715 Description: kernel BUG at ldiskfs2_ext_new_extent_cb
1716 Details : If insertion of an extent fails, then discard the inode
1717 preallocation and free data blocks else it can lead to duplicate
1722 Description: don't always update ctime in ext3_xattr_set_handle()
1723 Details : Current xattr code updates inode ctime in ext3_xattr_set_handle.
1724 In some cases the ctime should not be updated, for example for
1725 2.0->1.8 compatibility it is necessary to delete an xattr and it
1726 should not update the ctime.
1730 Bugzilla : 15713/16362
1731 Description: Assertion in iopen_connect_dentry in 1.6.3
1732 Details : looking up an inode via iopen with the wrong generation number can
1733 populate the dcache with a disconneced dentry while the inode
1734 number is in the process of being reallocated. This causes an
1735 assertion failure in iopen since the inode's dentry list contains
1736 both a connected and disconnected dentry.
1740 Description: assertion failure in ldlm_handle2lock()
1741 Details : fix a race between class_handle_unhash() and class_handle2object()
1742 introduced in lustre 1.6.5 by bug 13622.
1747 Description: Kernel BUG: sd_iostats_bump: unexpected disk index
1748 Details : remove the limit of 256 scsi disks in the sd_iostat patch
1753 Description: oops in sd_iostats_seq_show()
1754 Details : unloading/reloading the scsi low level driver triggers a kernel
1755 bug when trying to access the sd iostat file.
1760 Description: Kernel panics during QLogic driver reload
1761 Details : REQ_BLOCK_PC requests are not handled properly in the sd iostat
1762 patch, causing memory corruption.
1767 Description: journal_dev option does not work in b1_6
1768 Details : pass mount option during pre-mount.
1770 Severity : enhancement
1772 Description: Add a FIEMAP(FIle Extent MAP) ioctl
1773 Details : FIEMAP ioctl will allow an application to efficiently fetch the
1774 extent information of a file. It can be used to map logical blocks
1775 in a file to physical blocks in the block device.
1779 Description: LDLM soft lockups - improvement
1780 Details : It is be possible to send the lock handle along with each read
1781 or write request because the client is already doing a lock match
1782 itself so there isn't any reason the OST should have to re-do that
1786 Frequency : only on Cray X2
1788 Description: X2 build failures
1789 Details : fix build failures on Cray X2.
1793 Description: xid & resent requests
1794 Details : Initialize RPC XID from clock at startup (randomly if clock is
1797 Severity : enhancement
1799 Description: Add lustre_start utility to start or stop multiple Lustre servers
1804 Description: Lustre GPF in {:ptlrpc:ptlrpc_server_free_request+373}
1805 Details : In case of memory pressure, list_del() can be called twice on
1806 req->rq_history_list, causing a kernel oops.
1810 Description: kptllnd_peer_check_sends()) ASSERTION(!in_interrupt()) failed
1811 Details : fix stack overflow in the distributed lock manager by defering
1812 export eviction after a failed AST to the elt thread instead of
1813 handling it in the dlm interpret routine.
1817 Description: Convert some comments to new format.
1818 Details : Update documenting comments to match doxygen conventions.
1822 Description: Grammar fixes.
1823 Details : A couple of trivial spelling fixes.
1827 Description: OSD_COUNTERS-mandatory
1828 Details : Make previously optional ->oti_{w,r}_locks sanity checks mandatory
1829 to simplify the code.
1833 Description: simplify cmm_device freeing logic.
1834 Details : Call cmm_device_free() in the failure path of cmm_device_alloc().
1838 Description: Add lockdep support to dt_object_operations locking interface.
1839 Details : Augment ->do_{read,write}_lock() prototypes with a `role'
1840 parameter indicating lock ordering. Update mdd code to use new
1845 Description: Introduce failloc constants for lockless IO tests.
1846 Details : Add two new failloc constants to test lockless IO. Only one of
1847 them in implemented---another is checked in yet to be landed
1852 Description: Add lockdep support for inode mutex.
1853 Details : Introduce and use new LOCK_INODE_MUTEX_PARENT() macro to be used
1854 in the situations where Lustre has to lock more than one inode
1859 Description: Add optional invariants checking support.
1860 Details : Add new LINVRNT() macro, optional on new --enable-invariants
1861 configure switch. This macro is to be used for consistency and
1862 sanity checks that are too expensive to be left in `production'
1867 Description: Zap lock->l_granted_mode with explicit LCK_MINMODE.
1868 Details : Use LCK_MINMODE rather than 0 to reset lock->l_granted_mode to
1873 Description: Add lockdep support for ldlm_lock and ldlm_resource.
1874 Details : Use spin_lock_nested() in (the only) situation where more than
1875 one ldlm_lock is locked simultaneously. Also, fix possible
1876 dead-lock in ldlm_lock_change_resource() by enforcing particular
1881 Description: Use struct ldlm_callback_suite in ldlm_lock_create().
1882 Details : Instead of specifying each ldlm_lock call-back through separate
1883 parameter, wrap them into struct ldlm_callback_suite.
1887 Description: Kill join_lru obd method and its callers.
1888 Details : CLIO uses lock weighting policy to keep locks over mmapped regions
1889 in memory---a requirement implemented through ->o_join_lru() obd
1890 method in HEAD. Remove this method and its users.
1894 Description: Add asynchronous ldlm ENQUEUE completion handler.
1895 Details : CLIO posts ENQUEUE requests asynchronously through ptlrpcd---a
1896 case that stock ldlm_completion_ast() cannot handle as it waits
1897 until lock is granted. Introduce new ldlm_completion_ast_async()
1898 for this. Also comment ldlm_completion_ast().
1902 Description: ldlm_error <-> errno conversion.
1903 Details : Add functions to map (rather arbitrary) between LDLM error codes
1904 and standard errno values. CLIO needs this to prevent LDLM specific
1905 constants from escaping ldlm and osc.
1909 Description: Kill unused ldlm_handle2lock_ns() function.
1910 Details : Kill unused ldlm_handle2lock_ns() function.
1914 Description: Add lu_ref support to ldlm_lock
1915 Details : lu_ref support for ldlm_lock and ldlm_resource. See lu_ref patch.
1916 lu_ref fields ->l_reference and ->lr_reference are added to ldlm_lock
1917 and ldlm_resource. LDLM interface has to be changed, because code that
1918 releases a reference on a lock, has to "know" what reference this is.
1919 In the most frequent case
1921 lock = ldlm_handle2lock(handle);
1923 LDLM_LOCK_PUT(lock);
1925 no changes are required. When any other reference (received _not_ from
1926 ldlm_handle2lock()) is released, LDLM_LOCK_RELEASE() has to be called
1927 instead of LDLM_LOCK_PUT().
1929 Arguably, changes are pervasive, and interface requires some discipline
1930 for proper use. On the other hand, it was very instrumental in finding
1931 a few leaked lock references.
1935 Description: Add ldlm_lock_addref_try().
1936 Details : Introduce ldlm_lock_addref_try() function (used by CLIO) that
1937 attempts to addref a lock that might be being canceled
1942 Description: Add ldlm_weigh_callback().
1943 Details : Add new ->l_weigh_ast() call-back to ldlm_lock. It is called
1944 by ldlm_cancel_shrink_policy() to estimate lock "value", instead of
1945 hard-coded `number of pages' logic.
1949 Description: Add lockdep annotations to llog code.
1950 Details : Use appropriately tagged _nested() locking calls in the places
1951 where llog takes more than one ->lgh_lock lock.
1955 Description: Add loi_kms_set().
1956 Details : Wrap kms updates into a helper function.
1960 Description: Constify instances of struct lsm_operations.
1961 Details : Constify instances of struct lsm_operations.
1965 Description: lu_conf support.
1966 Details : On a server, a file system object is uniquely identified
1967 by a fid, which is sufficient to locate and load all object
1968 state (inode). On a client, on the other hand, more data are
1969 necessary instantiate an object. Change lu_object_find() and
1970 friends to take additional `lu_conf' argument describing object.
1971 Typically this includes layout information.
1975 Description: lu_context fixes.
1976 Details : Introduce new lu_context functions that are needed on the client
1977 side, where some system threads (ptlrpcd) are shared by multiple
1978 modules, and so cannot be stopped during module shutdown.
1982 Description: Add start and stop methods to lu_device_type_operations.
1983 Details : Introduce two new methods in lu_device_type_operations, that are
1984 invoked when first instance of a given type is created and last one
1985 is destroyed respectively. This is need by CLIO.
1989 Description: Add lu_ref support to struct lu_device.
1990 Details : Add lu_ref support to lu_object and lu_device. lu_ref is used to
1991 track leaked references.
1995 Description: Introduce lu_kmem_descr.
1996 Details : lu_kmem_descr and its companion interface allow to create
1997 and destroy a number of kmem caches at once.
2001 Description: Fix lu_object finalization race.
2002 Details : Fix a race between lu_object_find() finding an object and its
2003 concurrent finalization. This race is (most likely) not possible
2004 on the server, but might happen on the client.
2008 Description: Introduce lu_ref interface.
2009 Details : lu_ref is a debugging module allowing to track references to
2010 a given object. It is quite cpu expensive, and has to be
2011 explicitly enabled with --enable-lu_ref. See usage description
2016 Description: Factor lu_site procfs stats into a separate function.
2017 Details : Separate lu_site stats printing code into a separate function
2018 that can be reused on a client.
2022 Description: Constify instances of struct {lu,dt,md}_device_operations.
2023 Details : Constify instances of struct {lu,dt,md}_device_operations.
2027 Description: Introduce struct md_site and move meta-data specific parts of
2028 struct lu_site here.
2029 Details : Move md-specific fields out of struct lu_site into special struct
2030 md_site, so that lu_site can be used on a client.
2034 Description: Kill mdd_lov_destroy().
2035 Details : Remove unused mdd code.
2039 Description: Add st_block checking to multistat.c.
2040 Details : Add st_block checking to multistat.c.
2044 Description: Add lu_ref support to struct obd_device.
2045 Details : Add lu_ref tracking to obd_device.
2049 Description: Kill obd_set_fail_loc().
2050 Details : Remove unused code.
2054 Description: Add special type for ptlrpc_request interpret functions.
2055 Details : Add lu_env parameter to ->rq_interpreter call-back. NULL is passed
2056 there. Actual usage will be in CLIO.
2060 Description: Replace RW_LOCK_UNLOCKED() macro with rwlock_init().
2061 Details : Replace RW_LOCK_UNLOCKED() with rwlock_init() as the former
2062 doesn't work with lockdep.
2066 Description: Add rwv.c test program.
2067 Details : New testing program exercising readv(2) and writev(2) (Qian).
2071 Description: Add sendfile.c test program.
2072 Details : New testing program exercising sendfile(2) (Jay).
2076 Description: Ratelimit a message that can be very frequent.
2077 Details : Ratelimit a memory allocation failure message that can
2082 Description: Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
2083 Details : Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
2087 Description: Liblustre build fixes.
2088 Details : Liblustre build fixes.
2092 Description: libcfs: add cfs_{need,cond}_resched() interface.
2093 Details : libcfs: add cfs_{need,cond}_resched() definition and
2094 implementations for Linux, NT, and liblustre.
2096 Severity : enhancement
2098 Description: More exported tunables for mballoc
2099 Details : Add support for tunable preallocation window and new tunables for
2100 large/small requests
2104 Description: Detect corruption of block bitmap and checking for preallocations
2105 Details : Checks validity of on-disk block bitmap. Also it does better
2106 checking of number of applied preallocations. When corruption is
2107 found, it turns filesystem readonly to prevent further corruptions.
2111 Description: (rw.c:1323:ll_read_ahead_pages()) ASSERTION(page_idx > ria->ria_stoff) failed
2112 Details : Once the unmatched stride IO mode is detected, shrink the stride-ahead
2113 window to 0. If it does hit cache miss, and read-pattern is still
2114 stride-io mode, does not reset the stride window, but also does not
2115 increase the stride window length in this case.
2119 Frequency : only for big-endian servers
2120 Description: Check if system is big-endian while mounting fs with extents feature
2121 Details : Mounting a filesystem with extents feature will fail on big-endian
2122 systems since ext3-based ldiskfs is not supported on big-endian
2123 systems. This can be over-riden with "bigendian_extents" mount option.
2125 Severity : enhancement
2127 Description: The root squash functionality
2128 Details : A security feature, which is to prevent users from being able
2129 to mount lustre on their desktop, run as root, and delete
2130 all of the files in the filesystem. The goal is accomplished by
2131 remapping user id (UID) and group id (GID) of the root user to
2132 a UID and GID specified by the system administartor via Lustre
2133 configuration management server (MGS). The functionality also
2134 allows to specify sets of clients for which the remapping does
2139 Description: Excessive recovery window
2140 Details : With AT enabled, the recovery window can be excessively long (6000+
2141 seconds). To address this problem, we no longer use
2142 OBD_RECOVERY_FACTOR when extending the recovery window (the connect
2143 timeout no longer depends on the service time, it is set to
2144 INITIAL_CONNECT_TIMEOUT now) and clients report the old service
2145 time via pb_service_time.
2149 Description: Watchdog triggered on MDS failover
2150 Details : enable OBD_CONNECT_MDT flag when connecting from the MDS so that
2151 the OSTs know that the MDS "UUID" can be reused for the same export
2152 from a different NID, so we do not need to wait for the export to be
2156 Frequency : rare, only if using MMP with Linux RAID
2158 Description: MMP doesn't work with Linux RAID
2159 Details : While using HA for Lustre servers with Linux RAID, it is possible
2160 that MMP will not detect multiple mounts. To make this work we
2161 need to unplug the device queue in RAID when the MMP block is being
2162 written. Also while reading the MMP block, we should read it from
2163 disk and not the cached one.
2165 Severity : enhancement
2167 Description: open file using fid
2168 Details : A file can be opened using just its fid, like
2169 <mntpt>/.lustre/fid/SEQ:OID:VER - this is needed for HSM and replication
2172 Frequency : Only in RHEL5 when mounting multiple ext3 filesystems
2175 Description: "kmem_cache_create: duplicate cache jbd_4k" error message
2176 Details : add proper locking for creation of jbd_4k slab cache
2180 Description: MMP check in ext3_remount() fails without displaying any error
2181 Details : When multiple mount protection fails during remount, proper error
2184 Severity : enhancement
2186 Description: Allow stripe size to be up to 4G-64k
2187 Details : Fix math logic to allow large stripe sizes.
2191 Description: add check for >8TB ldiskfs filesystems
2192 Details : ext3-based ldiskfs does not support greater than 8TB LUNs.
2193 Don't allow >8TB ldiskfs filesystems to be mounted without
2194 force_over_8tb mount option
2199 Descriptoin: OSS complains "too many blocks" and crashes
2200 Details : In filter_commitrw_write():
2201 1. lquota_chkquota();
2202 2. obdo_from_inode();
2203 3. lquota_pending_commit();
2204 "2." has changed uid and gid in this case, then caused
2205 lquota_chkquota() and lquota_pending_commit() to use
2206 different uid and gid. That is the root of the bug.
2210 Descriptoin: getstripe doesn't report stripe count = -1 sanely
2211 Details : Detect and print out the 'stripe count = -1' case properly.
2213 --------------------------------------------------------------------------------
2215 2007-08-10 Cluster File Systems, Inc. <info@clusterfs.com>
2217 * Support for kernels:
2218 2.6.5-7.283 (SLES 9),
2219 2.6.9-55.EL (RHEL 4),
2220 2.6.16.46-0.14 (SLES 10),
2221 2.6.18.8 vanilla (kernel.org)
2222 * Client support for unpatched kernels:
2223 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
2224 2.6.16 - 2.6.22 vanilla (kernel.org)
2225 * Due to recently discovered recovery problems, we do not recommend
2226 using patchless RHEL 4 clients with this or any earlier release.
2227 * Recommended e2fsprogs version: 1.39.cfs8
2228 * Note that reiserfs quotas are disabled on SLES 10 in this kernel.
2229 * Starting with this release, the ldiskfs backing filesystem required
2230 by Lustre is now in its own package, lustre-ldiskfs. This package
2231 should be installed. It is versioned separately from Lustre and
2232 may be released separately in future.
2236 Description: block reactivating mgc import until all deactivates complete
2237 Details : Fix race when failing back MDT/MGS to itself (testing)
2240 Frequency : at statup only
2242 Description: mds_lov_synchronize race leads to various problems
2243 Details : simultaneous MDT->OST connections at startup can cause the
2244 sync to abort, leaving the OSC in a bad state.
2246 Severity : enhancement
2248 Description: add optional extra BUILD_VERSION info
2249 Details : add a new environment variable (namely LUSTRE_VERS) which allows
2250 to override the lustre version.
2253 Frequency : 2.6.18 servers only
2255 Description: ll_kern_mount() doesn't release the module reference
2256 Details : The ldiskfs module reference count never drops down to 0
2257 because ll_kern_mount() doesn't release the module reference.
2262 Description: server LBUG when using old ost_num_threads parameter
2263 Details : Accept the old ost_num_threads parameter but warn that it
2264 is deprecated, and fix an off-by-one error that caused an LBUG.
2269 Description: Transient SCSI error results in persistent IO issue
2270 Details : iobuf->dr_error is not reinitialized to 0 between two
2274 Frequency : sometimes when underlying device returns I/O errors
2276 Description: OSTs not going read-only during write failures
2277 Details : OSTs are not remounted read-only when the journal commit threads
2278 get I/O errors because fsfilt_ext3 calls journal_start/stop()
2279 instead of the ext3 wrappers.
2283 Description: poor connect scaling with increasing client count
2284 Details : Don't run filter_grant_sanity_check for more than 100 exports
2285 to improve scaling for large numbers of clients.
2288 Frequency : SLES10 only
2290 Description: sanity-quota.sh quotacheck failed: rc = -22
2291 Details : Quotas cannot be enabled on SLES10.
2294 Frequency : liblustre clients only
2296 Description: getdirentries does not give error when run on compute nodes
2297 Details : getdirentries does not fail when the size specified as an argument
2298 is too small to contain at least one entry
2300 Severity : enhancement
2302 Description: Add LNET router traceability for debug purposes
2303 Details : If a checksum failure occurs with a router as part of the
2304 IO path, the NID of the last router that forwarded the bulk data
2305 is printed so it can be identified.
2310 Description: OST "spontaneously" evicts client; client has imp_pingable == 0
2311 Details : Due to a race condition, liblustre clients were occasionally
2312 evicted incorrectly.
2314 Severity : enhancement
2316 Description: lfs setstripe use optional parameters instead of postional
2319 Severity : enhancement
2321 Description: Nanosecond timestamp support for ldiskfs
2322 Details : The on-disk ldiskfs filesystem has added support for nanosecond
2323 resolution timestamps. There is not yet support for this at
2324 the Lustre filesystem level.
2327 Frequency : during server recovery
2329 Description: MDS failing to send precreate requests due to OSCC_FLAG_RECOVERING
2330 Details : request with rq_no_resend flag not awake l_wait_event if they get
2334 Frequency : nfs export on patchless client
2336 Description: connectathon hang when test nfs export over patchless client
2337 Details : Disconnected dentry cannot be found with lookup, so we do not need
2338 to unhash it or make it invalid
2341 Description: fix llapi_lov_get_uuids() to allow many OSTs to be returned
2342 Details: : Change llapi_lov_get_uuids() to read the UUIDs from /proc instead
2343 of using an ioctl. This allows lfsck for > 160 OSTs to succeed.
2348 Description: open req refcounting wrong on reconnect
2349 Details : If reconnect happened between getting open reply from server and
2350 call to mdc_set_replay_data in ll_file_open, we will schedule
2351 replay for unreferenced request that we are about to free.
2352 Subsequent close will crash in variety of ways.
2353 Check that request is still eligible for replay in
2354 mdc_set_replay_data().
2359 Description: disable writes to filesystem when reading health_check file
2360 Details : the default for reading the health_check proc file has changed
2361 to NOT do a journal transaction and write to disk, because this
2362 can cause reads of the /proc file to hang and block HA state
2363 checking on a healthy but otherwise heavily loaded system. It
2364 is possible to return to the previous behaviour during configure
2365 with --enable-health-write.
2367 Severity : enhancement
2369 Description: 64-bit inode version
2370 Details: : Add a on-disk 64-bit inode version for ext3 to track changes made
2371 to the inode. This will be required for version-based recovery.
2376 Description: MDS fails to start if a duplicate client export is detected
2377 Details : in some rare cases it was possible for a client to connect to
2378 an MDS multiple times. Upon recovery the MDS would detect this
2379 and fail during startup. Handle this more gracefully.
