1 tbd Sun Microsystems, Inc.
4 2.6.16.60-0.37 (SLES 10),
5 2.6.18-128.1.6.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: 1.41.5.sun2
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.
16 Severity : enhancement
18 Description: Update kernel to RHEL5.3 2.6.18-128.1.6.el5.
20 Severity : enhancement
22 Description: Update kernel to SLES10 SP2 2.6.16.60-0.37.
26 Description: resolve race between obd_disconnect and class_disconnect_exports
27 Details : if obd_disconnect will be called to already disconnected export he
28 forget release one reference and osc module can't unloaded.
30 Severity : enhancement
32 Description: Allow tuning service thread via /proc
33 Details : For each service a new
34 /proc/fs/lustre/{service}/*/thread_{min,max,started} entry is
35 created that can be used to set min/max thread counts, and get the
36 current number of running threads.
40 Descriptoin: don't return error if have particaly created objects for file.
41 Details : lov_update_create_set uses set->set_success as index for created objects,
42 so if some requests will be failed, they will have hole at end of
43 array and we can use qos_shrink_lsm for allocate correct lsm.
45 Severity : enhancement
47 Description: Update OFED support to 1.4.1
51 Description: Reduce small size read RPC
52 Details : Set read-ahead limite for every file and only do read-ahead when
53 available read-ahead pages are bigger than 1M to avoid small size
56 Severity : enhancement
58 Description: move AT tunable parameters for more consistent usage
59 Details : add AT tunables under /proc/sys/lustre, add to conf_param parsing
61 Severity : enhancement
63 Description: add lazystatfs mount option to allow statfs(2) to skip down OSTs
64 Details : allow skip disconnected ost for send statfs request and hide error
70 Description: fix racy locking of mballoc block bitmaps causing BUG
71 Details : The locking of the mballoc buddy bitmap and the in-memory
72 block bitmap was using two different spin locks in some
73 cases. This made it possible to incorrectly access the
74 mballoc bitmap while another process was modifying it,
75 causing a sanity assertion to fail. While no on-disk corruption
76 was reported, there was some risk of this happening.
79 Frequency : rare, on llog test 6
81 Descriptoin: don't allow connect to already connected import
82 Details : allowing connect to already connected import is hide connecting problem.
87 Descriptoin: allow kill process which wait statahead result
88 Details : for some reasons 'ls' can stick in waiting result from statahead,
89 in this case need way for kill this process.
91 Severity : enhancement
93 Description: Add state history info file, enhance import info file
94 Details : Track import connection state changes in a new osc/mdc proc file;
95 add overview-type data to the osc/mdc import proc file.
97 Severity : enhancement
99 Description: MDS create should not wait for statfs RPC while holding DLM lock.
101 Severity : enhancement
103 Description: Update to RHEL5U3 kernel-2.6.18-128.1.1.el5.
108 Descriptoin: OST grant too much space to client even there are not enough space.
109 Details : Client will shrink its grant cache to OST if there are no write
110 activity over 6 mins (GRANT_SHRINK_INTERVAL), and OST will retrieve
111 this grant cache if there are already not enough avaible space
112 (left_space < total_clients * 32M).
115 Frequency : start MDS on uncleanly shutdowned MDS device
117 Descriptoin: ll_sync thread stay in waiting mds<>ost recovery finished
118 Details : stay in waiting mds<>ost recovery finished produce random bugs
119 due race between two ll_sync thread for one lov target. send
120 ACTIVATE event only if connect realy finished and import have
124 Frequency : rare, connect and disconnect target at same time
126 Descriptoin: ASSERTION(atomic_read(&imp->imp_inflight) == 0
127 Details : don't call obd_disconnect under lov_lock. this long time
128 operation and can block ptlrpcd which answer to connect request.
133 Descriptoin: don't lose wakeup for imp_recovery_waitq
134 Details : recover_import_no_retry or invalidate_import and import_close can
135 both sleep on imp_recovery_waitq, but we was send only one wakeup
139 Frequency : always with long access acl
141 Descriptoin: mds can't pack reply with long acl.
142 Details : mds don't control size of acl but they limited by reint/getattr
145 Severity : enhancement
147 Description: Update to SLES10 kernel-2.6.16.60-0.33.
149 Severity : enhancement
151 Description: Update to RHEL5 kernel-2.6.18-92.1.22.el5.
154 Frequency : start MDS on uncleanly shutdowned MDS device
156 Descriptoin: aborting recovery hang on MDS
157 Details : don't throttle destroy RPCs for the MDT.
160 Frequency : on remount
162 Description: external journal device not working after the remount
163 Details : clear dev_rdonly flag for external journal devices in
169 Description: shutdown vs evict race
170 Details : client_disconnect_export vs connect request race.
171 if client will evicted at this time - we start invalidate
172 thread without referece to import and import can be freed
176 Frequency : rare, need acl's on inode.
178 Description: client can't handle ost additional correctly
179 Details : if ost was added after client connected to mds client can have
180 hit lnet_try_match_md ... to big messages to wide striped files.
181 in this case need teach client to handle config events about add
182 lov target and update client max ea size at that event.
184 Severity : enhancement
186 Description: Changelogs
187 Details : Changelogs are a lightweight mechanism to track filesystem
188 metadata and namespace changes. The changelog is recorded
189 permanently on the MDTs, and is periodically "consumed" / purged
190 when records are no longer needed.
192 Severity : enhancement
194 Description: compact fld format with extents
195 Details : Store range of seq rather than every seq in FLD. Seq
196 controller update FLD rather than clients. In Case of CMD, mdt0
197 has FLD, all other metadata server act as non persistent proxy
198 for FLD queries and cache fld entries in fld cache.
203 Description: don't skip ost target if they assigned to file
204 Details : Drop slow OSCs if we can, but not for requested start idx.
205 This means "if OSC is slow and it is not the requested
206 start OST, then it can be skipped, otherwise skip it only
207 if it is inactive/recovering/out-of-space.
211 Description: more cleanup in mds_lov
212 Details : not send LOV EA under replay, we can't know about they size at this
213 time. Don't allow client connect to mds before any ost connected,
214 for avoid problems with LOV EA size and returning EIO to client.
216 Severity : enhancement
218 Description: Interoperability at server side (Disk interoperability)
220 Severity : enhancement
222 Description: Update to RHEL5 kernel-2.6.18-92.1.17.el5.
224 Severity : enhancement
226 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.31.
228 Severity : enhancement
230 Description: New client IO stack (CLIO).
232 Severity : enhancement
234 Description: Commit on sharing. Eliminate inter-client dependencies between
235 uncommitted transactions by doing transaction commits.
236 Thereby clients may recovery independently.
239 Frequency : Create a symlink file with a very long name
241 Description: ldlm_cancel_pack()) ASSERTION(max >= dlm->lock_count + count)
242 Details : If there is no extra space in the request for early cancels,
243 ldlm_req_handles_avail() returns 0 instead of a negative value.
245 Severity : enhancement
247 Description: Add /proc entry for import status
248 Details : The mdc, osc, and mgc import directories now have
249 an import directory that contains useful import data for debugging
252 Severity : enhancement
254 Description: Re-disable certain /proc logging
255 Details : Enable and disable client's offset_stats, extents_stats and
256 extents_stats_per_process stats logging on the fly.
259 Frequency : Only on FC kernels 2.6.22+
261 Description: oops in statahead
262 Details : Do not drop reference count for the dentry from VFS when lookup,
263 VFS will do that by itself.
265 Severity : enhancement
267 Description: Generic /proc file permissions
268 Details : Set /Proc file permissions in a more generic way to enable non-
269 root users operate on some /proc files.
273 Description: Hitting mdc_commit_close() ASSERTION
274 Details : Properly handle request reference release in
275 ll_release_openhandle().
279 Description: quota recovery deadlock during mds failover
280 Details : This patch includes att18982, att18236, att18237 in bz14840.
282 1. fix osts hang when mds does failover with quotaon
283 2. prevent watchdog storm when osts threads wait for the
288 Frequency : only patchless client
289 Description: add workaround for race between add/remove dentry from hash
291 Severity : enhancement
293 Description: Allow OST glimpses to return PW locks
297 Description: LBUG when llog conf file is full
298 Details : When llog bitmap is full, ENOSPC should be returned for plain log.
302 Description: Prevent import from entering FULL state when server in recovery
306 Description: service mount cannot take device name with ":"
307 Details : Only when device name contains ":/" will mount treat it as
313 Description: replace ptlrpcd with the statahead thread to interpret the async
314 statahead RPC callback
318 Frequency : on recovery
319 Description: I/O failures after umount during fail back
320 Details : if client reconnected to restarted server we need join to recovery
321 instead of find server handler is changed and process self
322 eviction with cancel all locks.
324 Severity : enhancement
326 Description: Update to RHEL5 kernel-2.6.18-92.1.10.el5.
328 Severity : enhancement
330 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.27.
332 Severity : enhancement
334 Description: Upcall on Lustre log has been dumped
335 Details : Allow for a user mode script to be called once a Lustre log has
336 been dumped. It passes the filename of the dumped log to the
337 script, the location of the script can be specified via
338 /proc/sys/lnet/debug_log_upcall.
343 Description: avoid idr_remove called for id which is not allocated.
344 Details : Move assigment s_dev for clustered nfs to end of initialization,
345 to avoid problem with error handling.
350 Description: avoid Already found the key in hash [CONN_UNUSED_HASH] messages
351 Details : When connection is reused this not moved from CONN_UNUSED_HASH
352 into CONN_USED_HASH and this prodice warning when put connection
353 again in unused hash.
356 Severity : enhancement
358 Description: File striping can now be set to use an arbitrary pool of OSTs.
360 Severity : enhancement
362 Description: Export bytes_read/bytes_write count on OSC/OST.
366 Description: Early reply size mismatch, MGC loses connection
367 Details : Apply the MGS_CONNECT_SUPPORTED mask at reconnect time so
368 the connect flags are properly negotiated.
373 Description: quotas are not honored with O_DIRECT
374 Details : all writes with the flag O_DIRECT will use grants which leads to
375 this problem. Now using OBD_BRW_SYNC to guard this.
379 Description: add quota statistics
380 Details : 1. sort out quota proc entries and proc code.
381 2. add quota statistics
383 Severity : enhancement
385 Description: enable quota support for HEAD.
389 Description: Properly propagate oinfo flags from lov to osc for statfs
390 Details : restore missing copy oi_flags to lov requests.
392 Severity : enhancement
394 Description: Add man pages for llobdstat(8), llstat(8), plot-llstat(8),
395 : l_getgroups(8), lst(8), routerstat(8)
396 Details : included man pages for llobdstat(8), llstat(8),
397 : plot-llstat(8), l_getgroups(8), lst(8), routerstat(8)
399 Severity : enhancement
401 Description: configure's --enable-quota should check the
402 : kernel .config for CONFIG_QUOTA
403 Details : configure is terminated if --enable-quota is passed but
404 : no quota support is in kernel
408 Description: Remove portals compatibility
409 Details : Remove portals compatibility, not interoperable with releases
414 Description: Resolve device initialization race
415 Details : Prevent proc handler from accessing devices added to the
416 obd_devs array but yet be intialized.
418 Severity : enhancement
420 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.23.
422 Severity : enhancement
424 Description: Update to RHEL5 kernel-2.6.18-92.1.6.el5.
429 Description: Using wrong pointer in osc_brw_prep_request
430 Details : Access to array[-1] can produce panic if kernel compiled with
431 CONFIG_PAGE_ALLOC enabled
435 Description: Client runs out of low memory
436 Details : Consider only lowmem when counting initial number of llap pages
440 Description: *optional* service tags registration
441 Details : if the "service tags" package is installed on a Lustre node
442 When the filesystem is mounted, a local-node service tag will
443 be created. See http://inventory.sun.com/ for more information
444 about the Service Tags asset management system.
448 Description: Kernel BUG tries to release flock
449 Details : Lustre does not destroy flock lock before last reference goes
450 away. So always drop flock locks when client is evicted and
451 perform unlock regardless of successfulness of speaking to MDS.
455 Description: add refcount protection for osc callbacks, avoid panic on shutdown
459 Description: sanity test 65a fails if stripecount of -1 is set
460 Details : handle -1 striping on filesystem in ll_dirstripe_verify
465 Description: ASSERTION(CheckWriteback(page,cmd)) failed
466 Details : badly clear PG_Writeback bit in ll_ap_completion can produce false
469 Severity : enhancement
471 Description: Update to RHEL5 kernel-2.6.18-53.1.21.el5.
475 Description: do not process already freed flock
476 Details : flock can possibly be freed by another thread before it reaches
477 to ldlm_flock_completion_ast.
481 Description: LBUG during stress test
482 Details : Need properly lock accesses the flock deadlock detection list.
486 Description: oops in page fault handler
487 Details : kernel page fault handler can return two special 'pages' in error
488 case, don't try dereference NOPAGE_SIGBUS and NOPAGE_OMM.
492 Description: timeout with invalidate import.
493 Details : ptlrpcd_check call obd_zombie_impexp_cull and wait request which
494 should be handled by ptlrpcd. This produce long age waiting and
495 -ETIMEOUT ptlrpc_invalidate_import and as result LASSERT.
497 Severity : enhancement
499 Description: Update to RHEL5 kernel-2.6.18-53.1.19.el5.
503 Description: enable MGS and MDT services start separately
504 Details : add a 'nomgs' option in mount.lustre to enable start a MDT with
505 a co-located MGS without starting the MGS, which is a complement
506 to 'nosvc' mount option.
510 Frequency : after recovery
511 Description: precreate to many object's after del orphan.
512 Details : del orphan st in oscc last_id == next_id and this triger growing
513 count of precreated objects. Set flag LOW to skip increase count
514 of precreated objects.
518 Frequency : rare, on clear nid stats
519 Description: ASSERTION(client_stat->nid_exp_ref_count == 0)
520 Details : when clean nid stats sometimes try destroy live entry,
521 and this produce panic in free.
525 Description: Stack overflow during MDS log replay
526 ease stack pressure by using a thread dealing llog_process.
530 Description: wait until IO finished before start new when do lock cancel.
531 Details : VM protocol want old IO finished before start new, in this case
532 need wait until PG_writeback is cleared until check dirty flag and
533 call writepages in lock cancel callback.
535 Severity : enhancement
537 Description: using special macro for print time and cleanup in includes.
541 Description: mds_mfd_close() ASSERTION(rc == 0)
542 Details : In mds_mfd_close(), we need protect inode's writecount change
543 within its orphan write semaphore to prevent possible races.
547 Description: Obsolete CURRENT_SECONDS and use cfs_time_current_sec() instead.
551 Frequency : rare, on shutdown ost
552 Description: don't hit live lock with umount ost.
553 Details : shrink_dcache_parent can be in long loop with destroy dentries,
554 use shrink_dcache_sb instead.
558 Description: don't panic with use echo client
559 Details : echo client pass NULL as client nid pointer and this produce null
564 Description: fix build on ppc32
565 Details : compile code with -m64 flag produce wrong object file for ppc32.
569 Description: add message levels for liblustreapi
573 Description: fix for occasional failure case of -ENOSPC in recovery-small tests
574 Details : Move the 'good_osts' check before the 'total_bavail' check. This
575 will result in an -EAGAIN and in the exit call path we call
576 alloc_rr() which will with increasing aggressiveness attempt to
577 aquire precreated objects on the minimum number of required OSCs.