2381 Severity : enhancement
2383 Description: Add -o localflock option to simulate old noflock
2385 Details : This will achieve local-only flock/fcntl locks
2391 Description: log_commit_thread vs filter_destroy race leads to crash
2392 Details : Take import reference before releasing llog record semaphore
2397 Description: Wrong request locking in request set processing
2398 Details : ptlrpc_check_set wrongly uses req->rq_lock for proctect add to
2399 imp_delayed_list, in this place should be used imp_lock.
2402 Frequency : when reconnection
2404 Description: Grant Leak when osc reconnect to OST
2405 Details : When osc reconnect ost, OST(filter) should check whether it
2406 should grant more space to client by comparing fed_grant and
2407 cl_avail_grant, and return the granted space to client instead
2408 of "new granted" space, because client will call osc_init_grant
2409 to update the client grant space info.
2412 Frequency : when client reconnect to OST
2414 Description: Grant Leak when osc do resend and replay bulk write
2415 Details : When osc reconnect to OST, OST(filter)should clear grant info of
2416 bulk write request, because the grant info while be sync between
2417 OSC and OST when reconnect, and we should ignore the grant info
2418 these of resend/replay write req.
2423 Description: Grant space more than avaiable left space sometimes.
2424 Details : When then OST is about to be full, if two bulk writing from
2425 different clients came to OST. Accord the avaliable space of the
2426 OST, the first req should be permitted, and the second one
2427 should be denied by ENOSPC. But if the seconde arrived before
2428 the first one is commited. The OST might wrongly permit second
2429 writing, which will cause grant space > avaiable space.
2432 Frequency : when client is evicted
2434 Description: Grant might be wrongly erased when osc is evicted by OST
2435 Details : when the import is evicted by server, it will fork another
2436 thread ptlrpc_invalidate_import_thread to invalidate the
2437 import, where the grant will be set to 0. While the original
2438 thread will update the grant it got when connecting. So if
2439 the former happened latter, the grant will be wrongly errased
2440 because of this race.
2445 Description: Checking Stale with correct fid
2446 Details : ll_revalidate_it should uses de_inode instead of op_data.fid2
2447 to check whether it is stale, because sometimes, we want the
2448 enqueue happened anyway, and op_data.fid2 will not be initialized.
2450 Severity : enhancement
2452 Description: update patchless client
2453 Details : Add support for patchless client with 2.6.20, 2.6.21 and RHEL 5
2456 Frequency : only with 2.4 kernel
2458 Description: random memory corruption
2459 Details : size of struct ll_inode_info is to big for union inode.u and this
2460 can be cause of random memory corruption.
2465 Description: Memory leak in recovery
2466 Details : Lov_mds_md was not free in an error handler in mds_create_object.
2467 It should also check obd_fail before fsfilt_start, otherwise if
2468 fsfilt_start return -EROFS,(failover mds during mds recovery).
2469 then the req will return with repmsg->transno = 0 and rc = EROFS.
2470 and we met hit the assert LASSERT(req->rq_reqmsg->transno ==
2471 req->rq_repmsg->transno) in ptlrpc_replay_interpret. Fcc should
2472 be freed no matter whether fsfilt_commit success or not.
2475 Frequency : only with huge count clients
2477 Description: Prevents from taking the superblock lock in llap_from_page for
2479 Details : using LL_ORIGIN_REMOVEPAGE origin flag instead of LL_ORIGIN_UNKNOW
2480 for llap_from_page call in ll_removepage prevents from taking the
2481 superblock lock for a soon died page.
2486 Description: Not check open intent error before release open handle
2487 Details : in some rare cases, the open intent error is not checked before
2488 release open handle, which may cause
2489 ASSERTION(open_req->rq_transno != 0), because it tries to release
2490 the failed open handle.
2495 Description: Set cat log bitmap only after create log success.
2496 Details : in some rare cases, the cat log bitmap is set too early. and it
2497 should be set only after create log success.
2501 Description: Accessing a block bevice can re-enable I/O when Lustre is
2502 tearing down a device.
2503 Details : dev_clear_rdonly(bdev) must be called in kill_bdev() instead of
2508 Description: service threads may hog cpus when there are a lot of requests
2509 Details : Insert cond_resched to give other threads a chance to use some CPU
2514 Description: the cat log was not initialized in recovery
2515 Details : When mds(mgs) do recovery, the tgt_count might be zero, so the
2516 unlink log on mds will not be initialized until mds post
2517 recovery. And also in mds post recovery, the unlink log will
2518 initialization will be done asynchronausly, so there will be race
2519 between add unlink log and unlink log initialization.
2523 Description: brw_stats were being printed incorrectly
2524 Details : brw_stats were being printed as log2 but all of them were not
2525 recorded as log2. Also remove some code duplication arising from
2526 filter_tally_{read,write}.
2530 Frequency : rare, only in recovery.
2531 Description: ASSERTION(req->rq_type != LI_POISON) failed
2532 Details : imp_lock should be held while iterating over imp_sending_list for
2533 prevent destroy request after get timeout in ptlrpc_queue_wait.
2537 Description: replay-single.sh test 52 fails
2538 Details : A lock's skiplist need to be cleanup when it being unlinked
2539 from its resource list.
2543 Description: Short directio read returns full requested size rather than
2545 Details : Direct I/O operations should return actual amount of bytes
2546 transferred rather than requested size.
2548 Severity : enhancement
2550 Description: metadata RPC reduction (e.g. for rm performance)
2551 Details : decrease the amount of synchronous RPC between clients and servers
2552 by canceling conflicing lock before the operation on the client
2553 and packing thier handles into the main operation RPC to server.
2555 Severity : enhancement
2557 Description: Async OSC create to avoid the blocking unnecessarily.
2558 Details : If a OST has no remain object, system will block on the creating
2559 when need to create a new object on this OST. Now, ways use
2560 pre-created objects when available, instead of blocking on an
2561 empty osc while others are not empty. If we must block, we block
2562 for the shortest possible period of time.
2566 Description: improve handling recoverable errors
2567 Details : if request processig with error which can be recoverable on server
2568 request should be resend, otherwise page released from cache and
2571 Severity : enhancement
2573 Description: refine locking for avoid write wrong info into lov_objid file
2574 Details : fix possible races with add new target and write/update data in
2577 --------------------------------------------------------------------------------
2579 2007-05-03 Cluster File Systems, Inc. <info@clusterfs.com>
2584 Frequency : on some architectures
2586 Description: 1.6 client sometimes fails to mount from a 1.4 MDT
2587 Details : Uninitialized flags sometimes cause configuration commands to
2591 Frequency : patchless clients only
2593 Description: missing __iget() symbol export
2594 Details : The __iget() symbol export is missing. To avoid the need for
2595 this on patchless clients the deathrow inode reaper is turned
2596 off, and we depend on the VM to clean up old inodes. This
2597 dependency was during via the fix for bug 12181.
2599 --------------------------------------------------------------------------------
2601 2007-04-19 Cluster File Systems, Inc. <info@clusterfs.com>
2603 * CONFIGURATION CHANGE. This version of Lustre WILL NOT
2604 INTEROPERATE with older versions automatically. In many cases a
2605 special upgrade step is needed. Please read the
2606 user documentation before upgrading any part of a live system.
2607 * WIRE PROTOCOL CHANGE from previous 1.6 beta versions. This
2608 version will not interoperate with 1.6 betas before beta5 (1.5.95).
2609 * WARNING: Lustre configuration and startup changes are required with
2610 this release. See https://mail.clusterfs.com/wikis/lustre/MountConf
2614 Severity : enhancement
2616 Description: MountConf
2617 Details : Lustre configuration is now managed via mkfs and mount
2618 commands instead of lmc and lconf. New obd types (MGS, MGC)
2619 are added for dynamic configuration management. See
2620 https://mail.clusterfs.com/wikis/lustre/MountConf for
2623 Severity : enhancement
2625 Description: dynamic OST addition
2626 Details : OSTs can now be added to a live filesystem
2628 Severity : enhancement
2630 Description: startup order invariance
2631 Details : MDTs and OSTs can be started in any order. Clients only
2632 require the MDT to complete startup.
2634 Severity : enhancement
2636 Description: parallel, asynchronous orphan cleanup
2637 Details : orphan cleanup is now performed in separate threads for each
2638 OST, allowing parallel non-blocking operation.
2640 Severity : enhancement
2642 Description: optimized stripe assignment
2643 Details : stripe assignments are now made based on ost space available,
2644 ost previous usage, and OSS previous usage, in order to try
2645 to optimize storage space and networking resources.
2647 Severity : enhancement
2649 Description: Permanently set tunables
2650 Details : All writable /proc/fs/lustre tunables can now be permanently
2651 set on a per-server basis, at mkfs time or on a live system.
2653 Severity : enhancement
2655 Description: Lustre message v2
2656 Details : Add lustre message format v2.
2658 Severity : enhancement
2660 Description: client OST exclusion list
2661 Details : Clients can be started with a list of OSTs that should be
2662 declared "inactive" for known non-responsive OSTs.
2666 Description: SPEC SFS validation failure on NFS v2 over lustre.
2667 Details : Changes the blocksize for regular files to be 2x RPC size,
2668 and not depend on stripe size.
2670 Severity : enhancement
2672 Description: Multiple MD RPCs in flight.
2673 Details : Further unserialise some read-only MDS RPCs - learn about intents.
2674 To avoid overly-overloading MDS, introduce a limit on number of
2675 MDS RPCs in flight for a single client and add /proc controls
2676 to adjust this limit.
2678 Severity : enhancement
2680 Description: client read/write statistics
2681 Details : Add client read/write call usage stats for performance
2682 analysis of user processes.
2683 /proc/fs/lustre/llite/*/offset_stats shows non-sequential
2684 file access. extents_stats shows chunk size distribution.
2685 extents_stats_per_process show chunk size distribution per
2688 Severity : enhancement
2690 Description: mds statistics
2691 Details : Add detailed mds operations statistics in
2692 /proc/fs/lustre/mds/*/stats.
2696 Description: Failure of copying files with lustre special EAs.
2697 Details : Client side always return success for setxattr call for lustre
2698 special xattr (currently only "trusted.lov").
2703 Description: Refcount LNET uuids
2704 Details : The global LNET uuid list grew linearly with every startup;
2705 refcount repeated list entries instead of always adding to
2708 Severity : enhancement
2710 Description: Dynamic service threads
2711 Details : Within a small range, start extra service threads
2712 automatically when the request queue builds up.
2715 Frequency : mixed-endian client/server environments
2717 Description: mixed-endian crashes
2718 Details : The new msg_v2 system had some failures in mixed-endian
2721 Severity : enhancement
2723 Description: Easy OST removal
2724 Details : OSTs can be permanently deactivated with e.g. 'lctl
2725 conf_param lustre-OST0001.osc.active=0'
2727 Severity : enhancement
2729 Description: MGS proc entries
2730 Details : Added basic proc entries for the MGS showing what filesystems
2733 Severity : enhancement
2735 Description: provide MGS failover
2736 Details : Added config lock reacquisition after MGS server failover.
2738 Severity : enhancement
2740 Description: add Linux 2.4 support
2741 Details : Added support for RHEL 2.4.21 kernel for 1.6 servers and clients
2745 Description: a large application tries to do I/O to the same resource and dies
2746 in the middle of it.
2747 Details : Check the req->rq_arrival time after the call to
2748 ost_brw_lock_get(), but before we do anything about
2749 processing it & sending the BULK transfer request. This
2750 should help move old stale pending locks off the queue as
2751 quickly as obd_timeout.
2754 Frequency : when an incorrect nid is specified during startup
2756 Description: ptlrpc connect to non-existant node causes kernel crash
2757 Details : LNET can't be re-entered from an event callback, which
2758 happened when we expire a message after the export has been
2759 cleaned up. Instead, hand the zombie cleanup off to another
2762 Severity : enhancement
2764 Description: plain/inodebits lock performance improvement
2765 Details : Grouping plain/inodebits in granted list by their request modes
2766 and bits policy, thus improving the performance of search through
2770 Frequency : only if OST filesystem is corrupted
2772 Description: client incorrectly hits assertion in ptlrpc_replay_req()
2773 Details : for a short time RPCs with bulk IO are in the replay list,
2774 but replay of bulk IOs is unimplemented. If the OST filesystem
2775 is corrupted due to disk cache incoherency and then replay is
2776 started it is possible to trip an assertion. Avoid putting
2777 committed RPCs into the replay list at all to avoid this issue.
2780 Frequency : liblustre (e.g. catamount) on a large cluster with >= 8 OSTs/OSS
2782 Description: System hang on startup
2783 Details : This bug allowed the liblustre (e.g. catamount) client to
2784 return to the app before handling all startup RPCs. This
2785 could leave the node unresponsive to lustre network traffic
2786 and manifested as a server ptllnd timeout.
2788 Severity : enhancement
2790 Description: Add "/proc/sys/lustre/debug_peer_on_timeout"
2791 Details : liblustre envirable: LIBLUSTRE_DEBUG_PEER_ON_TIMEOUT
2792 boolean to control whether to print peer debug info when a
2793 client's RPC times out.
2796 Frequency : only for kernels with patches from Lustre below 1.4.3
2798 Description: Remove old rdonly API
2799 Details : Remove old rdonly API which unused from at least lustre 1.4.3
2802 Frequency : only for devices with external journals
2804 Description: Set external device read-only also
2805 Details : During a commanded failover stop, we set the disk device
2806 read-only while the server shuts down. We now also set any
2807 external journal device read-only at the same time.
2810 Frequency : when upgrading from 1.4 while trying to change parameters
2812 Description: The wrong (new) MDC name was used when setting parameters for
2813 upgraded MDT's. Also allows changing of OSC (and MDC)
2814 parameters if --writeconf is specified at tunefs upgrade time.
2817 Frequency : when setting specific ost indicies
2819 Description: QOS code breaks on skipped indicies
2820 Details : Add checks for missing OST indicies in the QOS code, so OSTs
2821 created with --index need not be sequential.
2823 Severity : enhancement
2825 Description: Add uninit_groups feature to ldiskfs2 to speed up e2fsck
2826 Details : The uninit_groups feature works in conjunction with the kernel
2827 filesystem code (ldiskfs2 only) and e2fsprogs-1.39-cfs6 to speed
2828 up the pass1 processing of e2fsck. This is a read-only feature
2829 in ldiskfs2 only, so older kernels and current ldiskfs cannot
2830 mount filesystems that have had this feature enabled.
2832 Severity : enhancement
2834 Description: Improve multi-block allocation algorithm to avoid fragmentation
2835 Details : The mballoc3 code (ldiskfs2 only) adds new mechanisms to improve
2836 allocation locality and avoid filesystem fragmentation.
2838 ------------------------------------------------------------------------------
2840 2007-04-01 Cluster File Systems, Inc. <info@clusterfs.com>
2842 * Support for kernels:
2843 2.4.21-47.0.1.EL (RHEL 3)
2844 2.6.5-7.283 (SLES 9)
2845 2.6.9-42.0.10.EL (RHEL 4)
2846 2.6.12.6 vanilla (kernel.org)
2847 2.6.16.27-0.9 (SLES 10)
2848 * Recommended e2fsprogs version: 1.39.cfs5
2850 * Note that reiserfs quotas are disabled on SLES 10 in this kernel
2854 Frequency : occasional, depends on client load and configuration
2855 Bugzilla : 12181, 12203
2856 Description: data loss for recently-modified files
2858 Details : In some cases it is possible that recently written or created
2859 files may not be written to disk in a timely manner (this should
2860 normally be within 30s unless client IO load is very high).
2861 The problem appears as zero-length files or files that are a
2862 multiple of 1MB in size after a client crash or client eviction
2863 that are missing data at the end of the file.
2865 This problem is more likely to be hit on clients where files are
2866 repeatedly created and unlinked in the same directory, clients
2867 have a large amount of RAM, have many CPUs, the filesystem has
2868 many OSTs, the clients are rebooted frequently, and/or the files
2869 are not accessed by other nodes after being written.
2871 The presence of the problem can be detected by looking at
2872 /proc/sys/fs/inode-state. If the first number (nr_inodes) is
2873 smaller than the second (nr_unused) then dirty files will not
2874 be flushed automatically to disk. "sync; sleep 10" should be
2875 run several times on the node before unmounting it to update
2876 Lustre (this is also safe to run on nodes without this problem).
2878 There is also a related kernel bug in the RHEL4 4 2.6.9 kernel
2879 that can cause this same problem, so customers using that kernel
2880 also need to update the kernel in addition to Lustre. In order
2881 to properly fix this bug, the RHEL3 2.4.21 kernel is also updated.
2883 It is normal that files written just before a client crash (less
2884 than 30s) may not yet have been flushed to disk, even for local
2888 Frequency : frequent on thin XT3 nodes
2890 Description: UUID collision on thin XT3 Linux nodes
2891 Details : UUIDs on Compute Node Linux XT3 nodes were not generated
2892 randomly, since we relied on an insufficiently-seeded PRNG.
2897 Description: OSS hangs after "All ost request buffers busy"
2898 Details : A deadlock between quota and journal operations caused OSS
2899 hangs after printing "All ost request buffers busy."
2902 Frequency : always on liblustre builds
2904 Description: Cleanup compiler warnings on liblustre
2907 Frequency : always on liblustre builds on XT3
2909 Description: LC_CONFIG_CDEBUG don't run while build liblustre on XT3.
2913 Description: Addition of EXT3_FEATURE_RO_COMPAT_DIR_NLINKS flag for
2914 > 32000 subdirectories
2915 Details : Add EXT3_FEATURE_RO_COMPAT_DIR_NLINK flag to
2916 EXT3_FEATURE_RO_COMPAT_SUPP. This flag will be set whenever
2917 subdirectory count crosses 32000. This will aid e2fsck to
2918 correctly handle more than 32000 subdirectories.
2921 Frequency : liblustre (e.g. catamount) on a large cluster with >= 8 OSTs/OSS
2923 Description: System hang on startup
2924 Details : This bug allowed the liblustre (e.g. catamount) client to
2925 return to the app before handling all startup RPCs. This
2926 could leave the node unresponsive to lustre network traffic
2927 and manifested as a server ptllnd timeout.
2929 Severity : enhancement
2931 Description: Add "/proc/sys/lustre/debug_peer_on_timeout"
2932 (liblustre envirable: LIBLUSTRE_DEBUG_PEER_ON_TIMEOUT)
2933 boolean to control whether to print peer debug info when a
2934 client's RPC times out.
2939 Description: make O_SYNC working on 2.6 kernels
2940 Details : 2.6 kernels use different method for mark pages for write,
2941 so need add a code to lustre for O_SYNC work.
2946 Description: Failure to close file and release space on NFS
2947 Details : Put inode details into lock acquired in ll_intent_file_open.
2948 Use mdc_intent_lock in ll_intent_open to properly
2949 detect all kind of errors unhandled by mdc_enqueue.
2954 Description: proc file read during shutdown sometimes raced obd removal,
2956 Details : Add lock to prevent obd access after proc file removal.
2959 Frequency : Only for files larger than 4GB on 32-bit clients.
2961 Description: improperly doing page alignment of locks
2962 Details : Modify lustre core code to use CFS_PAGE_* defines instead of
2963 PAGE_*. Make CFS_PAGE_MASK a 64-bit mask.
2968 Description: RPCs being resent when they shouldn't be
2969 Details : Some RPCs that should not be resent are being resent. This
2970 can cause inconsistencies in the RPC state machine. Do not
2971 resend such requests.
2974 Frequency : rare, only with NFS export
2976 Description: Crash on NFS re-export node
2977 Details : under very unusual load conditions an assertion is hit in
2978 ll_intent_file_open()
2981 Frequency : only if OST filesystem is corrupted
2983 Description: client incorrectly hits assertion in ptlrpc_replay_req()
2984 Details : for a short time RPCs with bulk IO are in the replay list,
2985 but replay of bulk IOs is unimplemented. If the OST filesystem
2986 is corrupted due to disk cache incoherency and then replay is
2987 started it is possible to trip an assertion. Avoid putting
2988 committed RPCs into the replay list at all to avoid this issue.
2993 Description: large O_DIRECT requests fail under memory pressure/fragmentation
2994 Details : Large single O_DIRECT read and write calls can fail to allocate
2995 a sufficiently large buffer to process the request. In case of
2996 allocation failure the allocation is retried with a smaller
2997 buffer and broken into smaller requests.
2999 Severity : enhancement
3001 Description: Add -o localflock option to simulate old noflock behaviour.
3002 Details : This will achieve local-only flock/fcntl locks coherentness.
3007 Description: versioning check is incomplete
3008 Details : Checking the version difference of client vs. server, report
3009 error if the gap is too big.
3014 Description: add support PG_writeback bit
3015 Details : add support for PG_writeback bit for Lustre, for more carefull
3016 work with page cache in 2.6 kernel. This also fix some deadlocks
3017 and remove hack for work O_SYNC with 2.6 kernel.