581 Description: Use old size assignment to avoid deadlock
582 Details : Reverts the changes in bugs 2369 and bug 14138 that introduced
583 the scheduling while holding a spinlock. We do not need locking
584 for size in ll_update_inode() because size is only updated from
585 the MDS for directories or files without objects, so there is no
586 other place to do the update, and concurrent access to such inodes
587 are protected by the inode lock.
591 Description: resolve "_IOWR redefined" build error on SLES10
595 Description: dump the memory debugging after all modules are unloaded to
596 suppress false negative in conf_sanity test 39
598 Severity : enhancement
600 Description: build kernel-ib packages for OFED 1.3 in our release cycle
605 Description: fix SLES kernel versioning
606 Details : the kernel version for our SLES 10 kernel did not include a "-"
607 before the "smp" at the end. while this was not a problem in
608 general, it did mean that software trying to use the kernel
609 version to try to detect a vendor specific kernel would fail.
610 this was most evident by the OFED build scripts.
614 Description: Don't update lov_desc members until making sure they are valid
615 Details : When updating lov_desc members via proc fs, need fix their
616 validities before doing the real update.
620 Description: don't put request into delay list while invalidate in flight.
621 Details : ptlrpc_delay_request sometimes put in delay list while invalidate
622 import in flight. this produce timeout for invalidate and sometimes
623 can cause stale data.
627 Frequency : on ppc only
628 Description: not convert ost objects for directory because it's not exist.
629 Details : ll_dir_getstripe assume dirrectory has ost objects but this wrong.
633 Description: Add FMODE_EXEC file flag for SLES10 SP1 kernel.
635 Severity : enhancement
637 Description: Update to support 2.6.22.14 vanilla kernel.
641 Frequency : rare, on recovery
642 Description: read procfs can produce deadlock in some situation
643 Details : Holding lprocfs lock which send rpc can produce block for destroy
644 obd objects and this also block reconnect with -EALREADY. This
645 isn't fix all lprocfs bugs - but make it rare.
647 Severity : enhancement
649 Description: Update kernel to RHEL5 2.6.18-53.1.14.el5.
652 Frequency : frequent on X2 node
654 Description: mdc_set_open_replay_data LBUG
655 Details : Set replay data for requests that are eligible for replay.
659 Description: lustre_mgs: operation 101 on unconnected MGS
660 Details : When MGC is disconnected from MGS long enough, MGS will evict the
661 MGC, and late on MGC cannot successfully connect to MGS and a lot
662 of the error messages complaining that MGS is not connected.
666 Frequency : on network error
667 Description: panic with double free request if network error
668 Details : mdc_finish_enqueue is finish request if any network error ocuring,
669 but it's true only for synchronus enqueue, for async enqueue
670 (via ptlrpcd) this incorrect and ptlrpcd want finish request
673 Severity : enhancement
675 Description: client-side metadata stat-ahead during readdir(directory readahead)
676 Details : perform client-side metadata stat-ahead when the client detects
677 readdir and sequential stat of dir entries therein
680 Frequency : on start mds
682 Description: Implement get_info(last_id) in obdfilter.
685 Frequency : occasional
687 Description: Correctly check stale fid, not start epoch if ost not support SOM
688 Details : open with flag O_CREATE need set old fid in op_fid3 because
689 op_fid2 was overwritten with new generated fid, but mds can answer
690 with one of these two fids and both is not stale. Setattr
691 incorrectly started an epoch and assume will be called
692 done_writing, but without SOM done_writing ever being called.
695 Frequency : rare, depends on device drivers and load
697 Description: MDS or OSS nodes crash due to stack overflow
698 Details : Code changes in 1.8.0 increased the stack usage of some functions.
699 In some cases, in conjunction with device drivers that use a lot
700 of stack the MDS (or possibly OSS) service threads could overflow
701 the stack. One change which was identified to consume additional
702 stack has been reworked to avoid the extra stack usage.
705 Frequency : occasional
707 Description: Do not fail import if osc_interpret_create gets -EAGAIN
708 Details : If osc_interpret_create got -EAGAIN it immediately exits and
709 wakeup oscc_waitq. After wakeup oscc_wait_for_objects call
710 oscc_has_objects and see OSC has no objests and call
711 oscc_internal_create to resend create request.
713 Severity : enhancement
715 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.5.
717 Severity : enhancement
719 Description: Update to RHEL5 latest kernel-2.6.18-53.1.13.el5.
722 Frequency : very rare
724 Description: Fix replay if there is an un-replied request and open
725 Details : In some cases, older replay request will revert the
726 mcd->mcd_last_xid on MDS which is used to record the client's
729 Severity : enhancement
731 Description: Update to RHEL5 latest kernel-2.6.18-53.1.6.el5.
733 Severity : enhancement
735 Description: Add rhel5 support to HEAD.
737 Serverity : enhancement
739 Description: Update RHEL4 kernel to 2.6.9-67.0.4.
744 Description: Don't allow skipping OSTs if index has been specified.
745 Details : Don't allow skipping OSTs if index has been specified, make
746 locking in internal create lots better.
750 Description: LBUG in ptlrpc_check_set() bad phase ebc0de00
751 Details : access to bitfield in structure is always rounded to long
752 and this produce problem with not atomic change any bit.
756 Description: Lustre make rpms failed.
757 Details : Remove ldiskfs spec file to avoids rpmbuild be confused when
758 builds Lustre rpms from tarball.
761 Frequency : rare on shutdown ost
763 Description: If llog cancel was not send before clean_exports phase, this can
764 produce deadlock in llog code.
765 Details : If llog thread has last reference to obd and call class_import_put
766 this produce deadlock because llog_cleanup_commit_master wait when
767 last llog_commit_thread exited, but this never success because was
768 called from llog_commit_thread.
772 Description: allow userland application know is lost one of stripes.
773 Details : fill lvb_blocks with error code on ost and return it to
774 application if error flag found.
778 Description: NULL lov_tgts causing MDS oops
779 Details : more safe checks for NULL lov_tgts for avoid oops.
781 Severity : enhancement
783 Description: Update to RHEL4 latest kernel-2.6.9-67.0.1.EL.
787 Descriptoin: make lov_create() will not stuck in obd_statfs_rqset()
788 Details : If an OST is down the MDS will hang indefinitely in
789 obd_statfs_rqset() waiting for the statfs data. While for
790 MDS QOS usage of statfs, it should not stuck in waiting.
792 Severity : enhancement
794 Description: remote_acl support
795 Details : Support ACL-based permission check for remote user.
796 Support setfacl/getfacl for remote user with the utils
797 "lfs {l,r}{s,g}etfacl" which follow the same parameter format as
798 the system "{s,g}etfacl" utils.
800 Severity : enhancement
802 Description: Update to RHEL4 U6 kernel-2.6.9-67.EL.
804 Severity : enhancement
806 Description: Update to RHEL5 latest kernel-2.6.18-53.1.4.el5.
810 Description: make mgs_setparam() handle fsname containing dash
811 Details : fsname containing a dash does not work with lctl conf_param
813 Severity : enhancement
815 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.3.
817 Severity : enhancement
819 Description: Update to RHEL5 Update-1 kernel 2.6.18-53.el5.
823 Frequency : rare, at shutdown
824 Description: access already free / zero obd_namespace.
825 Details : if client_disconnect_export was called without force flag set,
826 and exist connect request in flight, this can produce access to
827 NULL pointer (or already free pointer) when connect_interpret
828 store ocd flags in obd_namespace.
832 Frequency : only at startup
833 Description: not alloc memory with spinlock held.
834 Details : allocation memory with GFP_KERNEL can produce sleep deadlock,
835 if any spinlock held.