3019 Severity : enhancement
3021 Description: Add uninit_groups feature to ldiskfs2 to speed up e2fsck
3022 Details : The uninit_groups feature works in conjunction with the kernel
3023 filesystem code (ldiskfs2 only) and e2fsprogs-1.39-cfs6 to speed
3024 up the pass1 processing of e2fsck. This is a read-only feature
3025 in ldiskfs2 only, so older kernels and current ldiskfs cannot
3026 mount filesystems that have had this feature enabled.
3028 Severity : enhancement
3030 Description: Improve multi-block allocation algorithm to avoid fragmentation
3031 Details : The mballoc3 code (ldiskfs2 only) adds new mechanisms to improve
3032 allocation locality and avoid filesystem fragmentation.
3034 ------------------------------------------------------------------------------
3036 2007-02-09 Cluster File Systems, Inc. <info@clusterfs.com>
3038 * Support for kernels:
3039 2.6.9-42.0.3.EL (RHEL 4)
3040 2.6.5-7.276 (SLES 9)
3041 2.4.21-47.0.1.EL (RHEL 3)
3042 2.6.12.6 vanilla (kernel.org)
3043 2.6.16.21-0.8 (SLES10)
3044 * Recommended e2fsprogs version: 1.39.cfs2-0
3046 * The backwards-compatible /proc/sys/portals symlink has been removed
3047 in this release. Before upgrading, please ensure that you change
3048 any configuration scripts or /etc/sysctl.conf files that access
3049 /proc/sys/portals/* or sysctl portals.* to use the corresponding
3050 entry in /proc/sys/lnet or sysctl lnet.*. This change can be made
3051 in advance of the upgrade on any system running Lustre 1.4.6 or
3052 newer, since /proc/sys/lnet was added in that version.
3053 * Note that reiserfs quotas are disabled on SLES 10 in this kernel
3057 Frequency : only when quota is used
3059 Description: avoid scanning export list for quota master
3060 Details : Change the algorithms to avoid scanning export list in order
3061 to improve the efficiency.
3064 Frequency : MDS failover only, very rarely
3066 Description: "went back in time" messages on mds failover
3067 Details : The greatest transno may be lost when the current operation
3068 finishes with an error (transno==0) and the client's last_rcvd
3069 record is over-written. Save the greatest transno in the
3070 mds_last_transno for this case.
3073 Frequency : always for specific kernels and striping counts
3075 Description: client may get "Matching packet too big" without ACL support
3076 Details : Clients compiled without CONFIG_FS_POSIX_ACL get an error message
3077 when trying to access files in certain configurations. The
3078 clients should in fact be denied when mounting because they do
3079 not understand ACLs.
3082 Frequency : Cray XT3 with more than 4000 clients and multiple jobs
3084 Description: many clients connecting with IO in progress causes connect timeouts
3085 Details : Avoid synchronous journal commits to avoid delays caused by many
3086 clients connecting/disconnecting when bulk IO is in progress.
3087 Queue liblustre connect requests on OST_REQUEST_PORTAL instead of
3088 OST_IO_PORTAL to avoid delays behind potentially many pending
3092 Frequency : occasionally with multiple writers to a single file
3094 Description: shared writes to file may result in wrong size reported by stat()
3095 Details : Allow growing of kms when extent lock is cancelled
3098 Frequency : always with random mmap IO to multi-striped file
3100 Description: mmap write might be lost if we are writing to a 'hole' in stripe
3101 Details : Only if the hole is at the end of OST object so that kms is too
3102 small. Fix is to increase kms accordingly in ll_nopage.
3105 Frequency : rare, only if OST filesystem is inconsistent with MDS filesystem
3107 Description: writes to a missing object would leak memory on the OST
3108 Details : If there is an inconsistency between the MDS and OST filesystems,
3109 such that the MDS references an object that doesn't exist, writes
3110 to that object will leak memory due to incorrect cleanup in the
3111 error handling path, eventually running out of memory on the OST.
3116 Description: Creating too long symlink causes lustre errors
3117 Details : Check symlink and name lengths before sending requests to MDS.
3120 Frequency : only if flock is enabled (not on by default)
3122 Description: posix locks not released on fd closure on 2.6.9+
3123 Details : We failed to add posix locks to list of inode locks on 2.6.9+
3124 kernels, this caused such locks not to be released on fd close and
3125 then assertions on fs unmount about still used locks.
3128 Frequency : MDS failover only, very rarely
3130 Description: clients may get ASSERTION(granted_lock != NULL)
3131 Details : When request was taking a long time, and a client was resending
3132 a getattr by name lock request. The were multiple lock requests
3133 with the same client lock handle and
3134 mds_getattr_name->fixup_handle_for_resent_request found one of the
3135 lock handles but later failed with ASSERTION(granted_lock != NULL).
3140 Description: handle->h_buffer_credits > 0, assertion failure
3141 Details : h_buffer_credits is zero after truncate, causing assertion
3142 failure. This patch extends the transaction or creates a new
3146 Frequency : NFS re-export or patchless client
3147 Bugzilla : 11179, 10796
3148 Description: Crash on NFS re-export node (__d_move)
3149 Details : We do not want to hash the dentry if we don't have a lock.
3150 But if this dentry is later used in d_move, we'd hit uninitialised
3151 list head d_hash, so we just do this to init d_hash field but
3152 leave dentry unhashed.
3155 Frequency : NFS re-export or patchless client
3157 Description: NFS exports has problem with symbolic link
3158 Details : lustre client didn't properly install dentry when re-exported
3159 to NFS or running patchless client.
3162 Frequency : NFS re-export or patchless client
3164 Description: Various nfs/patchless fixes.
3165 Details : fixes reuse disconected alias for lookup process - this fixes
3166 warning "find_exported_dentry: npd != pd",
3167 fix permission error with open files at nfs.
3168 fix apply umask when do revalidate.
3171 Frequency : occasional
3173 Description: Crash on NFS re-export node
3174 Details : calling clear_page() on the wrong pointer triggered oops in
3175 generic_mapping_read().
3178 Frequency : rarely, using O_DIRECT IO
3180 Description: unaligned directio crashes client with LASSERT
3181 Details : check for unaligned buffers before trying any requests.
3184 Frequency : rarely, using CFS RAID5 patches in non-standard kernel series
3186 Description: stale data returned from RAID cache
3187 Details : If only a small amount of IO is done to the RAID device before
3188 reading it again it is possible to get stale data from the RAID
3189 cache instead of reading it from disk.
3192 Frequency : always for sles10 kernel
3194 Description: sles10 support
3195 Details : ll_follow_link: compile fixes and using of nd_set_link
3196 under newer kernels.
3199 Frequency : depends on arch, kernel and compiler version, always on sles10
3202 Description: recursive or deep enough symlinks cause stack overflow
3203 Details : getting rid of large stack-allocated variable in
3207 Frequency : depends on hardware
3209 Description: lustre write performance loss in the SLES10 kernel
3210 Details : the performance loss is caused by using of write barriers in the
3211 ext3 code. The SLES10 kernel turns barrier support on by
3212 default. The fix is to undo that change for ldiskfs.
3214 ------------------------------------------------------------------------------
3216 2006-12-09 Cluster File Systems, Inc. <info@clusterfs.com>
3218 * Support for kernels:
3219 2.6.9-42.0.3EL (RHEL 4)
3220 2.6.5-7.276 (SLES 9)
3221 2.4.21-47.0.1.EL (RHEL 3)
3222 2.6.12.6 vanilla (kernel.org)
3226 Frequency : quota enabled and large files being deleted
3228 Description: releasing more than 4GB of quota at once hangs OST
3229 Details : If a user deletes more than 4GB of files on a single OST it
3230 will cause the OST to spin in an infinite loop. Release
3231 quota in < 4GB chunks, or use a 64-bit value for 1.4.7.1+.
3236 Description: statfs data retrieved from /proc may be stale or zero
3237 Details : When reading per-device statfs data from /proc, in the
3238 {kbytes,files}_{total,free,avail} files, it may appear
3239 as zero or be out of date.
3242 Frequency : systems with MD RAID1 external journal devices
3244 Description: lconf's call to blkid is confused by RAID1 journal devices
3245 Details : Use the "blkid -l" flag to locate the MD RAID device instead
3246 of returning all block devices that match the journal UUID.
3249 Frequency : always, for aggregate stripe size over 4GB
3251 Description: "lfs setstripe" fails assertion when setting 4GB+ stripe width
3252 Details : Using "lfs setstripe" to set stripe size * stripe count over 4GB
3253 will fail the kernel with "ASSERTION(lsm->lsm_xfersize != 0)"
3256 Frequency : always if "lfs find" used on a local file/directory
3258 Description: "lfs find" segfaults if used on a local file/directory
3259 Details : The case where a directory component was not specified wasn't
3260 handled correctly. Handle this properly.
3263 Frequency : always on ppc64
3265 Description: the write to an ext3 filesystem mounted with mballoc got stuck
3266 Details : ext3_mb_generate_buddy() uses find_next_bit() which does not
3267 perform endianness conversion.
3270 Frequency : rarely (truncate to non-zero file size after write under load)
3271 Bugzilla : 10730, 10687
3272 Description: Files padded with zeros to next 4K multiple
3273 Details : With filesystems mounted using the "extents" option (2.6 kernels)
3274 it is possible that files that are truncated to a non-zero size
3275 immediately after being written are filled with zero bytes beyond
3276 the truncated size. No file data is lost.
3278 Severity : enhancement
3280 Description: Allow recovery/failover for liblustre clients.
3281 Details : liblustre clients were unaware of failover configurations until
3284 Severity : enhancement
3286 Description: user file locks should fail when not mounting with flock option
3287 Details : Set up an error-returning stub in ll_file_operations.lock field
3288 to prevent incorrect behaviour when client is mounted without
3289 flock option. Also, set up properly f_op->flock field for
3293 Frequency : always on ia64
3295 Description: "lfs df" loops on printing out MDS statfs information
3296 Details : The obd_ioctl_data was not initialized and in some systems
3297 this caused a failure during the ioctl that did not return
3298 an error. Initialize the struct and return an error on failure.
3301 Frequency : SLES 9 only
3303 Description: Error of copying files with lustre special EAs as root
3304 Details : Client side always return success for setxattr call for lustre
3305 special xattr (currently only "trusted.lov").
3308 Frequency : rarely on clusters with both ia64+i386 clients
3310 Description: ia64+i686 clients doing shared IO on the same file may LBUG
3311 Details : In rare cases when both ia64+i686 (or other mixed-PAGE_SIZE)
3312 clients are doing concurrent writes to the same file it is
3313 possible that the ia64 clients may LASSERT because the OST
3314 extent locks are not PAGE_SIZE aligned. Ensure that grown
3315 locks are always aligned on the request boundary.
3318 Frequency : specific use, occasional
3320 Description: Overwriting in use executable truncates on-disk binary image
3321 Details : If one node attempts to overwrite an executable in use by
3322 another node, we now correctly return ETXTBSY instead of
3323 truncating the file.
3325 Severity : enhancement
3327 Description: Async OSC create to avoid the blocking unnecessarily.
3328 Details : If a OST has no remain object, system will block on the creating
3329 when need to create a new object on this OST. Now, ways use
3330 pre-created objects when available, instead of blocking on an
3331 empty osc while others are not empty. If we must block, we block
3332 for the shortest possible period of time.
3337 Description: chmod on Lustre root is propagated to other clients
3338 Details : Re-validate root's dentry in ll_lookup_it to avoid having it
3339 invalid by the follow_mount time.
3342 Frequency : liblustre clients only
3344 Description: Race in 'instant cancel' lock handling could lead to such locks
3345 never to be granted in case of SMP MDS
3346 Details : Do not destroy not yet granted but cbpending locks in
3350 Frequency : replay/resend of open
3352 Description: non null lock assetion failure in mds_intent_policy
3353 Details : Trying to replay/resend lockless open requests resulted in
3354 mds_open() returning 0 with no lock. Now it sets a flag if
3355 a lock is going to be returned.
3357 Severity : enhancement
3359 Description: Checksum enhancements
3360 Details : New checksum enhancements allow for resending RPCs that failed
3363 Severity : enhancement
3365 Description: Tunables on number of dirty pages in cacche
3366 Details : Allow to set limit on number of dirty pages cached.
3371 Description: client crash on unmount - lock still has references
3372 Details : In some error handling cases it was possible to leak a lock
3373 reference on a client while accessing a file. This was not
3374 harmful to the client during operation, but would cause the
3375 client to crash when the filesystem is unmounted.
3378 Frequency : specific case, rare
3380 Description: ETXTBSY on mds though file not in use by client
3381 Details : ETXTBSY is no longer incorrectly returned when attempting to
3382 chmod or chown a directory that the user previously tried to
3383 execute or a currently-executing binary.
3386 Frequency : extremely rare except on liblustre-based clients
3388 Description: Lustre space not freed when files are deleted
3389 Details : Clean up open-unlinked files after client eviction. Previously
3390 the unlink was skipped and the files remained as orphans.
3395 Description: OST failure "would be an LBUG" in waiting_locks_callback()
3396 Details : In some cases it was possible to send a blocking callback to
3397 a client doing a glimpse, even though that client didn't get
3398 a lock granted. When the glimpse lock is cancelled on the OST
3399 the freed lock is left on the waiting list and corrupted the list.
3402 Frequency : all core dumps
3404 Description: Broke core dumps to lustre
3405 Details : Negative dentry may be unhashed if parent does not have UPDATE
3406 lock, but some callers, e.g. do_coredump, expect dentry to be
3407 hashed after successful create, hash it in ll_create_it.
3409 ------------------------------------------------------------------------------
3411 2006-09-13 Cluster File Systems, Inc. <info@clusterfs.com>
3413 * Support for kernels:
3414 2.6.9-42.0.2.EL (RHEL 4)
3415 2.6.5-7.276 (SLES 9)
3416 2.4.21-40.EL (RHEL 3)
3417 2.6.12.6 vanilla (kernel.org)
3421 Frequency : always on RHEL 3
3423 Description: Number of open files grows over time
3424 Details : The number of open files grows over time, whether or not
3425 Lustre is started. This was due to a filp leak introduced
3426 by one of our kernel patches.
3428 ------------------------------------------------------------------------------
3430 08-20-2006 Cluster File Systems, Inc. <info@clusterfs.com>
3432 * Support for kernels:
3433 2.6.9-42.EL (RHEL 4)
3434 2.6.5-7.276 (SLES 9)
3435 2.4.21-40.EL (RHEL 3)
3436 2.6.12.6 vanilla (kernel.org)
3441 Bugzilla : 5719, 9635, 9792, 9684
3442 Description: OST (or MDS) trips assertions in (re)connection under heavy load
3443 Details : If a server is under heavy load and cannot reply to new
3444 connection requests before the client resends the (re)connect,
3445 the connection handling code can behave badly if two service
3446 threads are concurrently handing separate (re)connections from
3447 the same client. Add better locking to the connection handling
3448 code, and ensure that only a single connection will be processed
3449 for a given client UUID, even if the lock is dropped.
3451 Severity : enhancement
3453 Description: add TCP zero-copy support to kernel
3454 Details : Add support to the kernel TCP stack to allow zero-copy bulk
3455 sends if the hardware supports scatter-gather and checksumming.
3456 This allows socklnd to do client-write and server-read more
3457 efficiently and reduce CPU utilization from skbuf copying.
3460 Frequency : only if NFS exporting from client
3462 Description: NULL pointer deref in ll_iocontrol() if chattr mknod file
3463 Details : If setting attributes on a file created under NFS that had
3464 never been opened it would be possible to oops the client
3465 if the file had no objects.
3468 Frequency : always for liblustre
3470 Description: liblustre client does MDS+OSTs setattr RPC for each write
3471 Details : When doing a write from a liblustre client, the client
3472 incorrectly issued an RPC to the MDS and each OST the file was
3473 striped over in order to update the timestamps. When writing
3474 with small chunks and many clients this could overwhelm the MDS
3475 with RPCs. In all cases it would slow down the write because
3476 these RPCs are unnecessary.
3478 Severity : enhancement
3480 Description: allow number of MDS service threads to be changed at module load
3481 Details : It is now possible to change the number of MDS service threads
3482 running. Adding "options mds mds_num_threads={N}" to the MDS's
3483 /etc/modprobe.conf will set the number of threads for the next
3484 time Lustre is restarted (assuming the "mds" module is also
3485 reloaded at that time). The default number of threads will
3486 stay the same, 32 for most systems.
3491 Description: OST crash if filesystem is unformatted or corrupt
3492 Details : If an OST is started on a device that has never been formatted
3493 or if the filesystem is corrupt and cannot even mount then the
3494 error handling cleanup routines would dereference a NULL pointer.
3499 Description: NULL pointer deref in llap_from_page.
3500 Details : get_cache_page_nowait can return a page with NULL (or otherwise
3501 incorrect) mapping if the page was truncated/reclaimed while it was
3502 searched for. Check for this condition and skip such pages when
3503 doing readahead. Introduce extra check to llap_from_page() to
3504 verify page->mapping->host is non-NULL (so page is not anonymous).
3507 Frequency : Sometimes when using sys_sendfile
3509 Description: "page not covered by a lock" warnings from ll_readpage
3510 Details : sendfile called ll_readpage without right page locks present.
3511 Now we introduced ll_file_sendfile that does necessary locking
3512 around call to generic_file_sendfile() much like we do in
3516 Frequency : with certain MDS communication failures at client mount time
3518 Description: NULL pointer deref after failed client mount
3519 Details : a client connection request may delayed by the network layer
3520 and not be sent until after the PTLRPC layer has timed out the
3521 request. If the client fails the mount immediately it will try
3522 to clean up before the network times out the request. Add a
3523 reference from the request import to the obd device and delay
3524 the cleanup until the network drops the request.
3527 Frequency : occasionally during client (re)connect
3529 Description: assertion failure during client (re)connect
3530 Details : processing a client connection request may be delayed by the
3531 client or server longer than the client connect timeout. This
3532 causes the client to resend the connection request. If the
3533 original connection request is replied in this interval, the
3534 client may trip an assertion failure in ptlrpc_connect_interpret()
3535 which thought it would be the only running connect process.
3538 Frequency : only with obd_echo servers and clients that are rebooted
3540 Description: kernel BUG accessing uninitialized data structure
3541 Details : When running an obd_echo server it did not start the ping_evictor
3542 thread, and when a client was evicted an uninitialized data
3543 structure was accessed. Start the ping_evictor in the RPC
3544 service startup instead of the OBD startup.
3546 Severity : enhancement
3547 Bugzilla : 10193 (patchless)
3548 Description: Remove dependency on various unexported kernel interfaces.
3549 Details : No longer need reparent_to_init, exit_mm, exit_files,
3550 sock_getsockopt, filemap_populate, FMODE_EXEC, put_filp.
3553 Frequency : rare (only users of deprecated and unsupported LDAP config)
3555 Description: write_conf for zeroconf mount queried LDAP incorrectly for client
3556 Details : LDAP apparently contains 'lustreName' attributes instead of
3557 'name'. A simple remapping of the name is sufficient.
3560 Frequency : rare (only with non-default dump_on_timeout debug enabled)
3562 Description: waiting_locks_callback trips kernel BUG if client is evicted
3563 Details : Running with the dump_on_timeout debug flag turned on makes
3564 it possible that the waiting_locks_callback() can try to dump
3565 the Lustre kernel debug logs from an interrupt handler. Defer
3566 this log dumping to the expired_lock_main() thread.
3568 Severity : enhancement
3570 Description: Support NFS exporting on 2.6 kernels.
3571 Details : Implement non-rawops metadata methods for NFS server to use without
3572 changing NFS server code.
3575 Frequency : very rare (synthetic metadata workload only)
3577 Description: two racing renames might cause an MDS thread to deadlock
3578 Details : Running the "racer" program may cause one MDS thread to rename
3579 a file from being the source of a rename to being the target of
3580 a rename at exactly the same time that another thread is doing
3581 so, and the second thread has already enqueued these locks after
3582 doing a lookup of the target and is trying to relock them in
3583 order. Ensure that we don't try to re-lock the same resource.
3586 Frequency : only very large systems with liblustre clients
3588 Description: slow eviction of liblustre clients with the "evict_by_nid" RPC
3589 Details : Use asynchronous set_info RPCs to send the "evict_by_nid" to
3590 all OSTs in parallel. This allows the eviction of stale liblustre
3591 clients to proceed much faster than if they were done in series,
3592 and also offers similar improvements for other set_info RPCs.
3597 Description: excessive CPU usage during initial read phase on client
3598 Details : During the initial read phase on a client, it would agressively
3599 retry readahead on the file, consuming too much CPU and impacting
3600 performance (since 1.4.5.8). Improve the readahead algorithm
3601 to avoid this, and also improve some other common cases (read
3602 of small files in particular, where "small" is files smaller than
3603 /proc/fs/lustre/llite/*/max_read_ahead_whole_mb, 2MB by default).