837 Severity : enhancement
839 Description: make lustre randomly fail allocating memory
840 Details : Make lustre randomly failed allocating memory for testing purpose.
842 Severity : enhancement
844 Description: lost problems with lov objid file
845 Details : Fixes some scability and access to not inited memory problems
846 in work with lov objdid file.
851 Description: lfs find does not continue on file error
852 Details : Continue other files processing when a file/dir is absent.
856 Description: Inconsistent usage of lustre_pack_reply()
857 Details : Standardize the usage of lustre_pack_reply() such that it
858 always generate a CERROR on failure.
861 Frequency : occasional
863 Description: MDS hang or stay in waiting lock
864 Details : If client receive lock with CBPENDING flag ldlm need send lock
865 cancel as separate rpc, to avoid situation when cancel request
866 can't processed due all i/o threads stay in wait lock.
870 Description: Update to RHEL5 kernel 2.6.18-8.1.15.el5.
874 Description: Update to SLES10 SP1 kernel 2.6.16.53-0.16
878 Description: Update to SLES9 kernel-2.6.5-7.287.3.
882 Description: Update to RHEL4 latest kernel.
884 Severity : enhancement
886 Description: Build SLES10 patchless client fails
887 Details : The configure was broken by run ./configure with
888 --with-linux-obj=.... argument for patchless client. When the
889 configure use --with-linux-obj, the LINUXINCLUDE= -Iinclude
890 can't search header adequately. Use absolute path such as
891 -I($LINUX)/include instead.
895 Description: interrupt oig_wait produce painc on resend.
896 Details : brw_redo_request can be used for resend requests from ptlrpcd and
897 private set, and this produce situation when rq_ptlrpcd_data not
898 copyed to new allocated request and triggered LBUG on assert
899 req->rq_ptlrpcd_data != NULL. But this member used only for wakeup
900 ptlrpcd set if request is changed and can be safety changed to use
905 Description: LASSERT_{REQ,REP}SWAB macros are buggy
906 Details : If SWAB_PARANOIA is disabled, the LASSERT_REQSWAB and
907 LASSERT_REPSWAB macros become no-ops, which is incorrect. Drop
908 these macros and replace them with their difinitions instead.
912 Description: Update kernel patches for SLES10 2.6.16.53-0.8.
913 Details : Update which_patch & target file for SLES10 latest kernel.
916 Description: Remove client patches from SLES 10 kernel.
917 Details : This causes SLES 10 clients to behave as patchless clients
918 even on a Lustre-patched (server) kernel.
920 Severity : enhancement
922 Description: self-adjustable client's lru lists
923 Details : use adaptive algorithm for managing client cached locks lru
924 lists according to current server load, other client's work
925 pattern, memory activities, etc. Both, server and client
926 side namespaces provide number of proc tunables for controlling
929 Severity : enhancement
931 Description: light-weight GSS support
932 Details : Support krb5n and krb5a mode, which keep Kerberos 5 authentication
933 and reduce performance overhead.
935 Severity : enhancement
937 Description: Linux keyring support
938 Details : Support using service of Linux keyring for Lustre GSS internal
939 context refresh/cache mechanism.
943 Description: Fix errors in lfs documentation
944 Details : Fixes man pages
948 Description: don't use GFP_* in generic Lustre code.
949 Details : Use cfs_alloc_* functions and CFS_* flags for code portability.
953 Description: obdclass is limited by single OBD_ALLOC(idarray)
954 Details : replace OBD_ALLOC/OBD_FREE with OBD_VMALLOC/OBD_VFREE
958 Description: warnings with build patchless client with vanila 2.6.19 and up
959 Details : change old ctl_table style and replace ctl_table/ctl_table_header
960 with cfs_sysctl_table_t/cfs_sysctl_table_header_t
964 Frequency : Only for SLES
965 Description: sanity_quota fail test_1
966 Details : There are multiple occurences of $TSTUSR in SLES's /etc/group
967 file, which makes TSTID[2] inunique.
971 Frequency : Only for SLES9
972 Description: Kernel patches for SLES9 2.6.5-7.286 kernel
973 Details : Update target/ChangeLog/which_patch .
977 Frequency : Only for test_10 in sanity-quota.sh
978 Description: a bug in quota test code
979 Details : checking a test flag in a wrong place
983 Frequency : Only for test_13 in sanity-quota.sh
984 Description: a bug in quota test script
985 Details : shouldn't check overfull allocated quota
989 Description: jbd statistics
990 Details : Port older jbd statistics patch for sles10
992 Severity : enhancement
994 Description: Add printing inode info into message about error in writepage.
999 Description: reply_lock_interpret crash due to race with it and lock cancel.
1000 Details : Do not replay locks that are being cancelled. Do not reference
1001 locks by their address during replay, just by their handle.
1005 Frequency : When flocks are used.
1006 Description: assertion failure in ldlm_cli_enquque_fini for non NULL lock.
1007 Details : Flock locks might destroy just granted lock if it could be merged
1008 with another existing flock, this is done in completion handler,
1009 so teach ldlm_cli_enquque_fini that this is a valid case for
1015 Description: Oops in read and write path when failing to allocate lock.
1016 Details : Check if lock allocation failed and return error back.
1020 Description: lstripe command fails for valid OST index
1021 Details : The stripe offset is compared to lov->desc.ld_tgt_count
1022 instead of lov->desc.ld_active_tgt_count.
1026 Description: sanity.sh failed test 103
1027 Details : RHEL mis-interpret setfacl "-X" param, so we won't test setfacl
1032 Description: df doesn't work properly if diskfs blocksize != 4K
1033 Details : Choose biggest blocksize of OST's as the LOV's blocksize.
1037 Description: lfs find on -1 stripe looping in lsm_lmm_verify_common()
1038 Details : Avoid lov_verify_lmm_common() on directory with -1 stripe count.
1040 Severity : enhancement
1042 Description: Adaptive timeouts
1043 Details : RPC timeouts adapt to changing server load and network
1044 conditions to reduce resend attempts and improve recovery time.
1048 Description: llapi_file_create() does not allow some changes
1049 Details : add llapi_file_open() that allows specifying the mode and
1050 open flags, and also returns an open file handle.
1054 Description: merge and cleanup kernel patches.
1055 Details : Remove mnt_lustre_list in vfs_intent-2.6-rhel4.patch.
1059 Description: Add journal checksum support.(Kernel part)
1060 Details : The journal checksum feature adds two new flags i.e
1061 JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT and
1062 JBD2_FEATURE_COMPAT_CHECKSUM. JBD2_FEATURE_CHECKSUM flag
1063 indicates that the commit block contains the checksum for
1064 the blocks described by the descriptor blocks. Now commit
1065 record can be sent to disk without waiting for descriptor
1066 blocks to be written to disk. This behavior is controlled
1067 using JBD2_FEATURE_ASYNC_COMMIT flag.
1071 Description: OSS needs mutliple precreate threads
1072 Details : Add ability to start more than one create thread per OSS.
1076 Description: Sanity.sh test_65k failed: setstripe should have succeeded
1077 Details : execute lfs setstripe on client
1081 Description: mds_obd_create error creating tmp object
1082 Details : When the user sets quota on root, llog will be affected and can't
1083 create files and write files.
1086 Frequency : Always on ia64 patchless client, and possibly others.
1088 Description: Add EXPORT_SYMBOL check for node_to_cpumask symbol.
1089 Details : This allows the patchless client to be loaded on architectures
1090 without this export.
1094 Description: RedHat Update kernel for RHEL5
1095 Details : Add the kernel config file for RHEL5.
1099 Description: RedHat Update kernel for RHEL5
1100 Details : Modify the kernel config file more closer RHEL5.
1104 Description: Build failure against Centos5 (RHEL5)
1105 Details : Define PAGE_SIZE when it isn't present.
1109 Description: test_7 on sanity-quota.sh can't be used on separate machines
1110 Details : add facet to handle it
1114 Description: "ll_intent_file_open()) lock enqueue: err: -13" with nfs
1115 Details : with NFS, the anon dentry's parent was set to itself in
1116 d_alloc_anon(), so in MDS, we use rec->ur_fid1 to find the
1117 corresponding dentry other than use rec->ur_name.
1119 Severity : enhancement
1121 Description: lfs setstripe enhancement
1122 Details : Make lfs setstripe understand 'k', 'm' and 'g' for stripe size.
1126 Description: enable data checksumming by default
1127 Details : enable checksum by default, allow --disable-checksum
1128 configure option and "-o nochecksum" mount option. Checksums
1129 can also be disabled at runtime via $LPROC/osc/*/checksum_pages.
1133 Description: lustre support for RHEL5
1134 Details : Add support for RHEL5.
1138 Description: Client eviction due to failover config
1139 Details : after a connection loss, the lustre client should attempt to
1140 reconnect to the last active server first before trying the
1141 other potential connections.
1145 Description: when mds and osts use different quota unit(32bit and 64bit),
1146 quota will be released repeatly.
1147 Details : void sending multiple quota reqs to mds, which will keep the
1148 status between the reqs.
1152 Description: osts not allocated evenly to files
1153 Details : change the condition to increase offset_idx
1156 Frequency : Always for filesystems larger than 2TB on 32-bit systems.
1158 Description: Data corruption for OSTs that are formatted larger than 2TB
1160 Details : When generating the bio request for lustre file writes the
1161 sector number would overflow a temporary variable before being
1162 used for the IO. The data reads correctly from Lustre (which
1163 will overflow in a similar manner) but other file data or
1164 filesystem metadata may be corrupted in some cases.
1168 Description: Tune the kernel for good SCSI performance.
1169 Details : Set the value of /sys/block/{dev}/queue/max_sectors_kb
1170 to the value of /sys/block/{dev}/queue/max_hw_sectors_kb
1175 Description: rewrite ext2-derived code in obdclass/uuid.c
1176 Details : rewrite inherited code (uuid parsing code from ext2 utils)
1177 from scratch preserving functionality.
1181 Description: "lfs find -obd UUID" prints directories
1182 Details : "lfs find -obd UUID" will return all directory names instead
1183 of just file names. It is incorrect because the directories
1184 do not reside on the OSTs.
1188 Description: use i_size_read and i_size_write in 2.6 port
1189 Details : replace inode->i_size access with i_size_read/write()
1192 Frequency : when removing large files
1194 Description: scheduling issue during removal of large Lustre files
1195 Details : Don't take the BKL in fsfilt_ext3_setattr() for 2.6 kernels.
1196 It causes scheduling issues when removing large files (17TB in the
1200 Frequency : only with liblustre clients on XT3
1202 Description: evictions taking too long
1203 Details : allow llrd to evict clients directly on OSTs
1206 Frequency : only on ppc
1208 Description: /proc/fs/lustre/devices broken on ppc
1209 Details : The patch as applied to 1.6.2 doesn't look correct for all arches.
1210 We should make sure the type of 'index' is loff_t and then cast
1211 explicitly as needed below. Do not assign an explicitly cast
1215 Frequency : only for rhel5
1217 Description: Kernel patches update for RHEL5 2.6.18-8.1.10.el5.
1218 Details : Modify the target file & which_kernel.
1220 Severity : enhancement
1222 Description: omit set fsid for export NFS
1223 Details : fix set/restore device id for avoid EMFILE error and mark lustre
1224 fs as FS_REQUIRES_DEV for avoid problems with generate fsid.
1228 Frequency : Always, for kernels after 2.6.16
1229 Description: Fix warning idr_remove called for id=.. which is not allocated.
1230 Details : Last kernels save old s_dev before kill super and not allow
1231 to restore from callback - restore it before call kill_anon_super.
1235 Description: buffer overruns could theoretically occur
1236 Details : llapi_semantic_traverse() modifies the "path" argument by
1237 appending values to the end of the origin string, and a
1238 overrun may occur. Adding buffer overrun check in liblustreapi.
1242 Description: Fix error on 'ls .' at the top of the Lustre mount.
1243 Details : Don't revalidate dentry if it is a root dentry.
1247 Description: Kernel patches update for RHEL4 2.6.9-55.0.6.
1248 Details : Modify vm-tunables-rhel4.patch.
1252 Description: Kernel config for 2.6.18-vanilla.
1253 Details : Modify targets/2.6-vanilla.target.in.
1254 Add config file kernel-2.6.18-2.6-vanilla-i686.config.
1255 Add config file kernel-2.6.18-2.6-vanilla-i686-smp.config.
1256 Add config file kernel-2.6.18-2.6-vanilla-x86_64.config.
1257 Add config file kernel-2.6.18-2.6-vanilla-x86_64-smp.config.
1261 Description: improve handling recoverable errors
1262 Details : if request processig with error which can be recoverable on server
1263 request should be resend, otherwise page released from cache and
1268 Description: Kernel patches update for RHEL5 2.6.18-8.1.14.el5.
1269 Details : Modify target file & which_patch.
1270 A flaw was found in the IA32 system call emulation provided
1271 on AMD64 and Intel 64 platforms. An improperly validated 64-bit
1272 value could be stored in the %RAX register, which could trigger an
1273 out-of-bounds system call table access. An untrusted local user
1274 could exploit this flaw to run code in the kernel
1275 (ie a root privilege escalation). (CVE-2007-4573).
1279 Description: O_DIRECT bypasses client statistics.
1280 Details : When running with O_DIRECT I/O, neither the client rpc_stats nor
1281 read_ahead_stats were updated. copied the stats section from
1282 osc_send_oap_rpc() into async_internal().
1286 Description: Add jbd statistics patch for RHEL5 and 2.6.18-vanilla
1290 Description: change order of libsysio includes
1291 Details : '#include sysio.h' should always come before '#include xtio.h'
1295 Description: handle "serious error: objid * already exists" more gracefully
1296 Details : If LAST_ID value on disk is smaller than the objects existing in
1297 the O/0/d* directories, it indicates disk corruption and causes an
1298 LBUG(). If the object is 0-length, then we should use the existing
1299 object. This will help to avoid a full fsck in most cases.
1301 Severity : enhancement
1303 Description: adapt the lustre_config script to support the upgrade case
1304 Details : Add "-u" option for lustre_config script to support upgrading 1.4
1305 server targets to 1.6 in parallel.
1309 Description: To avoid grant space > avaible space when the disk is almost
1310 full. Without this patch you might see the error "grant XXXX >
1311 available" or some LBUG about grant, when the disk is almost
1313 Details : In filter_check_grant, for non_grant cache write, we should
1314 check the left space by if (*left > ungranted + bytes), instead
1315 of (*left > ungranted), because only we are sure the left space
1316 is enough for another "bytes", then the ungrant space should be
1317 increase. In client, we should update cl_avail_grant only there
1318 is OBD_MD_FLGRANT in the reply.
1322 Description: Update RHEL 4 kernel to fix local root privilege escalation.
1323 Details : Update to the latest RHEL 4 kernel to fix the vulnerability
1324 described in CVE-2007-4573. This problem could allow untrusted
1325 local users to gain root access.
1328 Frequency : when using O_DIRECT and quotas
1330 Description: Incorrect file ownership on O_DIRECT output files
1331 Details : block usage reported by 'lfs quota' does not take into account
1332 files that have been written with O_DIRECT.
1337 Description: touch file failed when fs is not full
1338 Details : OST in recovery should not be discarded by MDS in alloc_qos(),
1339 otherwise we can get ENOSP while fs is not full.