3608 Description: MDS crash when receiving packet with unknown intent.
3609 Details : Do not LBUG in unknown intent case, just return -EFAULT
3611 Severity : enhancement
3612 Bugzilla : 9293, 9385
3613 Description: MDS RPCs are serialised on client. This is unnecessary for some.
3614 Details : Do not serialize getattr (non-intent version) and statfs.
3617 Frequency : occasional, when OST network is overloaded/intermittent
3619 Description: client evicted by OST after bulk IO timeout
3620 Details : If a client sends a bulk IO request (read or write) the OST
3621 may evict the client if it is unresposive to its data GET/PUT
3622 request. This is incorrect if the network is overloaded (takes
3623 too long to transfer the RPC data) or dropped the OST GET/PUT
3624 request. There is no need to evict the client at all, since
3625 the pinger and/or lock callbacks will handle this, and the
3626 client can restart the bulk request.
3629 Frequency : Always when mmapping file with no objects
3631 Description: client crashes when mmapping file with no objects
3632 Details : Check that we actually have objects in a file before doing any
3633 operations on objects in ll_vm_open, ll_vm_close and
3639 Description: Request leak when working with deleted CWD
3640 Details : Introduce advanced request refcount tracking for requests
3641 referenced from lustre intent.
3643 Severity : Enhancement
3645 Description: Cache open file handles on client.
3646 Details : MDS now will return special lock along with openhandle, if
3647 requested and client is allowed to hold openhandle, even if unused,
3648 until such a lock is revoked. Helps NFS a lot, since NFS is opening
3649 closing files for every read/write openration.
3651 Severity : Enhancement
3653 Description: Cache open negative dentries on client when possible.
3654 Details : Guard negative dentries with UPDATE lock on parent dir, drop
3655 negative dentries on lock revocation.
3660 Description: Remounting a client read-only wasn't possible with a zconf mount
3661 Details : It wasn't possible to remount a client read-only with llmount.
3663 Severity : enhancement
3664 Description: Include MPICH 1.2.6 Lustre ADIO interface patch
3665 Details : In lustre/contrib/ or /usr/share/lustre in RPM a patch for
3666 MPICH is included to add Lustre-specific ADIO interfaces.
3667 This is based closely on the UFS ADIO layer and only differs
3668 in file creation, in order to allow the OST striping to be set.
3669 This is user-contributed code and not supported by CFS.
3674 Description: extended inode attributes (immutable, append-only) work improperly
3675 when 2.4 and 2.6 kernels are used on client/server or vice versa
3676 Details : Introduce kernel-independent values for these flags.
3678 Severity : enhancement
3681 Description: Allow fractional MB tunings for lustre in /proc/ filesystem.
3682 Details : Many of the /proc/ tunables can only be tuned at a megabyte
3683 granularity. Now, Fractional MB granularity is be supported,
3684 this is very useful for low memory system.
3686 Severity : enhancement
3688 Description: Getattr by fid
3689 Details : Getting a file attributes by its fid, obtaining UPDATE|LOOKUP
3690 locks, avoids extra getattr rpc requests to MDS, allows '/' to
3691 have locks and avoids getattr rpc requests for it on every stat.
3694 Frequency : Always, for filesystems larger than 2TB
3696 Description: ldiskfs crash at mount for filesystem larger than 2TB with mballoc
3697 Details : Kenrel kmalloc limits allocations to 128kB and this prevents
3698 filesystems larger than 2TB to be mounted with mballoc enabled.
3701 Frequency : Always, for 32-bit kernel without CONFIG_LBD and filesystem > 2TB
3703 Description: ldiskfs crash at mount for filesystem larger than 2TB with mballoc
3704 Details : If a 32-bit kernel is compiled without CONFIG_LBD enabled and a
3705 filesystems larger than 2TB is mounted then the kernel will
3706 silently corrupt the start of the filesystem. CONFIG_LBD is
3707 enabled for all CFS-supported kernels, but the possibility of
3708 this happening with a modified kernel config exists.
3710 Severity : enhancement
3712 Description: add client O_DIRECT support for 2.6 kernels
3713 Details : It is now possible to do O_DIRECT reads and writes to files
3714 in the Lustre client mountpoint on 2.6 kernel clients.
3716 Severity : enhancement
3718 Description: parallel glimpse, setattr, statfs, punch, destroy requests
3719 Details : Sends glimpse, setattr, statfs, punch, destroy requests to OSTs in
3720 parallel, not waiting for response from every OST before sending
3721 a rpc to the next OST.
3726 Description: setattr vs write race when updating file timestamps
3727 Details : Client processes that update a file timestamp into the past
3728 right after writing to the file (e.g. tar) it is possible that
3729 the updated file modification time can be reset to the current
3730 time due to a race between processing the setattr and write RPC.
3732 Severity : enhancement
3734 Description: Bring 'lfs find' closer in line with regular Linux find.
3735 Details : lfs find util supports -atime, -mtime, -ctime, -maxdepth, -print,
3736 -print0 options and obtains all the needed info through the lustre
3739 Severity : enhancement
3741 Description: support up to 1024 configured devices on one node
3742 Details : change obd_dev array from statically allocated to dynamically
3743 allocated structs as they are first used to reduce memory usage
3748 Description: Flush dirty partially truncated pages during truncate
3749 Details : Immediatelly flush partially truncated pages in filter_setattr,
3750 this way we completely avoid having any pages in page cache on OST
3751 and can retire ugly workarounds during writes to flush such pages.
3756 Description: i_sem vs transaction deadlock in mds_obd_destroy during unlink.
3757 Details : protect inode from truncation within vfs_unlink() context
3758 just take a reference before calling vfs_unlink() and release it
3759 when parent's i_sem is free.
3764 Description: last_id value checked outside lock on OST caused LASSERT failure
3765 Details : If there were multiple MDS->OST object precreate requests in
3766 flight, it was possible that the OST's last object id was checked
3767 outside a lock and incorrectly tripped an assertion. Move checks
3768 inside locks, and discard old precreate requests.
3771 Frequency : always, if extents are used on OSTs
3773 Description: index ei_leaf_hi (48-bit extension) is not zeroed in extent index
3774 Details : OSTs using the extents format would not zero the high 16 bits of
3775 the index physical block number. This is not a problem for any
3776 OST filesystems smaller than 16TB, and no kernels support ext3
3777 filesystems larger than 16TB yet. This is fixed in 1.4.7 (all
3778 new/modified files) and can be fixed for existing filesystems
3779 with e2fsprogs-1.39-cfs1.
3784 Description: import connection selection may be incorrect if timer wraps
3785 Details : Using a 32-bit jiffies timer with HZ=1000 may cause backup
3786 import connections to be ignored if the 32-bit jiffies counter
3787 wraps. Use a 64-bit jiffies counter.
3790 Frequency : very large clusters immediately after boot
3792 Description: LNET request buffers exhausted under heavy short-term load
3793 Details : If a large number of client requests are generated on a service
3794 that has previously never seen so many requests it is possible
3795 that the request buffer growth cannot keep up with the spike in
3796 demand. Instead of dropping incoming requests, they are held in
3797 the LND until the RPC service can accept more requests.
3800 Frequency : Sometimes during replay
3802 Description: Assertion failure in ll_local_open after replay.
3803 Details : If replay happened on an open request reply before we were able
3804 to set replay handler, reply will become not swabbed tripping the
3805 assertion in ll_local_open. Now we set the handler right after
3806 recognising of open request
3809 Frequency : very rare
3811 Description: kernel reports "badness in vsnprintf"
3812 Details : Reading from the "recovery_status" /proc file in small chunks
3813 may cause a negative length in lprocfs_obd_rd_recovery_status()
3814 call to vsnprintf() (which is otherwise harmless). Exit early
3815 if there is no more space in the output buffer.
3817 Severity : enhancement
3819 Description: clear OBD RPC statistics by writing to them
3820 Details : It is now possible to clear the OBD RPC statistics by writing
3821 to the "stats" file.
3826 Description: Client mtime is not the same on different clients after utimes
3827 Details : In some cases, the client was using the utimes() syscall on
3828 a file cached on another node. The clients now validate the
3829 ctime from the MDS + OSTs to determine which one is right.
3834 Description: Inability to activate failout mode
3835 Details : lconf script incorrectly assumed that in pythong string's numeric
3836 value is used in comparisons.
3839 Frequency : always with multiple stripes per file
3841 Description: Inefficient object allocation for mutli-stripe files
3842 Details : When selecting which OSTs to stripe files over, for files with
3843 a stripe count that divides evenly into the number of OSTs,
3844 the MDS is always picking the same starting OST for each file.
3845 Return the OST selection heuristic to the original design.
3850 Description: mount failures may take full timeout to return an error
3851 Details : Under some heavy load conditions it is possible that a
3852 failed mount can wait for the full obd_timeout interval,
3853 possibly several minutes, before reporting an error.
3854 Instead return an error as soon as the status is known.
3856 Frequency : quota enabled and large files being deleted
3858 Description: releasing more than 4GB of quota at once hangs OST
3859 Details : If a user deletes more than 4GB of files on a single OST it
3860 will cause the OST to spin in an infinite loop. Release
3861 quota in < 4GB chunks, or use a 64-bit value for 1.4.7.1+.
3866 Description: statfs data retrieved from /proc may be stale or zero
3867 Details : When reading per-device statfs data from /proc, in the
3868 {kbytes,files}_{total,free,avail} files, it may appear
3869 as zero or be out of date.
3872 Frequency : systems with MD RAID1 external journal devices
3874 Description: lconf's call to blkid is confused by RAID1 journal devices
3875 Details : Use the "blkid -l" flag to locate the MD RAID device instead
3876 of returning all block devices that match the journal UUID.
3879 Frequency : always, for aggregate stripe size over 4GB
3881 Description: assertion fails when trying to use 4GB stripe size
3882 Details : Use "setstripe" to set stripe size over 4GB will fail the kernel,
3883 complaining "ASSERTION(lsm->lsm_xfersize != 0)"
3886 Frequency : always on ppc64
3888 Description: the first write on an ext3 filesystem with mballoc got stuck
3889 Details : ext3_mb_generate_buddy() uses find_next_bit() which does not
3890 perform endianness conversion.
3892 ------------------------------------------------------------------------------
3894 02-14-2006 Cluster File Systems, Inc. <info@clusterfs.com>
3896 * WIRE PROTOCOL CHANGE. This version of Lustre networking WILL NOT
3897 INTEROPERATE with older versions automatically. Please read the
3898 user documentation before upgrading any part of a live system.
3899 * WARNING: Lustre networking configuration changes are required with
3900 this release. See https://bugzilla.clusterfs.com/show_bug.cgi?id=10052
3903 * Support for kernels:
3904 2.6.9-22.0.2.EL (RHEL 4)
3905 2.6.5-7.244 (SLES 9)
3906 2.6.12.6 vanilla (kernel.org)
3909 Severity : enhancement
3910 Bugzilla : 7981/8208
3911 Description: Introduced Lustre Networking (LNET)
3912 Details : LNET is new networking infrastructure for Lustre, it includes
3913 a reorganized network configuration mode (see the user
3914 documentation for full details) as well as support for routing
3915 between different network fabrics. Lustre Networking Devices
3916 (LNDs) for the supported network fabrics have also been
3917 created for this new infrastructure.
3919 Severity : enhancement
3920 Description: Introduced Access control lists
3921 Details : clients can set ACLs on files and directories in order to have
3922 more fine-grained permissions than the standard Unix UGO+RWX.
3923 The MDS must be started with the "-o acl" mount option.
3925 Severity : enhancement
3926 Description: Introduced filesystem quotas
3927 Details : Administrators may now establish per-user quotas on the
3930 Severity : enhancement
3932 Description: Configuration change for the XT3
3933 The PTLLND is now used to run Lustre over Portals on the XT3
3934 The configure option(s) --with-cray-portals are no longer used.
3935 Rather --with-portals=<path-to-portals-includes> is used to
3936 enable building on the XT3. In addition to enable XT3 specific
3937 features the option --enable-cray-xt3 must be used.
3942 Description: Running on many-way SMP OSTs can trigger oops in llcd_send()
3943 Details : A race between allocating a new llcd and re-getting the llcd_lock
3944 allowed another thread to grab newly-allocated llcd.
3946 Severity : enhancement
3948 Description: 2.6 OST async journal commit and locking fix to improve performance
3949 Details : The filter_direct_io()+filter_commitrw_write() journal commits for
3950 2.6 kernels are now async as they already were in 2.4 kernels so
3951 that they can commit concurrently with the network bulk transfer.
3952 For block-allocated files the filter allocation semaphore is held
3953 to avoid filesystem fragmentation during allocation. BKL lock
3954 removed for 2.6 xattr operations where it is no longer needed.
3959 Description: lconf incorrectly determined whether two IP networks could talk
3960 Details : In some more complicated routing and multiple-network
3961 configurations, lconf will avoid trying to make a network
3962 connection to a disjoint part of the IP space. It was doing the
3963 math incorrectly for one set of cases.
3968 Description: Fix for potential infinite loop processing records in an llog.
3969 Details : If an llog record is corrupted/zeroed, it is possible to loop
3970 forever in llog_process(). Validate the llog record length
3971 and skip the remainder of the block on error.
3974 Frequency : occasional (liblustre only)
3976 Description: liblustre could not open files whose last component is a symlink
3977 Details : sysio_path_walk() would incorrectly pass the open intent to
3978 intermediate path components.
3981 Frequency : rare (liblustre only with non-standard tuning)
3982 Bugzilla : 7201 (7350)
3983 Description: Tuning the MDC DLM LRU size to zero triggers client LASSERT
3984 Details : llu_lookup_finish_locks() tries to set lock data on a lock
3985 after it has been released, only do this for referenced locks
3987 Severity : enhancement
3989 Description: specifying an (invalid) directory default stripe_size of -1
3990 would reset the directory default striping
3991 Details : stripe_size -1 was used internally to signal directory stripe
3992 removal, now use "all default" to signal dir stripe removal
3993 as a directory striping of "all default" is not useful
3996 Frequency : common for large clusters running liblustre clients
3998 Description: doing an ls when liblustre clients are running is slow
3999 Details : sending a glimpse AST to a liblustre client waits for every AST
4000 to time out, as liblustre clients will not respond. Since they
4001 cannot cache data we refresh the OST lock LVB from disk instead.
4003 Severity : enhancement
4005 Description: doing an ls at the same time as file IO can be slow
4006 Details : enqueue and other "small" requests can be blocked behind many
4007 large IO requests. Create a new OST IO portal for non-IO
4008 requests so they can be processed faster.
4011 Frequency : rare (only HPUX clients mounting unsupported re-exported NFS vol)
4013 Description: an HPUX NFS client would get -EACCESS when ftruncate()ing a newly
4014 created file with mode 000
4015 Details : the Linux NFS server relies on an MDS_OPEN_OWNEROVERRIDE hack to
4016 allow an ftruncate() as a non-root user to a file with mode 000.
4017 Lustre now respects this flag to disable mode checks when
4018 truncating a file owned by the user
4021 Frequency : liblustre-only, when liblustre client dies unexpectedly or becomes
4024 Description: Revoking locks from clients that went dead or catatonic might take
4026 Details : New lock flags FL_CANCEL_ON_BLOCK used by liblustre makes
4027 cancellation of such locks instant on servers without waiting for
4028 any reply from clients. Clients drops these locks when cancel
4029 notification from server is received without replying.
4032 Frequency : liblustre-only, when liblustre client dies or becomes busy
4034 Description: Doing ls on Linux clients can take a long time with active
4036 Details : Liblustre client cannot handle ASTs in timely manner, so avoid
4037 granting such locks to it in the first place if possible. Locks
4038 are taken by proxy on the OST during the read or write and
4039 dropped immediately afterward. Add connect flags handling, do
4040 not grant locks to liblustre clients for glimpse ASTs.
4042 Severity : enhancement
4044 Description: Improve read-ahead algorithm to avoid excessive IO for random reads
4045 Details : Existing read-ahead algorithm is tuned for the case of streamlined
4046 sequential reads and behaves badly with applications doing random
4047 reads. Improve it by reading ahead at least read region, and
4048 avoiding excessive large RPC for small reads.
4050 Severity : enhancement
4052 Description: Creating more than 1000 files for a single job may cause a load
4053 imbalance on the OSTs if there are also a large number of OSTs.
4054 Details : qos_prep_create() uses an OST index reseed value that is an
4055 even multiple of the number of available OSTs so that if the
4056 reseed happens in the middle of the object allocation it will
4057 still utilize the OSTs as uniformly as possible.
4062 Description: OST or MDS may oops in ping_evictor_main()
4063 Details : ping_evictor_main() drops obd_dev_lock if deleting a stale export
4064 but doesn't restart at beginning of obd_exports_timed list
4067 Severity : enhancement
4069 Description: improve by-nid export eviction on the MDS and OST
4070 Details : allow multiple exports with the same NID to be evicted at one
4071 time without re-searching the exports list.
4074 Frequency : rare, only with supplementary groups enabled on SMP 2.6 kernels
4076 Description: MDS may oops in groups_free()
4077 Details : in rare race conditions a newly allocated group_info struct is
4078 freed again, and this can be NULL. The 2.4 compatibility code
4079 for groups_free() checked for a NULL pointer, but 2.6 did not.
4082 Frequency : common for liblustre clients doing little filesystem IO
4083 Bugzilla : 9352, 7313
4084 Description: server may evict liblustre clients accessing contended locks
4085 Details : if a client is granted a lock or receives a completion AST
4086 with a blocking AST already set it would not reply to the AST
4087 for LDLM_FL_CANCEL_ON_BLOCK locks. It now replies to such ASTs.
4090 Frequency : lfs setstripe, only systems with more than 160 OSTs
4092 Description: unable to set striping with a starting offset beyond OST 160
4093 Details : llapi_create_file() incorrectly limited the starting stripe
4094 index to the maximum single-file stripe count.
4097 Frequency : LDAP users only
4099 Description: lconf did not handle in-kernel recovery with LDAP properly
4100 Details : lconf/LustreDB get_refs() is searching the wrong namespace
4102 Severity : enhancement
4104 Description: bind OST threads to NUMA nodes to improve performance
4105 Details : all OST threads are uniformly bound to CPUs on a single NUMA
4106 node and do their allocations there to localize memory access
4108 Severity : enhancement
4110 Description: llmount can determine client NID directly from Myrinet (GM)
4111 Details : the client NID code from gmnalnid was moved directly into
4112 llmount, removing the need to use this or specifying the
4113 client NID explicitly when mounting GM clients with zeroconf
4116 Frequency : if client is started with down MDS
4118 Description: if client is started with down MDS mount hangs in ptlrpc_queue_wait
4119 Details : Having an LWI_INTR() wait event (interruptible, but no timeout)
4120 will wait indefinitely in ptlrpc_queue_wait->l_wait_event() after
4121 ptlrpc_import_delayed_req() because we didn't check if the
4122 request was interrupted, and we also didn't break out of the
4123 event loop if there was no timeout
4128 Description: data loss during non-page-aligned writes to a single file from
4129 both multiple nodes and multiple threads on one node at same time
4130 Details : updates to KMS and lsm weren't protected by common lock. Resulting
4131 inconsistency led to false short-reads, that were cached and later
4132 used by ->prepare_write() to fill in partially written page,
4133 leading to data loss.
4136 Frequency : always, if lconf --abort_recovery used
4138 Description: lconf --abort_recovery fails with 'Operation not supported'
4139 Details : lconf was attempting to abort recovery on the MDT device and not
4142 Severity : enhancement
4144 Description: remove cleanup logs
4145 Details : replace lconf-generated cleanup logs with lustre internal
4146 cleanup routines. Eliminates the need for client-cleanup and
4149 Severity : enhancement
4151 Description: add support for EAs (user and system) on lustre filesystems
4152 Details : it is now possible to store extended attributes in the Lustre
4153 client filesystem, and with the user_xattr mount option it
4154 is possible to allow users to store EAs on their files also
4156 Severity : enhancement
4158 Description: Add possibility (config option) to show minimal available OST free
4160 Details : When compiled with --enable-mindf configure option, statfs(2)
4161 (and so, df) will return least minimal free space available from
4162 all OSTs as amount of free space on FS, instead of summary of
4163 free spaces of all OSTs.
4165 Severity : enhancement
4167 Description: do not expand extent locks acquired on OST-side
4168 Details : Modify ldlm_extent_policy() to not expand local locks, acquired
4169 by server: they are not cached anyway.
4172 Frequency : when mmap is used/binaries executed from Lustre
4174 Description: Unmmap pages before throwing them away from read cache.
4175 Details : llap_shrink cache now attempts to unmap pages before discarding
4176 them (if unmapping failed - do not discard). SLES9 kernel has
4177 extra checks that trigger if this unmapping is not done first.
4182 Description: lconf didn't resolve symlinks before checking to see whether a
4183 given mountpoint was already in use
4186 Frequency : when migrating failover services
4187 Bugzilla : 6395, 9514
4188 Description: When migrating a subset of services from a node (e.g. failback
4189 from a failover service node) the remaining services would
4190 time out and evict clients.
4191 Details : lconf --force (implied by --failover) sets the global obd_timeout
4192 to 5 seconds in order to quickly disconnect, but this caused
4193 other RPCs to time out too quickly. Do not change the global
4194 obd_timeout for force cleanup, only set it for DISCONNECT RPCs.
4196 Severity : enhancement
4197 Frequency : if MDS is started with down OST
4198 Bugzilla : 9439,5706
4199 Description: Allow startup/shutdown of an MDS without depending on the
4200 availability of the OSTs.
4201 Details : Asynchronously call mds_lov_synchronize during MDS startup.
4202 Add appropriate locking and lov-osc refcounts for safe
4203 cleaning. Add osc abort_inflight calls in case the
4204 synchronize never started.
4207 Frequency : occasional (Cray XT3 only)
4209 Description: root not authorized to access files in CRAY_PORTALS environment
4210 Details : The client process capabilities were not honoured on the MDS in
4211 a CRAY_PORTALS/CRAY_XT3 environment. If the file had previously
4212 been accessed by an authorized user then root was able to access
4213 the file on the local client also. The root user capabilities
4214 are now allowed on the MDS, as this environment has secure UID.
4217 Frequency : occasional
4219 Description: ldiskfs "too long searching" message happens too often
4220 Details : A debugging message (otherwise harmless) prints too often on
4221 the OST console. This has been reduced to only happen when
4222 there are fragmentation problems on the filesystem.
4227 Description: Division by zero in statfs when all OSCs are inactive
4228 Details : lov_get_stripecnt() returns zero due to incorrect order of checks,
4229 lov_statfs divides by value returned by lov_get_stripecnt().
4233 Bugzilla : 9489, 3273
4234 Description: First write from each client to each OST was only 4kB in size,
4235 to initialize client writeback cache, which caused sub-optimal
4236 RPCs and poor layout on disk for the first writen file.
4237 Details : Clients now request an initial cache grant at (re)connect time
4238 and so that they can start streaming writes to the cache right
4239 away and always do full-sized RPCs if there is enough data.
4240 If the OST is rebooted the client also re-establishes its grant
4241 so that client cached writes will be honoured under the grant.
4246 Description: Slow ls (and stat(2) syscall) on files residing on IO-loaded OSTs
4247 Details : Now I/O RPCs go to different portal number and (presumably) fast
4248 lock requests (and glimses) and other RPCs get their own service
4249 threads pool that should be able to service those RPCs
4252 Severity : enhancement
4254 Description: Ability to exchange lustre version between client and servers and
4255 issue warnings at client side if client is too old. Also for
4256 liblustre clients there is ability to refuse connection of too old
4258 Details : New 'version' field is added to connect data structure that is
4259 filled with version info. That info is later checked by server and
4263 Frequency : rare, liblustre only.
4264 Bugzilla : 9296, 9581
4265 Description: Two simultaneous writes from liblustre at offset within same page
4266 might proceed at the same time overwriting eachother with stale
4268 Details : I/O lock withing llu_file_prwv was released too early, before data
4269 actually was hitting the wire. Extended lock-holding time until
4270 server acknowledges receiving data.
4273 Frequency : extremely rare. Never observed in practice.
4275 Description: avoid generating lustre_handle cookie of 0.
4276 Details : class_handle_hash() generates handle cookies by incrementing
4277 global counter, and can hit 0 occasionaly (this is unlikely, but
4278 not impossible, because initial value of cookie counter is
4279 selected randonly). Value of 0 is used as a sentinel meaning
4280 "unassigned handle" --- avoid it. Also coalesce two critical
4281 sections in this function into one.
4283 Severity : enhancement
4285 Description: allow liblustre clients to delegate truncate locking to OST
4286 Details : To avoid overhead of locking, liblustre client instructs OST to
4287 take extent lock in ost_punch() on client's behalf. New connection
4288 flag is added to handle backward compatibility.
4290 Severity : enhancement
4291 Bugzilla : 4928, 7341, 9758
4292 Description: allow number of OST service threads to be specified
4293 Details : a module parameter allows the number of OST service threads
4294 to be specified via "options ost ost_num_threads={N}" in the
4295 OSS's /etc/modules.conf or /etc/modprobe.conf.
4299 Bugzilla : 6146, 9635, 9895
4300 Description: servers crash with bad pointer in target_handle_connect()
4301 Details : In rare cases when a client is reconnecting it was possible that
4302 the connection request was the last reference for that export.
4303 We would temporarily drop the export reference and get a new
4304 one, but this may have been the last reference and the export
4305 was just destroyed. Get new reference before dropping old one.
4307 Severity : enhancement
4308 Frequency : if client is started with failover MDS
4310 Description: Allow multiple MDS hostnames in the mount command
4311 Details : Try to read the configuration from all specified MDS
4312 hostnames during a client mount in case the "primary"
4315 Severity : enhancement
4317 Description: Stop sending data to evicted clients as soon as possible.
4318 Details : Check if the client we are about to send or are sending data to
4319 was evicted already. (Check is done every second of waiting,
4320 for which l_wait_event interface was extended to allow checking
4321 of exit condition at specified intervals).
4324 Frequency : rare, normally only when NFS exporting is done from client
4326 Description: 'bad disk LOV MAGIC: 0x00000000' error when chown'ing files
4328 Details : Make mds_get_md() recognise empty md case and set lmm size to 0.
4331 Frequency : always, if srand() is called before liblustre initialization
4333 Description: Liblustre uses system PRNG disturbing its usage by user application
4334 Details : Introduce internal to lustre fast and high-quality PRNG for
4335 lustre usage and make liblustre and some other places in generic
4336 lustre code to use it.
4338 Severity : enhancement
4339 Bugzilla : 9477, 9557, 9870
4340 Description: Verify that the MDS configuration logs are updated when xml is
4341 Details : Check if the .xml configuration logs are newer than the config
4342 logs stored on the MDS and report an error if this is the case.
4343 Request --write-conf, or allow starting with --old_conf.
4345 Severity : enhancement
4347 Description: Handle symlinks in the path when checking if Lustre is mounted.
4348 Details : Resolve intermediate symlinks when checking if a client has
4349 mounted a filesystem to avoid duplicate client mounts.
4354 Description: lconf can hit an error exception but still return success.
4355 Details : The lconf command catches the Command error exception at the top
4356 level script context and will exit with the associated exit
4357 status, but doesn't ensure that this exit status is non-zero.
4362 Description: failure of ptlrpc thread startup can cause oops
4363 Details : Starting a ptlrpc service thread can fail if there are a large
4364 number of threads or the server memory is very fragmented.
4365 Handle this without oopsing.
4368 Frequency : always, only if liblustre and non-default acceptor port was used
4370 Description: liblustre cannot connect to servers with non-default acceptor port
4371 Details : tcpnal_set_default_params() was not called and was therefore
4372 ignoring the environment varaible TCPNAL_PORT, as well as other
4373 TCPNAL_ environment variables
4378 Description: two objects could be created on the same OST for a single file
4379 Details : If an OST is down, in some cases it was possible to create two
4380 objects on a single OST for a single file. No problems other
4381 than potential performance impact and spurious error messages.
4385 Bugzilla : 5681, 9562
4386 Description: Client may oops in ll_unhash_aliases
4387 Details : Client dcache may become inconsistent in race condition.
4388 In some cases "getcwd" can fail if the current directory is
4394 Description: Inode refcounting problems in NFS export code
4395 Details : link_raw functions used to call d_instantiate without obtaining
4396 extra inode reference first.
4400 Bugzilla : 9942, 9903
4401 Description: Referencing freed requests leading to crash, memleaks with NFS.
4402 Details : We used to require that call to ll_revalidate_it was always
4403 followed by ll_lookup_it. Also with revalidate_special() it is
4404 possible to call ll_revalidate_it() twice for the same dentry
4405 even if first occurence returned success. This fix changes semantic
4406 between DISP_ENQ_COMPLETE disposition flag to mean there is extra
4407 reference on a request referred from the intent.
4408 ll_intent_release() then releases such a request.
4411 Frequency : rare, normally benchmark loads only
4413 Description: unlinked inodes were kept in memory on the client
4414 Details : If a client is repeatedly creating and unlinking files it
4415 can accumulate a lot of stale inodes in the inode slab cache.
4416 If there is no other client load running this can cause the
4417 client node to run out of memory. Instead flush old inodes
4418 from client cache that have the same inode number as a new inode.
4421 Frequency : SLES9 2.6.5 kernel and long filenames only
4422 Bugzilla : 9969, 10379
4423 Description: utime reports stale NFS file handle
4424 Details : SLES9 uses out-of-dentry names in some cases, which confused
4425 the lustre dentry revalidation. Change it to always use the
4429 Frequency : rare, unless heavy write-truncate concurrency is continuous
4430 Bugzilla : 4180, 6984, 7171, 9963, 9331
4431 Description: OST becomes very slow and/or deadlocked during object unlink
4432 Details : filter_destroy() was holding onto the parent directory lock
4433 while truncating+unlinking objects. For very large objects this
4434 may block other threads for a long time and slow overall OST
4435 responsiveness. It may also be possible to get a lock ordering
4436 deadlock in this case, or run out of journal credits because of
4437 the combined truncate+unlink. Solution is to do object truncate
4438 first in one transaction without parent lock, and then do the
4439 final unlink in a new transaction with the parent lock. This
4440 reduces the lock hold time dramatically.
4443 Frequency : rare, 2.4 kernels only
4445 Description: MDS or OST cleanup may trip kernel BUG when dropping kernel lock
4446 Details : mds_cleanup() and filter_cleanup() need to drop the kernel lock
4447 before unmounting their filesystem in order to avoid deadlock.
4448 The kernel_locked() function in 2.4 kernels only checks whether
4449 the kernel lock is held, not whether it is this process that is
4450 holding it as 2.6 kernels do.
4455 Description: MDS or OST may oops/LBUG if a client is connecting multiple times
4456 Details : The client ptlrpc code may be trying to reconnect to a down
4457 server before a previous connection attempt has timed out.
4458 Increase the reconnect interval to be longer than the connection
4459 timeout interval to avoid sending duplicate connections to
4463 Frequency : echo_client brw_test command
4465 Description: fix echo_client to work with OST preallocated code
4466 Details : OST preallocation code (5137) didn't take echo_client IO path
4467 into account: echo_client calls filter methods outside of any
4468 OST thread and, hence, there is no per-thread preallocated
4469 pages and buffers to use. Solution: hijack pga pages for IO. As
4470 a byproduct, this avoids unnecessary data copying.
4474 Bugzilla : 3555, 5962, 6025, 6155, 6296, 9574
4475 Description: Client can oops in mdc_commit_close() after open replay
4476 Details : It was possible for the MDS to return an open request with no
4477 transaction number in mds_finish_transno() if the client was
4478 evicted, but without actually returning an error. Clients
4479 would later try to replay that open and may trip an assertion
4480 Simplify the client close codepath, and always return an error
4481 from the MDS in case the open is not successful.
4484 Frequency : rare, 2.6 OSTs only
4486 Description: OST may deadlock under high load on fragmented files
4487 Details : If there was a heavy load and highly-fragmented OST filesystems
4488 it was possible to have all the OST threads deadlock waiting on
4489 allocation of biovecs, because the biovecs were not released
4490 until the entire RPC IO was completed. Instead, release biovecs
4491 as soon as they are complete to ensure forward IO progress.
4493 Severity : enhancement
4495 Description: Support for specifying external journal device at mount
4496 Details : If an OST or MDS device is formatted with an external journal
4497 device, this device major/minor is stored in the ext3 superblock
4498 and may not be valid for failover. Allow detecting and
4499 specifying the external journal at mount time.
4504 Description: Mounting an MDS with pending unlinked files may cause oops
4505 Details : target_finish_recovery() calls mds_postrecov() which returned
4506 the number of orphans unlinked. mds_lov_connect->mds_postsetup()
4507 considers this an error and immediately begins cleaning up the
4508 lov, just after starting the mds_lov process
4510 Severity : enhancement
4512 Description: Implement 'lfs df' to report actual free space on per-OST basis
4513 Details : Add sub-command 'df' on 'lfs' to report the disk space usage of
4514 MDS/OSDs. Usage: lfs df [-i][-h]. Command Options: '-i' to report
4515 usage of objects; '-h' to report in human readable format.
4517 ------------------------------------------------------------------------------
4519 08-26-2005 Cluster File Systems, Inc. <info@clusterfs.com>
4526 Description: Mounting an ldiskfs file system with mballoc may crash OST node.
4527 Details : ldiskfs mballoc code may reference an uninitialized buddy struct
4528 at startup during orphan unlinking. Instead, skip buddy update
4529 before setup, as it will be regenerated after recovery is complete.
4534 Description: If an OST is inactive, its locks might reference stale inodes.
4535 Details : lov_change_cbdata() must iterate over all namespaces, even if
4536 they are inactive to clear inode references from the lock.
4538 Severity : enhancement
4539 Frequency : occasional, if non-standard max_dirty_mb used
4541 Description: Client will block write RPCs if not enough grant
4542 Details : If a client has max_dirty_mb smaller than max_rpcs_in_flight,
4543 then the client will block writes while waiting for another RPC
4544 to complete instead of consuming its dirty limit. With change
4545 we get improved performance when max_dirty_mb is small.
4547 Severity : enhancement
4548 Bugzilla : 3389, 6253
4549 Description: Add support for supplementary groups on the MDS.
4550 Details : The MDS has an upcall /proc/fs/lustre/mds/{mds}/group_upcall
4551 (set to /usr/sbin/l_getgroups if enabled) which will do MDS-side
4552 lookups for user supplementary groups into a cache.
4556 Description: O_CREAT|O_EXCL open flags in liblustre always return -EEXIST
4557 Details : Make libsysio to not enforce O_EXCL by clearing the flag,
4558 for liblustre O_EXCL is enforced by MDS.
4562 Description: readdir never returns NULL in liblustre.
4563 Details : Corrected llu_iop_getdirentries logic, to return offset of next
4564 dentry in struct dirent.
4568 Frequency : liblustre only, depends on application IO pattern
4569 Description: liblustre clients evicted if not contacting servers
4570 Details : Don't put liblustre clients into the ping_evictor list, so
4571 they will not be evicted by the pinger ever.
4573 Severity : enhancement
4575 Description: Add ability to evict clients by NID from MDS.
4576 Details : By echoing "nid:$NID" string into
4577 /proc/fs/lustre/mds/.../evict_client client with nid that equals to
4578 $NID would be instantly evicted from this MDS and from all active
4579 OSTs connected to it.
4583 Description: Do not query file size twice, somewhat slowing stat(2) calls.
4584 Details : lookup_it_finish() used to query file size from OSTs that was not
4589 Description: service threads change working directory to that of init
4590 Details : Starting lustre service threads may pin the working directory
4591 of the parent thread, making that filesystem busy. Threads
4592 now change to the working directory of init to avoid this.
4596 Frequency : during shutdown only
4597 Description: shutdown with a failed MDS or OST can cause unmount to hang
4598 Details : Don't resend DISCONNECT messages in ptlrpc_disconnect_import()
4603 Frequency : 2.6 only
4604 Description: chmod/chown may include an extra supplementary group
4605 Details : ll{,u}_mdc_pack_op_data() does not properly initialize the
4606 supplementary group and if none is specified this is used.
4609 Bugzilla : 5479 (6816)
4611 Description: Racing open + rm can assert client in mdc_set_open_replay_data()
4612 Details : If lookup is in progress on a file that is unlinked we might try
4613 to revalidate the inode and fail in revalidate after lookup is
4614 complete and ll_file_open() enqueues the open again but
4615 it_open_error() was not checking DISP_OPEN_OPEN errors correctly.
4618 Frequency : always, if lconf --abort_recovery used
4620 Description: lconf --abort_recovery fails with 'Operation not supported'
4621 Details : lconf was attempting to abort recovery on the MDT device and not
4624 ------------------------------------------------------------------------------
4626 2005-08-08 Cluster File Systems, Inc. <info@clusterfs.com>
4631 Frequency : rare (only unsupported configurations with a node running as an
4633 Bugzilla : 6514, 5137
4634 Description: Mounting a Lustre file system on a node running as an OST could
4636 Details : OSTs now preallocates memory needed to write out data at
4637 startup, instead of when needed, to avoid having to
4638 allocate memory in possibly low memory situations.
4639 Specifically, if the file system is mounted on on OST,
4640 memory pressure could force it to try to write out data,
4641 which it needed to allocate memory to do. Due to the low
4642 memory, it would be unable to do so and the node would
4643 become unresponsive.
4645 Severity : enhancement
4647 Description: Addition of lconf --service command line option
4648 Details : lconf now accepts a '--service <arg>' option, which is
4649 shorthand for 'lconf --group <arg> --select <arg>=<hostname>'
4651 Severity : enhancement
4653 Description: Failover mode is now the default for OSTs.
4654 Details : By default, OSTs will now run in failover mode. To return to
4655 the old behaviour, add '--failout' to the lmc line for OSTs.
4657 Severity : enhancement
4659 Description: Health checks are now provided for MDS and OSTs
4660 Details : Additional detailed health check information on MSD and OSTs
4661 is now provided through the procfs health_check value.
4664 Frequency : occasional, depends on IO load
4666 Description: Disk fragmentation on the OSTs could eventually cause slowdowns
4667 after numerous create/delete cycles
4668 Details : The ext3 inode allocation policy would not allocate new inodes
4669 very well on the OSTs because there are no new directories
4670 being created. Instead we look for groups with free space if
4671 the parent directories are nearly full.
4676 Description: Network or server problems during mount may cause partially
4677 mounted clients instead of returning an error.
4678 Details : The config llog parsing code may overwrite the error return
4679 code during mount error handling, returning success instead
4685 Description: MDS can fail to allocate large reply buffers
4686 Details : After long uptimes the MDS can fail to allocate large reply
4687 buffers (e.g. zconf client mount config records) due to memory
4688 fragmentation or consumption by the buffer cache. Preallocate
4689 some large reply buffers so that these replies can be sent even
4690 under memory pressure.
4694 Frequency : rare (liblustre)
4695 Description: fsx running with liblustre complained that using truncate() to
4696 extend the file doesn't work. This patch corrects that issue.
4697 Details : This is the liblustre equivalent of the fix for bug 6196. Fixes
4698 ATTR_SIZE and lsm use in llu_setattr_raw.
4702 Frequency : rare, only 2.6 kernels
4703 Description: Unusual file access patterns on the MDS may result in inode
4704 data being lost in very rare circumstances.
4705 Details : Bad interaction between the ea-in-inode patch and the "no-read"
4706 code in the 2.6 kernel caused the inode and/or EA data not to
4707 be read from disk, causing single-file corruption.
4711 Frequency : rare, only 2.6 filesystems using extents
4712 Description: Heavy concurrent write and delete load may cause data corruption.
4713 Details : It was possible under high-load situations to have an extent
4714 metadata block in the block device cache from a just-unlinked
4715 file overwrite a newly-allocated data block. We now unmap any
4716 metadata buffers that alias just-allocated data blocks.
4720 Frequency : filesystems with default stripe_count larger than 77
4721 Description: lconf+mke2fs fail when formatting filesystem with > 77 stripes
4722 Details : lconf specifies an inode size of 4096 bytes when the default
4723 stripe_count is larger than 77. This conflicts with the default
4724 inode density of 1 per 4096 bytes. Allocate smaller inodes in
4725 this case to avoid pinning too much memory for large EAs.
4727 ------------------------------------------------------------------------------
4729 2005-07-07 Cluster File Systems, Inc. <info@clusterfs.com>
4734 Frequency : rare (extremely heavy IO load with hundreds of clients)
4736 Description: Client is evicted, gets IO error writing to file
4737 Details : lock ordering changes for bug 5492 reintroduced bug 3267 and
4738 caused clients to be evicted for AST timeouts. The fixes in
4739 bug 5192 mean we no longer need to have such short AST timeouts
4740 so ldlm_timeout has been increased.