1343 Description: parallel lock callbacks
1344 Details : Instead of sending blocking and completion callbacks as separated
1345 requests, adding them to a set and sending in parallel.
1348 Frequency : only for Cray XT3
1349 Bugzilla : 12829/13455
1350 Description: Changing primary group doesn't change the group lustre assigns to
1352 Details : When CRAY_XT3 is defined, the fsgid supplied by the client is
1353 overridden with the primary group provided by the group upcall,
1354 whereas the supplied fsgid can be trusted if it is in the list of
1355 supplementary groups returned by the group upcall.
1357 Severity : enhancement
1359 Description: Allow masking D_WARNING, D_ERROR messages from console
1360 Details : Console messages can now be disabled via lnet.printk.
1364 Description: User code with malformed file open parameter crashes client node
1365 Details : Before packing join_file req, all the related reference should be
1366 checked carefully in case some malformed flags cause fake
1367 join_file req on client.
1371 Description: LDLM_ENQUEUE races with LDLM_CP_CALLBACK
1372 Details : ldlm_completion_ast() assumes that a lock is granted when the req
1373 mode is equal to the granted mode. However, it should also check
1374 that LDLM_FL_CP_REQD is not set.
1378 Description: Heavy nfs access might result in deadlocks
1379 Details : After ELC code landed, it is now improper to enqueue any mds
1380 locks under och_sem, because enqueue might want to decide to
1381 cancel open locks for same inode we are holding och_sem for.
1385 Description: Client eviction while running blogbench
1386 Details : A lot of unlink operations with concurrent I/O can lead to a
1387 deadlock causing evictions. To address the problem, the number of
1388 oustanding OST_DESTROY requests is now throttled to
1389 max_rpcs_in_flight per OSC and LDLM_FL_DISCARD_DATA blocking
1390 callbacks are processed in priority.
1394 Description: enable ACLs on MDS by default
1395 Details : ACLs must be enabled on MDS by default.
1398 Frequency : PPC/PPC64 only
1400 Description: conflicts between asm-ppc64/types.h and lustre_types.h
1401 Details : fix duplicated definitions between asm-ppc64/types.h and
1402 lustre_types.h on PPC.
1405 Frequency : PPC/PPC64 only
1407 Description: asm-ppc/segment.h does not exist
1408 Details : fix compile issue on PPC.
1412 Description: better handle error messages in extents code
1415 Frequency : RHEL4 only
1417 Description: mkfs is very slow on IA64/RHEL4
1418 Details : A performance regression has been discovered in the MPT Fusion
1419 driver between versions 3.02.73rh and 3.02.99.00rh. As a
1420 consequence, we have downgraded the MPT Fusion driver in the RHEL4
1421 kernel from 3.02.99.00 to 3.02.73 until this problem is fixed.
1423 Severity : enhancement
1425 Description: SNMP support enhancement
1426 Details : Adding total number of sampled request for an MDS node in snmp
1429 Severity : enhancement
1431 Description: Optimize ldlm waiting list processing for PR extent locks
1432 Details : When processing waiting list for read extent lock and meeting read
1433 lock that is same or wider to it that is not contended, skip
1434 processing rest of the list and immediatelly return current
1435 status of conflictness, since we are guaranteed there are no
1436 conflicting locks in the rest of the list.
1440 Description: Time out and refuse to reconnect
1441 Details : When the failover node is the primary node, it is possible
1442 to have two identical connections in imp_conn_list. We must
1443 compare not conn's pointers but NIDs, otherwise we can defeat
1444 connection throttling.
1448 Description: port llog fixes from b1_6 into HEAD
1449 Details : Port llog reference couting and some llog cleanups from b1_6
1450 (bug 10800) into HEAD, for protect from panic and access to already
1451 free llog structures.
1455 Description: Detect stride IO mode in read-ahead
1456 Details : When a client does stride read, read-ahead should detect that and
1457 read-ahead pages according to the detected stride pattern.
1461 Description: data checksumming impacts single node performance
1462 Details : add support for several checksum algorithm. Currently, only CRC32
1463 and Adler-32 are supported. The checksum type can be changed on
1464 the fly via /proc/fs/lustre/osc/*/checksum_type.
1468 Description: use adler32 for page checksums
1469 Details : when available, use the Adler-32 algorithm instead of CRC32 for
1474 Description: build for x2 fails
1475 Details : fix compile issue on Cray systems.
1479 Description: Properly match for duplicate locks
1480 Details : Due to different lock order from skiplists code, we need to
1481 traverse entire list for now
1484 Frequency : only on PPC/SLES10
1486 Description: "BITS_PER_LONG is not 32 or 64" in linux/idr.h
1487 Details : On SLES10/PPC, fs.h includes idr.h which requires BITS_PER_LONG to
1488 be defined. Add a hack in mkfs_lustre.c to work around this compile
1493 Description: LASSERT on MDS when client holding flock lock dies
1494 Details : ldlm pool logic depends on number of granted locks equal to
1495 number of released locks which is not true for flock locks, so
1496 just exclude such locks from consideration.
1500 Description: MDS deadlock with many ll_sync_lov threads and I/O stalled
1501 Details : Use fsfilt_sync() for both the whole filesystem sync and
1502 individual file sync to eliminate dangerous inode locking
1503 with I_LOCK that can lead to a deadlock.
1507 Description: performance in 1.6.3
1508 Details : Force q->max_phys_segments to MAX_PHYS_SEGMENTS on SLES10 to be
1509 sure that 1MB requests are not fragmented by the block layer.
1511 Severity : enhancement
1513 Description: organize the server-side client stats on per-nid basis
1514 Details : Change the structure of stats under obdfilter and mds to
1522 The "uuid"s file would list the uuids of _active_ exports.
1523 And the clear entry is to clear all stats and stale nids.
1525 Severity : enhancement
1527 Description: eliminate client locks in face of contention
1528 Details : file contention detection and lockless i/o implementation
1529 for contended files.
1533 Description: Reinitialize optind to 0 so that interactive lfs works in all cases
1536 Frequency : very rare, if additional xattrs are used on kernels >= 2.6.12
1538 Description: MDS may lose file striping (and hence file data) in some cases
1539 Details : If there are additional extended attributes stored on the MDS,
1540 in particular ACLs, SELinux, or user attributes (if user_xattr
1541 is specified for the client mount options) then there is a risk
1542 of attribute loss. Additionally, the Lustre file striping
1543 needs to be larger than default (e.g. striped over all OSTs),
1544 and an additional attribute must be stored initially in the
1545 inode and then increase in size enough to be moved to the
1546 external attribute block (e.g. ACL growing in size) for file
1551 Description: skiplist implementation simplification
1552 Details : skiplists are used to group compatible locks on granted list
1553 that was implemented as tracking first and last lock of each
1554 lock group the patch changes that to using doubly linked lists
1558 Description: MDS LBUG: ASSERTION(!IS_ERR(dchild))
1559 Details : Change LASSERTs to client eviction (i.e. abort client's recovery)
1560 because LASSERT on both the data supplied by a client, and the
1561 data on disk is dangerous and incorrect.
1563 Severity : enhancement
1565 Description: Slow truncate/writes to huge files at high offsets.
1566 Details : Directly associate cached pages to lock that protect those pages,
1567 this allows us to quickly find what pages to write and remove
1568 once lock callback is received.
1572 Description: more ldlm soft lockups
1573 Details : In ldlm_resource_add_lock(), call to ldlm_resource_dump()
1574 starve other threads from the resource lock for a long time in
1575 case of long waiting queue, so change the debug level from
1576 D_OTHER to the less frequently used D_INFO.
1578 Severity : enhancement
1580 Description: add -gid, -group, -uid, -user options to lfs find
1584 Description: Hung threads in invalidate_inode_pages2_range
1585 Details : The direct IO path doesn't call check_rpcs to submit a new RPC
1586 once one is completed. As a result, some RPCs are stuck in the
1587 queue and are never sent.