4743 Frequency : occasional during --force or --failover shutdown under load
4744 Bugzilla : 5949, 4834
4745 Description: Server oops/LBUG if stopped with --force or --failover under load
4746 Details : a collection of import/export refcount and cleanup ordering
4747 issues fixed for safer force cleanup
4750 Frequency : only filesystems larger than 120 OSTs
4751 Bugzilla : 5990, 6223
4752 Description: lfs getstripe would oops on a very large filesystem
4753 Details : lov_getconfig used kfree on vmalloc'd memory
4756 Frequency : only filesystems exporting via NFS to Solaris 10 clients
4757 Bugzilla : 6242, 6243
4758 Description: reading from files that had been truncated to a non-zero size
4759 but never opened returned no data
4760 Details : ll_file_read() reads zeros from no-object files to EOF
4765 Description: A bug in MDS/OSS recovery could cause the OSS to fail an assertion
4766 Details : There's little harm in aborting MDS/OSS recovery and letting it
4767 try again, so I removed the LASSERT and return an error instead.
4769 Severity : enhancement
4771 Description: New debugging infrastructure for tracking down data corruption
4772 Details : The I/O checksum code was replaced to: (a) control it at runtime,
4773 (b) cover more of the client-side code path, and (c) try to narrow
4774 down where problems occurred
4778 Bugzilla : 3819, 4364, 4397, 6313
4779 Description: Racing close and eviction MDS could cause assertion in mds_close
4780 Details : It was possible to get multiple mfd references during close and
4781 client eviction, leading to one thread referencing a freed mfd.
4783 Severity: : enhancement
4784 Bugzilla : 3262, 6359
4785 Description: Attempts to reconnect to servers are now more aggressive.
4786 Details : This builds on the enhanced upcall-less recovery that was added
4787 in 1.4.2. When trying to reconnect to servers, clients will
4788 now try each server in the failover group every 10 seconds. By
4789 default, clients would previously try one server every 25 seconds.
4794 Description: After recovery, certain operations trigger a failed
4795 assertion on a client.
4796 Details : Failing over an mds, using lconf -d --failover, while a
4797 client was doing a readdir() call would cause the client to
4798 LBUG after recovery completed and the readdir() was resent.
4800 Severity : enhancement
4802 Description: Default groups are now added by lconf
4803 Details : You can now run lconf --group <servicename> without having to
4804 manually add groups with lmc.
4807 Frequency : occasional
4809 Description: Nodes with an elan id of 0 trigger a failed assertion
4812 Frequency : always when accessing e.g. tty/console device nodes
4814 Description: tty and some other devices nodes cannot be used on lustre
4815 Details : file's private_data field is used by device data and lustre
4816 values in there got lost. New field was added to struct file to
4817 store fs-specific private data.
4820 Frequency : when exporting Lustre via NFS
4822 Description: NFSD failed occasionally when looking up a path component
4823 Details : NFSD is looking up ".." which was broken in ext3 directories
4824 that had grown large enough to become hashed.
4827 Frequency : Clusters with multiple interfaces not on the same subnet
4829 Description: Nodes will repeatedly try to reconnect to an interface which it
4830 cannot reach and report an error to the log.
4831 Details : Extra peer list entries will be created by lconf with some peers
4832 unreachable. lconf now validates the peer before adding it.
4835 Frequency : Only if a default stripe is set on the filesystem root.
4837 Description: Setting a default stripe on the filesystem root prevented the
4838 filesystem from being remounted.
4839 Details : The client was sending extra request flags in the root getattr
4840 request and did not allocate a reply buffer for the dir EA.
4843 Frequency : occasional, higher if lots of files are accessed by one client
4844 Bugzilla : 6159, 6097
4845 Description: Client trips assertion regarding lsm mismatch/magic
4846 Details : While revalidating inodes the VFS looks up inodes with ifind()
4847 and in rare cases can find an inode that is being freed.
4848 The ll_test_inode() code will free the lsm during ifind()
4849 when it finds an existing inode and then the VFS later attaches
4850 this free lsm to a new inode.
4854 Bugzilla : 6422, 7030
4855 Description: MDS deadlock between mkdir and client eviction
4856 Details : Creating a new file via mkdir or mknod (starting a transaction
4857 and getting the ns lock) can deadlock with client eviction
4858 (gets ns lock and trying to finish a synchronous transaction).
4861 Frequency : occasional
4862 Description: While starting a server, the fsfilt_ext3 module could not be
4864 Details : CFS's improved ext3 filesystem is named ldiskfs for 2.6
4865 kernels. Previously, lconf would still use the ext3 name
4866 when trying to load modules. Now, it will correctly use
4867 ext3 on 2.4 and ldiskfs on 2.6.
4869 Severity : enhancement
4870 Description: The default stripe count has been changed to 1
4871 Details : The interpretation of the default stripe count (0, to lfs
4872 or lmc) has been changed to mean striping across a single
4873 OST, rather than all available. For general usage we have
4874 found a stripe count of 1 or 2 works best.
4876 Severity : enhancement
4877 Description: Add support for compiling against Cray portals.
4878 Details : Conditional compiling for some areas that are different
4882 Frequency : occasional
4883 Bugzilla : 6409, 6834
4884 Description: Creating files with an explicit stripe count may lead to
4885 a failed assertion on the MDS
4886 Details : If some OSTs are full or unavailable, creating files may
4887 trigger a failed assertion on the MDS. Now, Lustre will
4888 try to use other servers or return an error to the
4892 Frequency : occasional
4894 Description: Multiple concurrent overlapping read+write on multiple SMP nodes
4895 caused lock timeout during readahead (since 1.4.2).
4896 Details : Processes doing readahead might match a lock that hasn't been
4897 granted yet if there are overlapping and conflicting lock
4898 requests. The readahead process waits on ungranted lock
4899 (original lock is CBPENDING), while OST waits for that process
4900 to cancel CBPENDING read lock and eventually evicts client.
4902 Severity : enhancement
4904 Description: Initial enabling of flock support for clients
4905 Details : Implements fcntl advisory locking and file status functions.
4906 This feature is provided as an optional mount flag (default
4907 off), and is NOT CURRENTLY SUPPORTED. Not all types of record
4908 locking are implemented yet, and those that are are not guaranteed
4909 to be completely correct in production environments.
4910 mount -t lustre -o [flock|noflock] ...
4913 Frequency : occasional
4915 Description: OSTs running 2.4 kernels but with extents enabled might trip an
4916 assertion in the ext3 JBD (journaling) layer.
4917 Details : The b_committed_data struct is protected by the big kernel lock
4918 in 2.4 kernels, serializing journal_commit_transaction() and
4919 ext3_get_block_handle->ext3_new_block->find_next_usable_block()
4920 access to this struct. In 2.6 kernels there is finer grained
4921 locking to improve SMP performance of the JBD layer.
4925 Description: Changes the "SCSI I/O Stats" kernel patch to default to "enabled"
4930 Description: merge and cleanup kernel patches.
4933 -----------------------------------------------------------------------------
4935 2005-05-05 Cluster File Systems, Inc. <info@clusterfs.com>
4937 NOTE: Lustre 1.4.2 uses an incompatible network protocol than previous
4938 versions of Lustre. Please update all servers and clients to
4939 version 1.4.2 or later at the same time. You must also run
4940 "lconf --write-conf {config}.xml" on the MDS while it is stopped
4941 to update the configuration logs.
4943 - fix for HPUX NFS client breakage when NFS exporting Lustre (5781)
4944 - mdc_enqueue does not need max_mds_easize request buffer on send (5707)
4945 - swab llog records of type '0' so we get proper header size/idx (5861)
4946 - send llog cancel req to DLM cancel portal instead of cb portal (5515)
4947 - fix rename of one directory over another leaking an inode (5953)
4948 - avoid SetPageDirty on 2.6 (5981)
4949 - don't re-add just-being-destroyed locks to the waiting list (5653)
4950 - when creating new directories, inherit the parent's custom
4951 striping settings if present parent (3048)
4952 - flush buffers from cache before direct IO in 2.6 obdfilter (4982)
4953 - don't hold i_size_sem in ll_nopage() and ll_ap_refresh_count (6077)
4954 - don't hold client locks on temporary worklist from l_lru (5666)
4955 - handle IO errors in 2.6 obdfilter bio completion routine (6046)
4956 - automatically evict dead clients (5921)
4957 - Update file size properly in create+truncate+fstat case (6196)
4958 - Do not unhash mountpoint dentries, do not allow removal of
4960 - Avoid lock ordering deadlock issue with write/truncate (6203,5654)
4961 - reserve enough journal credits in fsfilt_start_log for setattr (4554)
4962 - ldlm_enqueue freed-export error path would always LBUG (6149,6184)
4963 - don't reference lr_lvb_data until after we hold lr_lvb_sem (6170)
4964 - don't overwrite last_rcvd if there is a *_client_add() error (6086)
4965 - Correctly handle reads of files with no objects (6243)
4966 - lctl recover will also mark a device active if deactivate used (5933)
4968 - by default create 1 inode per 4kB space on MDS, per 16kB on OSTs
4969 - allow --write-conf on an MDS with different nettype than client (5619)
4970 - don't write config llogs to MDS for mounts not from that MDS (5617)
4971 - lconf should create multiple TCP connections from a client (5201)
4972 - init scripts are now turned off by default; run chkconfig --on
4973 lustre and chkconfig --on lustrefs to use them
4974 - upcalls are no longer needed for clients to recover to failover
4976 - add --abort-recovery option to lconf to abort recovery on device
4978 - add support for an arbitrary number of OSTs (3026)
4979 - Quota support protocol changes.
4980 - forward compatibility changes to wire structs (6007)
4981 - rmmod NALs that might be loaded because of /etc/modules.conf (6133)
4982 - support for mountfsoptions and clientoptions to the Lustre LDAP (5873)
4983 - improved "lustre status" script
4984 - initialize blocksize for non-regular files (6062)
4985 - added --disable-server and --disable-client configure options (5782)
4986 - introduce a lookup cache for lconf to avoid repeated DB scans (6204)
4987 - Vanilla 2.4.29 support
4988 - increase maximum number of obd devices to 520 (6242)
4989 - remove the tcp-zero-copy patch from the suse-2.4 series (5902)
4990 - Quadrics Elan drivers are now included for the RHEL 3 2.4.21 and
4991 SLES 9 2.6.5 kernels
4992 - limit stripes per file to 160 (the maximum EA size) (6093)
4994 2005-03-22 Cluster File Systems, Inc. <info@clusterfs.com>
4997 - don't LASSERT in ll_release on NULL lld with NFS export (4655, 5760)
4998 - hold NS lock when calling handle_ast_error->del_waiting_lock (5746)
4999 - fix setattr mtime regression from lovcleanup merge (4829, 5669)
5000 - workaround for 2.6 crash in ll_unhash_aliases (5687, 5210)
5001 - small ext3 extents cleanups and fixes (5733)
5002 - improved mballoc code, several small races and bugs fixed (5733, 5638)
5003 - kernel version 43 - fix remove_suid bugs in both 2.4 and 2.6 (5695)
5004 - avoid needless client->OST connect, fix handle mismatch (5317)
5005 - fix DLM error path that led to out-of-sync client, long delays (5779)
5006 - support common vfs-enforced mount options (nodev,nosuid,noexec) (5637)
5007 - fix several locking issues related to i_size (5492,5624,5654,5672)
5008 - don't move pending lock onto export if it is already evicted (5683)
5009 - fix kernel oops when creating .foo in unlinked directory (5548)
5010 - fix deadlock in obdfilter statistics vs. object create (5811)
5011 - use time_{before,after} to avoid timer jiffies wrap (5882)
5012 - shutdown --force/--failover stability (3607,3651,4797,5203,4834)
5013 - Do not leak request if server was not able to process it (5154)
5014 - If mds_open unable to find parent dir, make that negative lookup(5154)
5015 - don't create new directories with extent-mapping (5909, 5936)
5017 - fix lustre/lustrefs init scripts for SuSE (patch from Scali, 5702)
5018 - don't hold the pinger_sem in ptlrpc_pinger_sending_on_import
5019 - change obd_increase_kms to obd_adjust_kms (up or down) (5654)
5020 - lconf, lmc search both /usr/lib and /usr/lib64 for Python libs (5800)
5021 - support for RHEL4 kernel on i686 (5773)
5022 - provide error messages when incompatible logs are encountered (5898)
5024 2005-02-18 Cluster File Systems, Inc. <info@clusterfs.com>
5025 * version 1.4.0.10 (1.4.1 release candidate 1)
5027 - don't keep a lock reference when lock is not granted (4238)
5028 - unsafe list practices (rarely) led to infinite eviction loop (4908)
5029 - add per-fs limit of Lustre pages in page cache, avoid OOM (4699)
5030 - drop import inflight refcount on signal_completed_replay error (5255)
5031 - unlock page after async write error during send (3677)
5032 - handle missing objects in filter_preprw_read properly (5265)
5033 - no transno return for symlink open, don't save no-trasno open (3440)
5034 - don't try to complete elan receive that already failed (4012)
5035 - free RPC server reply state on error (5406)
5036 - clean up thread from ptlrpc_start_thread() on error (5160)
5037 - readahead could read extra page into cache that wasn't ejected (5388)
5038 - prevent races in class_attach/setup/cleanup/detach (5260)
5039 - don't dereference de->d_inode after l_dput of de (5458)
5040 - use "int" for stripe value returned from lock_to_stripe (5544)
5041 - mballoc allocation and error-checking fixes in 2.6 (5504)
5042 - block device patches to fix I/O request sizes in 2.6 (5482)
5043 - look up hostnames for IB nals (5602)
5044 - 2.6 changed lock ordering of 2 semaphores, caused deadlock (5654)
5045 - don't start multiple acceptors for the same port (5277)
5046 - fix incorrect LASSERT in mds_getattr_name (5635)
5047 - export a proc file for general "ping" checking (5628)
5048 - fix "lfs check" to not block when the MDS is down (5628)
5050 - service request history (4965)
5051 - put {ll,lov,osc}_async_page structs in a single slab (4699)
5052 - create an "evict_client" /proc entry on OSTs, like the MDS has
5053 - fix mount usage message, return errors per mount(8) (5168)
5054 - change grep [] to grep "[]" in tests so they work in more UMLs
5055 - fix ppc64/x86_64 spec to use %{_libdir} instead of /usr/lib (5389)
5056 - remove ancient LOV_MAGIC_V0 EA support (5047)
5057 - add "disk I/Os in flight" and "I/O req time" stats in obdfilter
5058 - align r/w RPCs to PTLRPC_MAX_BRW_SIZE boundary for performance (3451)
5059 - allow readahead allocations to fail when low on memory (5383)
5060 - mmap locking landed again, after considerable improvement (2828)
5061 - add get_hostaddr() to lustreDB.py for LDAP support (5459)
5063 2004-11-23 Cluster File Systems, Inc. <info@clusterfs.com>
5066 - send OST transaction number in read/write reply to free req (4966)
5067 - don't ASSERT in ptl_send_rpc() if we run out of memory (5119)
5068 - lock /proc/sys/portals/routes internal state, avoiding oops (4827)
5069 - the watchdog thread now runs as interruptible (5246)
5070 - flock/lockf fixes (but it's still disabled, pending 5135)
5071 - don't use EXT3 constants in llite code (5094)
5072 - memory shortage at startup could cause assertion (5176)
5074 - reorganization of lov code
5075 - single portals codebase
5077 - add extents/mballoc support (5025)
5078 - direct I/O reads in the obdfilter (4048)
5079 - kernel patches from LNXI for 2.6 (bluesmoke, perfctr, mtd, kexec)
5081 tbd Cluster File Systems, Inc. <info@clusterfs.com>
5084 - send OST transaction number in read/write reply to free req (4966)
5085 - don't ASSERT in ptl_send_rpc() if we run out of memory (5119)
5086 - lock /proc/sys/portals/routes internal state, avoiding oops (4827)
5087 - the watchdog thread now runs as interruptible (5246)
5088 - handle missing objects in filter_preprw_read properly (5265)
5089 - unsafe list practices (rarely) led to infinite eviction loop (4908)
5090 - drop import inflight refcount on signal_completed_replay error (5255)
5091 - unlock page after async write error during send (3677)
5092 - return original error code on reconstructed replies (3761)
5093 - no transno return for symlink open, don't save no-trasno open (3440)
5095 - add pid to ldlm debugging output (4922)
5096 - bump the watchdog timeouts -- we can't handle 30sec yet
5097 - extra debugging for orphan dentry/inode bug (5259)
5099 2004-11-16 Cluster File Systems, Inc. <info@clusterfs.com>
5102 - fix TCP_NODELAY bug, which caused extreme perf regression (5134)
5103 - allocate qswnal tx descriptors singly to avoid fragmentation (4504)
5104 - don't LBUG on obdo_alloc() failure, use OBD_SLAB_ALLOC() (4800)
5105 - fix NULL dereference in /proc/sys/portals/routes (4827)
5106 - allow failed mdc_close() operations to be interrupted (4561)
5107 - stop precreate on OST before MDS would time out on it (4778)
5108 - don't send partial-page writes before EOF from client (4410)
5109 - discard client grant for sub-page writes on large-page clients (4520)
5110 - don't free dentries not owned by NFS code, check generation (4806)
5111 - fix lsm leak if mds_create_objects() fails (4801)
5112 - limit debug_daemon file size, always print CERROR messages (4789)
5113 - use transno after validating reply (3892)
5114 - process timed out requests if import state changes (3754)
5115 - update mtime on OST during writes, return in glimpse (4829)
5116 - add mkfsoptions to LDAP (4679)
5117 - use ->max_readahead method instead of zapping global ra (5039)
5118 - don't interrupt __l_wait_event() during strace
5120 - add software watchdogs to catch hung threads quickly (4941)
5121 - make lustrefs init script start after nfs is mounted
5122 - fix CWARN/ERROR duplication (4930)
5123 - return async write errors to application if possible (2248)
5124 - add /proc/sys/portal/memused (bytes allocated by PORTALS_ALLOC)
5125 - print NAL number in %x format (4645)
5126 - update barely-supported suse-2.4.21-171 series (4842)
5127 - support for sles 9 %post scripts
5128 - support for building 2.6 kernel-source packages
5129 - support for sles km_* packages
5131 2004-10-07 Cluster File Systems, Inc. <info@clusterfs.com>
5134 - ignore -ENOENT errors in osc_destroy (3639)
5135 - notify osc create thread that OSC is being cleaned up (4600)
5136 - add nettype argument for llmount in #5d in conf-sanity.sh (3936)
5137 - reconstruct ost_handle() like mds_handle() (4657)
5138 - create a new thread to do import eviction to avoid deadlock (3969)
5139 - let lconf resolve symlinked-to devices (4629)
5140 - don't unlink "objects" from directory with default EA (4554)
5141 - hold socknal file ref over connect in case target is down (4394)
5142 - allow more than 32000 subdirectories in a single directory (3244)
5143 - fix blocks count for O_DIRECT writes (3751)
5144 - OST returns ENOSPC from object create when no space left (4539)
5145 - don't send truncate RPC if file size isn't changing (4410)
5146 - limit OSC precreate to 1/2 of value OST considers bogus (4778)
5147 - bind to privileged port in socknal and tcpnal (3689)
5149 - rate limit CERROR/CWARN console message to avoid overload (4519)
5150 - GETFILEINFO dir ioctl returns LOV EA + MDS stat in 1 call (3327)
5151 - basic mmap support (3918)
5152 - kernel patch series update from b1_4 (4711)
5154 2004-09-16 Cluster File Systems, Inc. <info@clusterfs.com>
5157 - avoid crash during MDS cleanup with OST shut down (2775)
5158 - fix loi_list_lock/oig_lock inversion on interrupted IO (4136)
5159 - don't use bad inodes on the MDS (3744)
5160 - dynamic object preallocation to improve recovery speed (4236)
5161 - don't hold spinlock over lock dumping or change debug flags (4401)
5162 - don't zero obd_dev when it is force cleaned (3651)
5163 - print grants to console if they go negative (4431)
5164 - "lctl deactivate" will stop automatic recovery attempts (3406)
5165 - look for existing locks in ldlm_handle_enqueue() (3764)
5166 - don't resolve lock handle twice in recovery avoiding race (4401)
5167 - revalidate should check working dir is a directory (4134)
5169 - don't always mark "slow" obdfilter messages as errors (4418)
5171 2004-08-24 Cluster File Systems, Inc. <info@clusterfs.com>
5174 - don't close LustreDB during write_conf until it is done (3860)
5175 - fix typo in lconf for_each_profile (3821)
5176 - allow dumping logs from multiple threads at one time (3820)
5177 - don't allow multiple threads in OSC recovery (3812)
5178 - fix debug_size parameters (3864)
5179 - fix mds_postrecov to initialize import for llog ctxt (3121)
5180 - replace config semaphore with spinlock (3306)
5181 - be sure to send a reply for a CANCEL rpc with bad export (3863)
5182 - don't allow enqueue to complete on a destroyed export (3822)
5183 - down write_lock before checking llog header bitmap (3825)
5184 - recover from lock replay timeout (3764)
5185 - up llog sem before sending rpc (3652)
5186 - reduce ns lock hold times when setting kms (3267)
5187 - change a dlm LBUG to LASSERTF, to maybe learn something (4228)
5188 - fix NULL deref and obd_dev leak on setup error (3312)
5189 - replace some LBUG about llog ops with error handling (3841)
5190 - don't match INVALID dentries from d_lookup and spin (3784)
5191 - hold dcache_lock while marking dentries INVALID and hashing (4255)
5192 - fix invalid assertion in ptlrpc_set_wait (3880)
5194 - add libwrap support for the TCP acceptor (3996)
5195 - add /proc/sys/portals/routes for non-root route listing (3994)
5196 - allow setting MDS UUID in .xml (2580)
5197 - print the stack of a process that LBUGs (4228)
5199 2004-07-14 Cluster File Systems, Inc. <info@clusterfs.com>
5202 - don't cleanup request in ll_file_open() on failed MDS open (3430)
5203 - make sure to unset replay flag from failed open requests (3440)
5204 - if default stripe count is 0, use OST count for inode size (3636)
5205 - update parent mtime/ctime on client for create/unlink (2611)
5206 - drop dentry ref in ext3_add_link from open_connect_dentry (3266)
5207 - free recovery state on server during a forced cleanup (3571)
5208 - unregister_reply for resent reqs (3063)
5209 - loop back devices mounting and status check on 2.6 (3563)
5210 - fix resource-creation race that can provoke i_size == 0 (3513)
5211 - don't try to use bad inodes returned from MDS/OST fs lookup (3688)
5212 - more debugging for page-accounting assertion (3746)
5213 - return -ENOENT instead of asserting if ost getattr+unlink race (3558)
5214 - avoid deadlock after precreation failure (3758)
5215 - fix race and lock order deadlock in orphan handling (3450, 3750)
5216 - add validity checks when grabbing inodes from l_ast_data (3599)
5218 - add /proc/.../recovery_status to obdfilter (3428)
5219 - lightweight CDEBUG infrastructure, debug daemon (3668)
5220 - change default OSC RPC parameters to be better on small clusters
5221 - turn off OST read cache for files smaller than 32MB
5222 - install man pages and include them in rpms (3100)
5223 - add new init script for (un)mounting lustre filesystems (2593)
5224 - run chkconfig in %post for init scripts (3701)
5225 - drop scimac NAL (unmaintained)
5227 2004-06-17 Cluster File Systems, Inc. <info@clusterfs.com>
5230 - clean kiobufs before and after use (3485)
5231 - strip trailing '/'s before comparing paths with /proc/mounts (3486)
5232 - remove assertions to work around "in-flight rpcs" recovery bug (3063)
5233 - change init script to fail more clearly if not run as root (1528)
5234 - allow clients to reconnect during replay (1742)
5235 - fix ns_lock/i_sem lock ordering deadlock for kms update (3477)
5236 - don't do DNS lookups on NIDs too small for IP addresses (3442)
5237 - re-awaken ptlrpcd if new requests arrive during check_set (3554)
5238 - fix cond_resched (3554)
5239 - only evict unfinished clients after recovery (3515)
5240 - allow bulk resend, prevent data loss (3570)
5241 - dynamic ptlrpc request buffer allocation (2102)
5242 - don't allow unlinking open directory if it isn't empty (2904)
5243 - set MDS/OST threads to umask 0 to not clobber client modes (3359)
5244 - remove extraneous obd dereference causing LASSERT failure (3334)
5245 - don't use get_cycles() when creating temp. files on the mds (3156)
5246 - hold i_sem when setting i_size in ll_extent_lock() (3564)
5247 - handle EEXIST for set-stripe, set proper directory name (3336)
5249 - servers can dump a log evicting a client - lustre.dump_on_timeout=1
5250 - fix ksocknal_fmb_callback() error messages (2918)
5252 2004-05-27 Cluster File Systems, Inc. <info@clusterfs.com>
5255 - don't copy lvb into (possibly NULL) reply on error (2983)
5256 - don't deref dentry after dput, don't free lvb on error (2922)
5257 - use the kms to determine writeback rpc length (2947)
5258 - increment oti_logcookies when osc is inactive (2948)
5259 - update client's i_blocks count via lvb messages (2543)
5260 - handle intent open/close of special files properly (1557)
5261 - mount MDS with errors=remount-ro, like obdfilter (2009)
5262 - initialize lock handle to avoid ASSERT on error cleanup (3057)
5263 - don't use cancelling-locks' kms values (2947)
5264 - use highest lock extent for kms, not last one (2925)
5265 - don't dereference ERR_PTR() dentry in error handling path (3107)
5266 - fix thread race in portals_debug_dumplog() (3122)
5267 - create lprocfs device entries at setup instead of at attach (1519)
5268 - common AST error handler, don't evict client on completion race (3145)
5269 - zero nameidata in detach_mnt in 2.6 (3118)
5270 - verify d_inode after revalidate_special is valid in 2.6 (3116)
5271 - use lustre_put_super() to handle zconf unmounts in 2.6 (3064)
5272 - initialize RPC timeout timer earlier for 2.6 (3219)
5273 - don't dereference NULL reply buffer if mdc_close was never sent (2410)
5274 - print nal/nid for unknown nid (3258)
5275 - additional checks for oscc recovery before doing precreate (3284)
5276 - fix ll_extent_lock() error return code for 64-bit systems (3043)
5277 - don't crash in mdc_close for bad permissions on open (3285)
5278 - zero i_rdev for non-device files (3147)
5279 - clear page->private before handing to FS, better assertion (3119)
5280 - tune the read pipeline (3236)
5281 - fix incorrect decref of invalidated dentry (2350)
5282 - provide read-ahead stats and refine rpc in flight stats (3328)
5283 - don't hold journal transaction open across create RPC (3313)
5284 - update atime on MDS at close time (3265)
5285 - close LDAP connection when recovering to avoid server load (3315)
5286 - update iopen-2.6 patch with fixes from 2399,2517,2904 (3301)
5287 - don't leak open file on MDS after open resend (3325)
5288 - serialize filter_precreate and filter_destroy_precreated (3329)
5289 - loop device shouldn't call sync_dev() for nul device (3092)
5290 - clear page cache after eviction (2766)
5291 - resynchronize MDS->OST in background (2824)
5292 - refuse to mount the same filesystem twice on same mountpoint (3394)
5293 - allow llmount to create routes for mounting behind routers (3320)
5294 - push lock cancellation to blocking thread for glimpse ASTs (3409)
5295 - don't call osc_set_data_with_check() for TEST_LOCK matches (3159)
5296 - fix rare problem with rename on htree directories (3417)
5298 - allow default OST striping configuration per directory (1414)
5299 - fix compilation for qswnal for 2.6 kernels (3125)
5300 - increase maximum number of MDS request buffers for large systems
5301 - change liblustreapi to be useful for external progs like lfsck (3098)
5302 - increase local configuration timeout for slow disks (3353)
5303 - allow configuring ldlm AST timeout - lustre.ldlm_timeout=<seconds>
5305 2004-03-22 Cluster File Systems, Inc. <info@clusterfs.com>
5308 - fixes for glimpse AST timeouts / incorrectly 0-sized files (2818)
5309 - don't overwrite extent policy data in reply if lock was blocked (2901)
5310 - drop filter export grants atomically with removal from device (2663)
5311 - del obd_self_export from work_list in class_disconnect_exports (2908)
5312 - don't LBUG if MDS recovery times out during orphan cleanup (2530)
5313 - swab reply message in mdc_close, other PPC fixes (2464)
5314 - fix destroying of named logs (2325)
5315 - overwrite old logs when running lconf --write_conf (2264)
5316 - bump LLOG_CHUNKSIZE to 8k to allow for larger clusters (2306)
5317 - fix race in target_handle_connect (2898)
5318 - mds_reint_create() should take same inode create lock (2926)
5319 - correct journal credits calculated for CANCEL_UNLINK_LOG (2931)
5320 - don't close files for self_export to avoid uninitialized obd (2936)
5321 - allow MDS with the same name as client node (2939)
5322 - hold dentry reference for closed log files for unlink (2325)
5323 - reserve space for all logs during transactions (2059)
5324 - don't evict page beyond end of stripe extent (2925)
5325 - don't oops on a deleted current working directory (2399)
5326 - handle hard links to targets without a parent properly (2517)
5327 - don't dereference NULL lock when racing during eviction (2867)
5328 - don't grow lock extents when lots of conflicting locks (2919)
5330 2004-03-04 Cluster File Systems, Inc. <info@clusterfs.com>
5333 - account for cache space usage on clients to avoid data loss (974)
5334 - lfsck support in lustre kernel code (2349)
5335 - reduce journal credits needed for BRW writes (2370)
5336 - orphan handling to avoid losing space on client/server crashes
5337 - ptlrpcd can be blocked, stopping ALL progress (2477)
5338 - use lock value blocks to assist in proper KMS, faster stat (1021)
5339 - takes i_sem instead of DLM locks internally on obdfilter (2720)
5340 - recovery for initial connections (2355)
5341 - fixes for mds_cleanup_orphans (1934)
5342 - abort_recovery crashes MDS in b_eq (mds_unlink_orphan) (2584)
5343 - block all file creations until orphan recovery completes (1901)
5344 - client remove rq_connection from request struct (2423)
5345 - conf-sanity test_5, proper cleanup in umount log not availale (2640)
5346 - recovery timer race (2670)
5347 - mdc_close recovey bug (2532)
5348 - ptlrpc cleanup bug (2710)
5349 - mds timeout on local locks (2588)
5350 - namespace lock held during RPCs (2431)
5351 - handle interrupted sync write properly (2503)
5352 - don't try to handle a message that hasn't been replied to (2699)
5353 - client assert failure during cleanup after abort recovery (2701)
5354 - leak mdc device after failed mount (2712)
5355 - ptlrpc_check_set allows timedout requests to complete (2714)
5356 - wait for inflight reqs when ptlrpcd finishes (2710)
5357 - make sure unregistered services are removed from the srv_list
5358 - reset bulk XID's when resending them (caught by 1138 test)
5359 - unregister_bulk after timeout
5360 - fix lconf error (2694)
5361 - handle write after unfinished setstripe, stripe-only getstripe (2388)
5362 - readahead locks pages, leaves pending causing memory pressure (2673)
5363 - increase OST request buffers to 4096 on large machines (2729)
5364 - fix up permission of existing directories in simple_mkdir (2661)
5365 - init deleted item, add assertions ptlrpc_abort_inflight() (2725)
5366 - don't assign transno to errored transactions (2742)
5367 - don't delete objects on OST if given a bogus objid from MDS (2751)
5368 - handle large client PAGE_SIZE readdir on small PAGE_SIZE MDS (2777)
5369 - if rq_no_resend, then timeout request after recovery (2432)
5370 - fix MDS llog_logid record size, 64-bit array alignment (2733)
5371 - don't call usermode_helper from ptlrpcd, DEFAULT upcall (2773)
5372 - put magic in mount.lustre data, check for bad/NULL mount data (2529)
5373 - MDS recovery shouldn't delete objects that it has given out (2730)
5374 - if enqueue arrives after completion, don't clobber LVB (2819)
5375 - don't unlock pages twice when trigger_group_io returns error (2814)
5376 - don't deref NULL rq_repmsg if ldlm_handle_enqueue failed (2822)
5377 - don't write pages to disk if there was an error (1450)
5378 - don't ping imports that have recovery disabled (2676)
5379 - take buffered bytes into account when balancing socknal conn (2817)
5380 - hold a DLM lock over readdir always, use truncate_inode_pages (2706)
5381 - reconnect unlink llog connection after MDS reconnects to OST (2816)
5382 - remove little-endian swabbing of llog records (1987)
5383 - set/limit i_blksize to LL_MAX_BLKSIZE on client (2884)
5384 - retry reposting request buffers if they fail (1191)
5385 - grow extent at grant time to avoid granting a revoked lock (2809)
5386 - lock revoke doesn't evict page if covered by a second lock (2765)
5387 - disable VM readahead to avoid reading outside lock extents (2805)
5389 - return LL_SUPER_MAGIC from statfs for the filesystem type (1972)
5390 - updated kernel patches for hp-2.4.20 kernel (2681)
5392 2004-02-07 Cluster File Systems, Inc. <info@clusterfs.com>
5395 - fix truncated write corruption (2366)
5396 - fix for failed assertion in iopen_connect_dentry (1792,2517)
5398 - don't flag the ptlrpcd thread with PF_MEMALLOC (2636)
5399 - ensure len(uuid) < 37 in lmc (1171)
5400 - fix ia64 OOPS in llog_test (2255)
5401 - zero end of page at obdfilter for partial page writes (2648)
5402 - don't leave stale dentries around after renames (bug 2428)
5403 - fix timeouts when evicting a client with a single lock held (2642)
5404 - set deadline for the initial HELLO message to drain (2634)
5405 - print out dotted-quad IP addresses in the socknal (2302)
5407 - additional debugging for MDS client eviction problem (2443)
5408 - fix mkfsoptions support for osts (2603, 2604)
5410 2004-01-27 Cluster File Systems, Inc. <info@clusterfs.com>
5413 - add series for the vanilla 2.6.0 kernel
5414 - add series for the vanilla 2.4.24 kernel
5415 - add series for a cray x86/64 UL kernel drop
5416 - fix xattr patches for the vanilla 2.4.19 series
5418 - generate true UUIDs in lmc (1171)
5419 - have portals stack dumping break in UML (2466)
5420 - avoid bad dchild deref; avoid inum lock w/o creation (2362)
5421 - allocate with _NOFS in ldlm to avoid deadlock (1933)
5422 - wake callback waiting threads on client eviction (2460)
5423 - Add --ptldebug and --subsystem to lmc (1719)
5424 - update assertion to allow safe interrupt allocation
5425 - set rq_no_resend for cancel requests (2432)
5426 - recalculate ptlrpcd timeout after resend (2494)
5427 - call vfs_rmdir when removing pending directories (2368)
5428 - fix renaming a file to itself (2429)
5429 - lmc creates a default one-stripe lov (2454)
5430 - expand procfs space to handle large clusters (2326)
5431 - increase UML stack to avoid overflow
5432 - update lconf's list of debug and subsystem masks
5433 - fix lfs find --obd (2510)
5434 - /proc tunable for disabling filter read caching (2591)
5435 - stop rpm packages from altering slapd.conf (2301)
5436 - disable nagle in the socknal under 0conf (2578)
5437 - choose mds inode size based on stripe count (2572)
5438 - fix kernel-source rpm problems (2516)
5440 - add --disable-doc to avoid pdf generation (2421)
5441 - update documentation, tests, type-os, comments
5442 - avoid format warnings on ia64
5443 - remove the TOE NAL
5444 - tiny code cleanups by removing unused fields
5446 2004-01-07 Cluster File Systems, Inc. <info@clusterfs.com>
5449 - fix obvious semaphore misuse in as-yet-unused setattr path (2348)
5450 - remove the most blatant lies from BUILDING file (2371)
5451 - change default debug level to reasonable production setting
5452 - reduce client side cache size to reduce cache flush time
5453 - reduce max RPCs in flight to avoid unnecessary file fragmentation
5454 - make TCP zerocopy and pinger support enabled by default (2476)
5455 - sync writes completed after process exits caused crashes (2319)
5456 - maintain correct mount count on the MDS (2356)
5457 - backout 1557, because 2316 wasn't really fixed
5458 - better file I/O statistics gathering in /proc
5459 - don't take unnecessary, deadlock-inducing bug in readpage (2383)
5460 - another kernel patch to fix zero-copy TCP function export
5461 - don't take duplicate lock when processing re-sent getattr (2420)
5462 - lctl uses obd_self_export instead of creating new conn (2353)
5463 - MDS/OST recovery case which requires object creation asserted (2425)
5464 - move lfs from /usr/sbin to /usr/bin in packages
5465 - fix race between mds_client_add and mds_client_free (2417)
5466 - use kmalloc instead of slabs in portals (2430)
5467 - don't create duplicate records when a failover MDS is present (2442)
5468 - remove unnecessary mount age check (2332)
5469 - don't remove directory inodes from locks prematurely (2451)
5470 - don't break if MDS service name is the same as hostname (2103)
5471 - fix races in client write RPC generation when cache full (2482)
5473 2003-12-13 Cluster File Systems, Inc. <info@clusterfs.com>
5476 - remove now-unused request->rq_obd (278)
5477 - if an allocation fails, print out how much memory we've used (1933)
5478 - use PORTAL_SLAB_ALLOC for structures, to get GFP_MEMALLOC (1933)
5479 - add the "configurable stack size" patch to most series files (1256)
5480 - ability to write large log records, for 100+ OST configs (2306)
5481 - fix NULL deref when filter_prep fails (2314)
5482 - fix operator precedence error in filter_sync
5483 - dynamic allocation of socknal TX descriptors (2315)
5484 - fix a missed case in the GFP_MEMALLOC patch, can cause deadlock (2310)
5485 - fix gcc 2.96 compilation problem in xattr kernel patch (2294)
5486 - ensure that CWARN messages in Portals always get to the syslog
5487 - __init/__exit are not for prototype decls (ldlm_init/exit)
5488 - x86-64 compile warning fixes
5489 - fix gateway LMC keyword conflict (2318)
5490 - fix MDS lock inversions in getattr/reint paths (1844)
5491 - fix a rare lock re-ordering bug, which caused deadlock (2322)
5492 - fix i_sem/journal inversion in fsfilt_ext3_write_record (2306)
5493 - DLM race condition prevented some lock evictions (2328)
5494 - ENOMEM detection and retry on socknal sends (2230)
5495 - use GFP_NOFS throughout Lustre, to combat ENOMEM (2230)
5496 - move osc_rpcd into ptlrpc, for use in MDC and others (2329)
5497 - protect MDS inode fsdata with stronger locking; fixes assertion (2313)
5498 - better error messages when a client is rejected during recovery (1505)
5499 - avoid cancelling locks which were never granted, after failure (2330)
5500 - fix i_sem/journal inversion in mds_client_add (2333)
5501 - fix truncate/getattr lock cycle deadlock (2334)
5502 - use rpcd to send close; allows resend after timeout, avoid leak (1897)
5503 - fix two rare exit paths which could leak an l_lock() ref (2321)
5504 - fencepost error in MDS/OST orphan recovery (2226)
5505 - make log record alignment 8 bytes (1988)
5506 - lstripe now fails when requested offset > ost_count (2237)
5507 - ensure that all kernel series have a complete list.h (1607)
5508 - fix crashes in special-file operations (2316)
5509 - lctl create/brw OID mismatch, caused by obsolete filter loop (2339)
5511 - allow configurable automake binary, for testing new versions
5512 - small update to the lfs documentation
5514 2003-12-03 Cluster File Systems, Inc. <info@clusterfs.com>
5516 * fix negative export reference count in fsfilt_sync (2312)
5518 2003-12-01 Cluster File Systems, Inc. <info@clusterfs.com>
5519 * release candidate 0.9.1
5521 - orphans are moved into the PENDING directory for possible recovery
5522 - replayed opens now open by fid for orphan/rename safety (1042)
5523 - last close of an orphan inode generates a transno (683)
5524 - chdir() and mount() now pin the directory entry (1020)
5525 - avoid CERROR in normal ll_setattr_raw() error case (1500)
5526 - discard very old requests without processing them (1502)
5527 - remove some common, well-understood CERRORs (1505)
5528 - require O_DIRECT I/O to be page-sized to workaround IA64 crash (1609)
5529 - clear "grant" flags in OST replies until OST grant code lands (1644)
5530 - fix read performance by not clobbering i_blksize on client (1598)
5531 - fix __ldlm_handle2lock oops by not dereferencing lock after PUT (1625)
5532 - make LRU size a /proc tunable, clears locks when reduced (707)
5533 - fix some lprocfs rot that prevented ptlbd from loading (1732)
5534 - server locks take references on exports now (1558)
5535 - build fixes for 2.4.20-rh trees (1663)
5536 - return an error from lov_create if all OSCs are inactive (1751)
5537 - fix import levels when a reconnect happens without a timeout (1597)
5538 - exit early from mds_open if we get a lookup error (1749)
5539 - partial page read at EOF wouldn't wait for disk before sending (1642)