1591 Description: filter threads hungs on waiting journal commit
1592 Details : Cleanup filter group llog code, then only filter group llog will
1593 be only created in the MDS/OST syncing process.
1597 Description: Procfs and llog threads access destoryed import sometimes.
1598 Details : Sync the import destoryed process with procfs and llog threads by
1599 the import refcount and semaphore.
1601 Severity : enhancement
1603 Description: openlock cache of b1_6 port to HEAD
1608 Description: kernel BUG at ldiskfs2_ext_new_extent_cb
1609 Details : If insertion of an extent fails, then discard the inode
1610 preallocation and free data blocks else it can lead to duplicate
1615 Description: don't always update ctime in ext3_xattr_set_handle()
1616 Details : Current xattr code updates inode ctime in ext3_xattr_set_handle.
1617 In some cases the ctime should not be updated, for example for
1618 2.0->1.8 compatibility it is necessary to delete an xattr and it
1619 should not update the ctime.
1623 Bugzilla : 15713/16362
1624 Description: Assertion in iopen_connect_dentry in 1.6.3
1625 Details : looking up an inode via iopen with the wrong generation number can
1626 populate the dcache with a disconneced dentry while the inode
1627 number is in the process of being reallocated. This causes an
1628 assertion failure in iopen since the inode's dentry list contains
1629 both a connected and disconnected dentry.
1633 Description: assertion failure in ldlm_handle2lock()
1634 Details : fix a race between class_handle_unhash() and class_handle2object()
1635 introduced in lustre 1.6.5 by bug 13622.
1640 Description: Kernel BUG: sd_iostats_bump: unexpected disk index
1641 Details : remove the limit of 256 scsi disks in the sd_iostat patch
1646 Description: oops in sd_iostats_seq_show()
1647 Details : unloading/reloading the scsi low level driver triggers a kernel
1648 bug when trying to access the sd iostat file.
1653 Description: Kernel panics during QLogic driver reload
1654 Details : REQ_BLOCK_PC requests are not handled properly in the sd iostat
1655 patch, causing memory corruption.
1660 Description: journal_dev option does not work in b1_6
1661 Details : pass mount option during pre-mount.
1663 Severity : enhancement
1665 Description: Add a FIEMAP(FIle Extent MAP) ioctl
1666 Details : FIEMAP ioctl will allow an application to efficiently fetch the
1667 extent information of a file. It can be used to map logical blocks
1668 in a file to physical blocks in the block device.
1672 Description: LDLM soft lockups - improvement
1673 Details : It is be possible to send the lock handle along with each read
1674 or write request because the client is already doing a lock match
1675 itself so there isn't any reason the OST should have to re-do that
1679 Frequency : only on Cray X2
1681 Description: X2 build failures
1682 Details : fix build failures on Cray X2.
1686 Description: xid & resent requests
1687 Details : Initialize RPC XID from clock at startup (randomly if clock is
1690 Severity : enhancement
1692 Description: Add lustre_start utility to start or stop multiple Lustre servers
1697 Description: Lustre GPF in {:ptlrpc:ptlrpc_server_free_request+373}
1698 Details : In case of memory pressure, list_del() can be called twice on
1699 req->rq_history_list, causing a kernel oops.
1703 Description: kptllnd_peer_check_sends()) ASSERTION(!in_interrupt()) failed
1704 Details : fix stack overflow in the distributed lock manager by defering
1705 export eviction after a failed AST to the elt thread instead of
1706 handling it in the dlm interpret routine.
1710 Description: Convert some comments to new format.
1711 Details : Update documenting comments to match doxygen conventions.
1715 Description: Grammar fixes.
1716 Details : A couple of trivial spelling fixes.
1720 Description: OSD_COUNTERS-mandatory
1721 Details : Make previously optional ->oti_{w,r}_locks sanity checks mandatory
1722 to simplify the code.
1726 Description: simplify cmm_device freeing logic.
1727 Details : Call cmm_device_free() in the failure path of cmm_device_alloc().
1731 Description: Add lockdep support to dt_object_operations locking interface.
1732 Details : Augment ->do_{read,write}_lock() prototypes with a `role'
1733 parameter indicating lock ordering. Update mdd code to use new
1738 Description: Introduce failloc constants for lockless IO tests.
1739 Details : Add two new failloc constants to test lockless IO. Only one of
1740 them in implemented---another is checked in yet to be landed
1745 Description: Add lockdep support for inode mutex.
1746 Details : Introduce and use new LOCK_INODE_MUTEX_PARENT() macro to be used
1747 in the situations where Lustre has to lock more than one inode
1752 Description: Add optional invariants checking support.
1753 Details : Add new LINVRNT() macro, optional on new --enable-invariants
1754 configure switch. This macro is to be used for consistency and
1755 sanity checks that are too expensive to be left in `production'
1760 Description: Zap lock->l_granted_mode with explicit LCK_MINMODE.
1761 Details : Use LCK_MINMODE rather than 0 to reset lock->l_granted_mode to
1766 Description: Add lockdep support for ldlm_lock and ldlm_resource.
1767 Details : Use spin_lock_nested() in (the only) situation where more than
1768 one ldlm_lock is locked simultaneously. Also, fix possible
1769 dead-lock in ldlm_lock_change_resource() by enforcing particular
1774 Description: Use struct ldlm_callback_suite in ldlm_lock_create().
1775 Details : Instead of specifying each ldlm_lock call-back through separate
1776 parameter, wrap them into struct ldlm_callback_suite.
1780 Description: Kill join_lru obd method and its callers.
1781 Details : CLIO uses lock weighting policy to keep locks over mmapped regions
1782 in memory---a requirement implemented through ->o_join_lru() obd
1783 method in HEAD. Remove this method and its users.
1787 Description: Add asynchronous ldlm ENQUEUE completion handler.
1788 Details : CLIO posts ENQUEUE requests asynchronously through ptlrpcd---a
1789 case that stock ldlm_completion_ast() cannot handle as it waits
1790 until lock is granted. Introduce new ldlm_completion_ast_async()
1791 for this. Also comment ldlm_completion_ast().
1795 Description: ldlm_error <-> errno conversion.
1796 Details : Add functions to map (rather arbitrary) between LDLM error codes
1797 and standard errno values. CLIO needs this to prevent LDLM specific
1798 constants from escaping ldlm and osc.
1802 Description: Kill unused ldlm_handle2lock_ns() function.
1803 Details : Kill unused ldlm_handle2lock_ns() function.
1807 Description: Add lu_ref support to ldlm_lock
1808 Details : lu_ref support for ldlm_lock and ldlm_resource. See lu_ref patch.
1809 lu_ref fields ->l_reference and ->lr_reference are added to ldlm_lock
1810 and ldlm_resource. LDLM interface has to be changed, because code that
1811 releases a reference on a lock, has to "know" what reference this is.
1812 In the most frequent case
1814 lock = ldlm_handle2lock(handle);
1816 LDLM_LOCK_PUT(lock);
1818 no changes are required. When any other reference (received _not_ from
1819 ldlm_handle2lock()) is released, LDLM_LOCK_RELEASE() has to be called
1820 instead of LDLM_LOCK_PUT().
1822 Arguably, changes are pervasive, and interface requires some discipline
1823 for proper use. On the other hand, it was very instrumental in finding
1824 a few leaked lock references.
1828 Description: Add ldlm_lock_addref_try().
1829 Details : Introduce ldlm_lock_addref_try() function (used by CLIO) that
1830 attempts to addref a lock that might be being canceled
1835 Description: Add ldlm_weigh_callback().
1836 Details : Add new ->l_weigh_ast() call-back to ldlm_lock. It is called
1837 by ldlm_cancel_shrink_policy() to estimate lock "value", instead of
1838 hard-coded `number of pages' logic.