5540 - avoid NULL deref in obdfilter when reading page past EOF (1592)
5541 - avoid LASSERT in ll_intent_lock if server failed very early (1090)
5542 - fix LBUG in ll_it_open_error with rc = -2 (1861)
5543 - write/truncate lock inversion (1639)
5544 - Don't auto-load obdclass, portals modules during cleanup (1495)
5545 - fix timestamps from jumping to "now" (1763)
5546 - extra journal assertions (1648)
5547 - add an extra multiunlink test (1771)
5548 - fix read_record/write_record API (1776)
5549 - fix leak of offset_extent, possible incorrect i_size later (1772)
5550 - fix lasserts in mis-matched transnos during open-unlink testing (1541)
5551 - Debugging for the kqswnal_get_idle_tx problems (1820)
5552 - Allow recovery to be attempted multiple times (1536)
5553 - Write out MDS last_rcvd file after it is first created (1600)
5554 - Fix tx_descriptor leak in failed transmit situations (1827)
5555 - ext3 journaling fixes for assertion failure after IO error (1871)
5556 - class_export_put() on freed export after completion AST error (1896)
5557 - Fix revalidate looping in VFS (1322)
5558 - Don't access a freed export during MDS_REINT timeout (1521)
5559 - Add open-unlink recovery support on the MDS (1673,1764)
5560 - Return an error if no MDS data was read from last_rcvd (1946)
5561 - Fix for lookup "." or ".." crash on error (1932,1931,1935)
5562 - Don't setup a disk device that doesn't match exported UUID (317)
5563 - Reduce bulk RPC timeout to avoid cascading client/OST failures (1845)
5564 - avoid committing NULL handle in force close
5565 - local.sh is now a one-stripe LOV configuration
5566 - POSIX utime.4 -EPERM on FIFO not owned by user (56)
5567 - fix ext3 htree duplicate directory entry corruption (1516)
5568 - POSIX creat.13, fstat.1, open.18, stat.3 new file atime/mtime (2020)
5569 - update to new LOV EA format (2097)
5570 - interoperability for different PAGE_SIZE/wordsize (686,1821,1343,2042)
5572 2003-06-15 Phil Schwan <phil@clusterfs.com>
5575 - imports and exports cleanup too early, need refcounts (349, 879, 1045)
5576 - per-import/export recovery handling (958, 931, 959)
5577 - multiple last-rcvd slots, for serving multiple FSes (949)
5578 - connections are again shared between multiple imp/exports (963, 964)
5579 - "umount -f" would hang if any requests needed to be sent (393, 978)
5580 - avoid pinning large req buffer by copying for queued messages (989)
5581 - add "uuid" to "lctl device" command to help upcalls (991)
5582 - "open" RPCs with transnos would confuse recovery counters (1037)
5583 - do proper endian conversion of all wire messages (288, 340, 891)
5584 - remove OST bulk get LBUGs, fix ost_brw_write cleanup (1126)
5585 - call waiting locks callback from LDLM recovery thread (1127, 1151)
5586 - fix ptlrpc_connection leak in target_handle_connect (1174)
5587 - fix import refcounting bug in OST and MDS cleanup (1134)
5588 - if an invalid-at-open-time OSC returned before close(), LBUG (1150)
5589 - fix very unlikely obd_types race condition (501)
5590 - remove osc_open hack for echo_client (1187)
5591 - we leaked exports/dlmimps for forcibly disconnected clients (1143)
5592 - a failure in read_inode2 leads to deadlock (1139)
5593 - cancel ack-locks as soon as transaction is committed (1072)
5594 - fix major leaks and crashes in the bulk I/O path (937, 1057)
5595 - make sure to commitrw after any preprw to avoid deadlock (1162)
5596 - failing to execute a file in a lustre FS would lock inode (1203)
5597 - small DEBUG_REQ fix to avoid dereferencing a NULL (1227)
5598 - don't ASSERT while cleaning up an incompletely-setup obd (1248)
5599 - obd_uuid2tgt would walk off the end of the list (1255)
5600 - on IA64 the osc would give portals incorrect bulk size (1258)
5601 - fix debug daemon ioctl interface; allows daemon on ia64 (1274)
5602 - fix lock inversion caused by new llite matching code (1282)
5603 - limit the number of dirty pages on a client to 10MB (1286)
5604 - timed out locks were not being corrected cancelled (1289)
5605 - fix O_DIRECT above 4GB on IA-32 (1292)
5606 * major user-visible changes
5607 - fail out/fail over policy now controlled by the upcall (993)
5609 - add OBD_PING to check server availability and failure (954)
5610 - lustre messages are now sent in sending host order (288, 340, 891)
5611 - add eadatalen to MDS getattr reply (340)
5612 - OST read replies may contain second buffer, with per-page status (593)
5614 2003-03-11 Phil Schwan <phil@clusterfs.com>
5617 - LDLM_DEBUG macro fix, for gcc 3.2 (850)
5618 - failed open()s could cause deadlock; fixed (867, 869)
5619 - stop cancelling OST locks when files are closed (481)
5620 - overlapping XID spaces caused network corruption (851, 853)
5621 - fix unsafe fsfilt counter arithmetic; change to atomic_t
5622 - setattr_raw added, to do single-RPC, server-side setattrs
5623 - lmc/lconf syntax change for OST UUIDs
5624 - fix crashy race condition between ptlrpc_free_req and osc_close
5625 - don't use request in mdc_enqueue if we hit a timeout (889)
5626 - don't set the inode i_size for regular files from the MDS (896)
5627 - handle out of order completion AST (842)
5628 - don't LBUG if a lock request times out after receiving AST (913)
5629 - avoid d_rehash race in ll_find_alias by rehashing inside dcache_lock
5630 - if a bad lock AST arrives, send an error instead of dropping entirely
5631 - return 0 from revalidate2 if ll_intent_lock returns -EINTR (912)
5632 - fix leak in bulk IO when only partially completed (899, 900, 926)
5633 - fix O_DIRECT for ia64 (55)
5634 - (almost) eliminate Lustre-kernel-thread effects on load average (722)
5635 - C-z after timeout could hang a process forever; fixed (977)
5637 - client-side I/O cache (678, 924, 929, 941, 970)
5639 - READPAGE and SETATTRs which don't take server-side locks get
5642 2003-02-11 Phil Schwan <phil@clusterfs.com>
5645 - Fix ldlm_lock_match on the MDS to avoid matching remote locks (592)
5646 - Fix fsfilt_extN_readpage() to read a full page of directory
5647 entries, or fake the remainder if PAGE_SIZE != blocksize (500)
5648 - Avoid extra mdc_getattr() in ll_intent_lock when possible (534, 604)
5649 - Fix imbalanced LOV object allocation and out-of-bound access (469)
5650 - Most intent operations were removed, in favour of a new RPC mode
5651 that does a single RPC to the server and bypasses most of the VFS
5652 - All LDLM resource ID arrays were removed in favour of ldlm_res_id
5653 - Aggressively cancel local locks on DLM servers
5654 - mds_reint_unlink sends EA to the client if it's the last nlink.
5655 client uses that EA to unlink OST objects.
5656 - mds_reint_{rename,unlink,link} were rewritten to take ordered locks
5657 - recursive symlinks were fixed (439)
5658 - fixed NULL deref in DEBUG_REQ
5659 - filter_update_lastobjid no longer calls sync, which annoyed extN
5660 - fixed multi-client small-writes to a single file problem (445)
5661 - fixed mtime updates during file writes (607)
5662 - fixed vector writes on obdfilter causing problems when ENOSPC (670)
5663 - fixed bug in obd_brw_read/write() (under guise of testing 367)
5664 - fixed Linux OST size reporting problem (444, 656)
5665 - OST now updates object mtime with writes or setattr (607, 619)
5666 - client verifies file size before zeroing page past EOF (445)
5667 - OST now writes last allocated objid to disk with allocation (108)
5668 - LOV on echo now works (409)
5670 - mds_reint_unlink sends a new buffer, with the EA included. this
5671 buffer is only valid if body->valid & OBD_MD_FLEASIZE, which is only
5672 set if a regular file was being unlinked, and it was the last link
5673 - use PtlGet from the target for bulk writes (315)
5674 - OST now updates object mtime with writes or setattr (607, 619)
5675 - LDLM now has a grant-time callback to revalidate locked items, if
5677 - Many MDS operations were reorganized to combat race conditions
5679 - Merge b_intel branch (updated lprocfs code) - now at /proc/fs/lustre
5680 - configure check to avoid gcc version 2.96 20000731-2.96-98) (606)
5682 2003-01-06 Andreas Dilger <adilger@clusterfs.com>
5685 - Fully reactivate OST imports after reconnection (512, others)
5686 - Make sure client sees our -ENOTCONN from mds_handle (513 - partial)
5687 - More graceful error handling for truncating on dead OST (515)
5688 - Don't error out unless we're actually accessing dead stripes (474)
5689 - Fix garbage sizes when stripes are missing (410)
5690 - LRU counters were broken, causing constant lock purge (433, 432)
5691 - garbage on read from stripes with failed OSTs (441)
5692 - mark OSCs as active before reconnecting during recovery (438)
5693 - lov_enqueue and lov_cancel need to handle inactive OSTs (403)
5694 - lfind did not preserve OST order in output (443)
5695 - symlinks cause hung clients, incorrect data (439)
5696 - stop dereferencing request after dropping refcount (457)
5697 - don't LASSERT(spin_is_locked) on non-SMP (455)
5698 - fixes for many rename() bugs
5699 - fstat didn't correctly synchronize attributes (399)
5700 - server must handle lock cancellation during blocking AST prep (487)
5701 - bulk descriptors were free()d too soon (511)
5702 - fix paths in lconf, which would load incorrect modules (451, 507)
5703 - fix confusing lconf 'host not found' error message (386)
5704 - fix lock order deadlock on OST (O/R i_sem before journal ops, 478)
5705 - fix race condition in mdc_blocking_ast() for inode access (526)
5706 - fix lov_unpackmd() unpacking wrong number of stripes (537)
5707 - fix lov_set_osc_active() marking wrong OSC inactive (440)
5708 - fix bad lstripe lov_unpackmd() assertion (fix layering too) (527)
5709 - fix multiple writes of stripe MD to MDS (358, maybe 519)
5710 - fix lstripe in several ways (kernel side) (527)
5711 - fix request leak in ldlm_cli_enqueue (262)
5712 - incorrect OSC was marked inactive after OST failure
5713 - call mds_fs_cleanup before unmounting filesystem (524)
5714 - fix races between taking ns_lock and ldlm_lock_change_resource
5715 - fix races updating LOV export open file list
5716 - fix lov_enqueue error path, avoid decref-ing bad lock handle
5717 - fix recovery NULL deref in ldlm_cli_cancel_unused
5718 - fix some DLM races by using new hash table for lock handles (419)
5719 - permit the client to specify desired inodes, at replay
5720 - duplicate requests when we queue them for replay reintegration
5721 - fix last_rcvd offset calculation
5722 - sync after each recovered transaction, so we always make progress
5723 - never, not always, ERESTART requests without transnos
5724 - store the lov_desc in the MDS, so we don't depend on getlovinfo to
5726 - skip replay if the MDS says that the client is already connected
5727 - don't check for a recovery-enabled export to match lctl's UUID
5728 - don't INC_USE_COUNT for phantom exports
5729 - don't crash when cleaning up phantom exports (567)
5730 - don't double-finish or set replay data for errored mdc_open requests
5731 - abort requests when they time out, so we don't get old replies
5732 - send/receive replies for AST messages again
5733 - if the client says that it doesn't have the lock, cancel it on the
5735 - if we timeout during I/O, don't try to cancel an in-use lock; instead
5736 mark it as destroyed, it will all work out when decref is called
5737 - fix module use counts (22, 581)
5739 - ASTs now expect a reply (server cancels lock on error reply)
5741 2002-12-02 Andreas Dilger <adilger@clusterfs.com>
5744 - fix many simultaneous client startup (392)
5745 - fix dentry->d_it clobbering
5746 - credentials weren't being shipped for readdir/getattr operations
5747 - remove invalid assertions triggered during some concurrent MD
5749 - proper Lustre versions added (336, 389)
5750 - fix memory leak for create error case (398)
5751 - fix LOV locking bug that would get cli/srv out of sync
5752 - fix echo client over LOV (409)
5753 - fix dbench 2, extN refcount problem (170, 258, 356, 418)
5754 - fix double-O_EXCL intent crash (424)
5755 - avoid sending multiple lock CANCELs (352)
5757 - MDS can do multi-client recovery (modulo bugs in new code)
5759 - many updates, edits, cleanups
5761 2002-11-18 Phil Schwan <phil@clusterfs.com>
5764 - fix null d_it dereference (346)
5765 - fix full OST/dbench hang (333)
5766 - fix permission problem with file removal (286)
5767 - fix removal of OSCs from LOV when they fail
5768 - fix NULL deref during bulk timeout (214)
5769 - fix problems related to multiple filesystems on one MDS (241)
5770 - fixed serious subtle metadata locking bugs
5771 - free locks on clients when inodes are removed due to memory
5773 - fix inode pointer in lock data (285)
5774 - partial support for multiple MDS on a single host (241)
5775 - data locks weren't cancelled at clear_inode time (290, 311)
5776 - intent locks could lead to unbounded lock growth (205)
5777 - added a maximum lock count, an LRU list, and a flusher
5778 - fix multiple rename (365)
5779 - properly abstracted the echo client
5780 - OSC locked 1 byte too many; fixed
5781 - rewrote brw callback code:
5782 - fixed recovery bugs related to LOVs (306)
5783 - fixed too-many-pages-in-one-write crash (191)
5784 - fixed (again) crash in sync_io_timeout (214)
5785 - probably fixed callback-related race (385)
5787 - Add capability to MDS protocol
5788 - LDLM cancellations and callbacks on different portals
5790 2002-10-28 Andreas Dilger <adilger@clusterfs.com>
5793 - limit client IOV size to PTL_MD_MAX_IOV (611336, 191)
5794 - defer open object destruction to close time (601981, 138)
5795 - open/close OST file handle in obdo (OBD_MD_FLHANDLE) (601981, 138)
5796 - move LDLM_ENQUEUE/CONVERT back to MDS portal (625069)
5797 - abstract ll_lookup2, fix ll_revalidate2 to use abstraction (256)
5798 - don't call obd_setattr in ll_file_release for destroyed objects
5799 * protocol change to lustre_msg: move |version| and add |flags|
5800 * protocol change to osc_punch: "start" in "o_size", "end" in "o_blocks"
5801 * lock replay: for LDLM_FL_REPLAY trust client to do right thing
5802 * added replay of create, unlink, link and rename operations during
5803 MDS failover; recovery should be much more robust now
5804 * remove failed OSCs from LOVs (only lov_create uses this so far)
5805 * the lustre-HOWTO was brought (more) up to date (582544)
5807 2002-10-23 Phil Schwan <phil@clusterfs.com>
5810 - in-use dentries weren't being reused properly (617851)
5811 - prevent multiple LDLM setup (599178)
5812 - fix LOV size calculations for truncate (617853)
5813 - fix client handling of MDS intent errors (POSIX)
5814 - fix permission bug in lovstripe.c test (624321)
5815 - fix MDS thread deadlock - move LDLM handler to DLM portal (625069)
5816 - truncate past end of file could corrupt data
5817 - proper cleanup after timeouts, crashes, etc (592524, 550815)
5818 - a race in recovery could return ETIMEDOUT to apps (623947)
5819 - building outside the source directory was fixed
5820 * the lustre-HOWTO was brought (more) up to date (582544)
5821 * major progress was made on recovery functionality
5823 2002-10-10 Phil Schwan <phil@clusterfs.com>
5826 - recovery deadlock fix
5827 - rm -rf causes LBUG fix (617817)
5828 - file open by multiple tasks fix (618962)
5829 - directory permissions bugs (602707 and 620007)
5830 - journal_stop fixed with locking (611313)
5831 - O_APPEND failures resolved (618273, perhaps 614459)
5832 - lconf PATH fix (619770)
5833 - IA64 build fix (621450)
5834 - RPC buffer sizes scale with amount of memory
5836 2002-10-01 Phil Schwan <phil@clusterfs.com>
5839 - locks would be cancelled without throwing away data pages,
5840 resulting in inconsistent data (605627)
5841 - inode attributes were not always being refreshed (605627, 612449)
5842 - lconf now continues to cleanup after lctl reports an error
5843 - MDS now enforces user permissions (602707)
5844 - lprocfs cleanup fixed, but not yet enabled (614157)
5845 - fixed infinite server hang, should a client not respond to an AST
5846 - avoid going into recovery if user calls readlink() with a buffer
5847 that's too small (613941)
5848 - AST RPCs no longer require replies (614867) -- this may be changed
5849 - don't crash server if client sends an IOV that's too big (611336)
5850 - fixed lock conversion deadlock (611892)
5851 - fixed the following of symlinks (614622)
5852 * recovery: the server can remove locks from a client that dies, other
5853 clients can make progress
5854 * more extN patch fixes
5855 * compile-time configurable ptlrpc buffer allocations
5857 - collaborative read cache document
5858 - Lustre Lite Performance CDR document-in-progress
5860 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
5863 - fix typo in patch-2.4.18
5865 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
5868 - clear ptlrpc request each time in handle_incoming_request()
5869 - unlink of files now destroys the object on the OST
5871 2002-09-19 Peter Braam <braam@clusterfs.com>
5873 * add hard link support
5874 * change obdfile creation method
5875 * kernel patch changed
5877 2002-09-19 Peter Braam <braam@clusterfs.com>
5880 - stack overflow bug in extN fixed
5882 2002-09-18 Andreas Dilger <adilger@clusterfs.com>
5884 * documentation updates
5885 - add man pages for config tools
5886 - update tests/README to describe testing with new config tools
5887 - finish metadata API descriptions
5888 * bug fixes and cleanups
5889 - statfs workaround for 16TB limit
5890 - LOV stripe allocation improved, can stripe on subset of OSTs
5891 - LOV file size/IO offset was wrong for files > 4GB in size
5892 - object EA data was being dropped, caused files to be unreadable
5893 - memory overflow with non-LOV OST caused memory corruption
5894 - fixed regression tests to work with new config tools, obdfilter
5895 - fixed bug when directory size became larger than 1 block
5896 - fixed bug (for single client case) when PWD was deleted
5897 - invalidate local directory pages when doing intent-based ops
5898 - avoid LDLM oops when lock callback contained bad data
5900 2002-09-09 Andreas Dilger <adilger@clusterfs.com>
5902 * documentation updates
5903 * bug fixes and cleanups
5904 - configuration tools
5907 - 64-bit compile warnings
5908 - 64-bit internal statfs data
5910 * test_brw on persistent OST devices
5912 * lprocfs (disabled)
5914 2002-09-04 Andreas Dilger <adilger@clusterfs.com>
5916 * documentation updates
5917 * bug fixes and cleanups
5918 * configuration tools
5920 2002-08-30 Peter J. Braam <braam@clusterfs.com>
5923 * many small fixes to 0_5_4
5924 * io/network handling
5925 * thinkos in MDS operations
5927 2002-08-24 Peter J. Braam <braam@clusterfs.com>
5930 * crucial basic fixes to 0.5.3
5931 * IOR, Iozone work over Elan
5934 2002-08-07 Phil Schwan <phil@clusterfs.com>
5935 * version 0_5_3, our first alpha
5936 * we use the new Portals iovs
5937 * documentation updates
5938 * bug fixes and cleanups
5939 * small changes in the DLM wire protocol
5941 2002-07-25 Peter J. Braam <braam@clusterfs.com>
5942 * version 0_5_1 with some initial stability,
5943 * locking on MD and file I/O.
5944 * documentation updates
5945 * several bug fixes since 0.5.0
5946 * small changes in wire protocol
5948 2002-07-18 Phil Schwan <phil@clusterfs.com>
5950 * delivered as Lustre Light Alpha
5951 * fixed a crash after handling invalid MDS requests
5952 * fixed directory pages for architectures with non-4k pages sizes
5954 2002-07-11 Andreas Dilger <adilger@clusterfs.com>
5955 * release version v0_4_4
5956 * Moves TCP acceptor to be on port 2432 (unused Coda port) instead
5958 * Fixes a number of interruption problems with OST operations.
5959 * Update documentation for portals header changes
5960 * Move all wire protocol structs/defines to lustre_idl.h
5961 * Fixes symlink length bug.
5962 * Add tcpdump to repository.
5964 2002-07-05 Andreas Dilger <adilger@clusterfs.com>
5965 * release version v0_4_3
5966 * Fixes statfs for inodes on extN.
5967 * Fixes bug in runtests which would delete /etc/hosts.
5968 * Use 64-bit object IDs wherever possible (not into VFS though)
5969 Remove ost_get_info, which is unused by lustre, and out of date.
5971 2002-07-03 Peter Braam <braam@clusterfs.com>
5972 * release version v0_4_2 Fixes a lookup error (type not passed)
5973 * move forward to head of Portals
5974 * move forward to latest Lustre kernel
5976 2002-06-25 Peter Braam <braam@clusterfs.com>
5977 * release version v0_4_1. Hopefully stable on single node use.
git://git.whamcloud.com / fs / lustre-release.git / blob