1842 Description: Add lockdep annotations to llog code.
1843 Details : Use appropriately tagged _nested() locking calls in the places
1844 where llog takes more than one ->lgh_lock lock.
1848 Description: Add loi_kms_set().
1849 Details : Wrap kms updates into a helper function.
1853 Description: Constify instances of struct lsm_operations.
1854 Details : Constify instances of struct lsm_operations.
1858 Description: lu_conf support.
1859 Details : On a server, a file system object is uniquely identified
1860 by a fid, which is sufficient to locate and load all object
1861 state (inode). On a client, on the other hand, more data are
1862 necessary instantiate an object. Change lu_object_find() and
1863 friends to take additional `lu_conf' argument describing object.
1864 Typically this includes layout information.
1868 Description: lu_context fixes.
1869 Details : Introduce new lu_context functions that are needed on the client
1870 side, where some system threads (ptlrpcd) are shared by multiple
1871 modules, and so cannot be stopped during module shutdown.
1875 Description: Add start and stop methods to lu_device_type_operations.
1876 Details : Introduce two new methods in lu_device_type_operations, that are
1877 invoked when first instance of a given type is created and last one
1878 is destroyed respectively. This is need by CLIO.
1882 Description: Add lu_ref support to struct lu_device.
1883 Details : Add lu_ref support to lu_object and lu_device. lu_ref is used to
1884 track leaked references.
1888 Description: Introduce lu_kmem_descr.
1889 Details : lu_kmem_descr and its companion interface allow to create
1890 and destroy a number of kmem caches at once.
1894 Description: Fix lu_object finalization race.
1895 Details : Fix a race between lu_object_find() finding an object and its
1896 concurrent finalization. This race is (most likely) not possible
1897 on the server, but might happen on the client.
1901 Description: Introduce lu_ref interface.
1902 Details : lu_ref is a debugging module allowing to track references to
1903 a given object. It is quite cpu expensive, and has to be
1904 explicitly enabled with --enable-lu_ref. See usage description
1909 Description: Factor lu_site procfs stats into a separate function.
1910 Details : Separate lu_site stats printing code into a separate function
1911 that can be reused on a client.
1915 Description: Constify instances of struct {lu,dt,md}_device_operations.
1916 Details : Constify instances of struct {lu,dt,md}_device_operations.
1920 Description: Introduce struct md_site and move meta-data specific parts of
1921 struct lu_site here.
1922 Details : Move md-specific fields out of struct lu_site into special struct
1923 md_site, so that lu_site can be used on a client.
1927 Description: Kill mdd_lov_destroy().
1928 Details : Remove unused mdd code.
1932 Description: Add st_block checking to multistat.c.
1933 Details : Add st_block checking to multistat.c.
1937 Description: Add lu_ref support to struct obd_device.
1938 Details : Add lu_ref tracking to obd_device.
1942 Description: Kill obd_set_fail_loc().
1943 Details : Remove unused code.
1947 Description: Add special type for ptlrpc_request interpret functions.
1948 Details : Add lu_env parameter to ->rq_interpreter call-back. NULL is passed
1949 there. Actual usage will be in CLIO.
1953 Description: Replace RW_LOCK_UNLOCKED() macro with rwlock_init().
1954 Details : Replace RW_LOCK_UNLOCKED() with rwlock_init() as the former
1955 doesn't work with lockdep.
1959 Description: Add rwv.c test program.
1960 Details : New testing program exercising readv(2) and writev(2) (Qian).
1964 Description: Add sendfile.c test program.
1965 Details : New testing program exercising sendfile(2) (Jay).
1969 Description: Ratelimit a message that can be very frequent.
1970 Details : Ratelimit a memory allocation failure message that can
1975 Description: Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
1976 Details : Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
1980 Description: Liblustre build fixes.
1981 Details : Liblustre build fixes.
1985 Description: libcfs: add cfs_{need,cond}_resched() interface.
1986 Details : libcfs: add cfs_{need,cond}_resched() definition and
1987 implementations for Linux, NT, and liblustre.
1989 Severity : enhancement
1991 Description: More exported tunables for mballoc
1992 Details : Add support for tunable preallocation window and new tunables for
1993 large/small requests
1997 Description: Detect corruption of block bitmap and checking for preallocations
1998 Details : Checks validity of on-disk block bitmap. Also it does better
1999 checking of number of applied preallocations. When corruption is
2000 found, it turns filesystem readonly to prevent further corruptions.
2004 Description: (rw.c:1323:ll_read_ahead_pages()) ASSERTION(page_idx > ria->ria_stoff) failed
2005 Details : Once the unmatched stride IO mode is detected, shrink the stride-ahead
2006 window to 0. If it does hit cache miss, and read-pattern is still
2007 stride-io mode, does not reset the stride window, but also does not
2008 increase the stride window length in this case.
2012 Frequency : only for big-endian servers
2013 Description: Check if system is big-endian while mounting fs with extents feature
2014 Details : Mounting a filesystem with extents feature will fail on big-endian
2015 systems since ext3-based ldiskfs is not supported on big-endian
2016 systems. This can be over-riden with "bigendian_extents" mount option.
2018 Severity : enhancement
2020 Description: The root squash functionality
2021 Details : A security feature, which is to prevent users from being able
2022 to mount lustre on their desktop, run as root, and delete
2023 all of the files in the filesystem. The goal is accomplished by
2024 remapping user id (UID) and group id (GID) of the root user to
2025 a UID and GID specified by the system administartor via Lustre
2026 configuration management server (MGS). The functionality also
2027 allows to specify sets of clients for which the remapping does
2032 Description: Excessive recovery window
2033 Details : With AT enabled, the recovery window can be excessively long (6000+
2034 seconds). To address this problem, we no longer use
2035 OBD_RECOVERY_FACTOR when extending the recovery window (the connect
2036 timeout no longer depends on the service time, it is set to
2037 INITIAL_CONNECT_TIMEOUT now) and clients report the old service
2038 time via pb_service_time.
2042 Description: Watchdog triggered on MDS failover
2043 Details : enable OBD_CONNECT_MDT flag when connecting from the MDS so that
2044 the OSTs know that the MDS "UUID" can be reused for the same export
2045 from a different NID, so we do not need to wait for the export to be
2049 Frequency : rare, only if using MMP with Linux RAID
2051 Description: MMP doesn't work with Linux RAID
2052 Details : While using HA for Lustre servers with Linux RAID, it is possible
2053 that MMP will not detect multiple mounts. To make this work we
2054 need to unplug the device queue in RAID when the MMP block is being
2055 written. Also while reading the MMP block, we should read it from
2056 disk and not the cached one.
2058 Severity : enhancement
2060 Description: open file using fid
2061 Details : A file can be opened using just its fid, like
2062 <mntpt>/.lustre/fid/SEQ:OID:VER - this is needed for HSM and replication
2065 Frequency : Only in RHEL5 when mounting multiple ext3 filesystems
2068 Description: "kmem_cache_create: duplicate cache jbd_4k" error message
2069 Details : add proper locking for creation of jbd_4k slab cache
2073 Description: MMP check in ext3_remount() fails without displaying any error
2074 Details : When multiple mount protection fails during remount, proper error
2077 --------------------------------------------------------------------------------
2079 2007-08-10 Cluster File Systems, Inc. <info@clusterfs.com>
2081 * Support for kernels:
2082 2.6.5-7.283 (SLES 9),
2083 2.6.9-55.EL (RHEL 4),
2084 2.6.16.46-0.14 (SLES 10),
2085 2.6.18.8 vanilla (kernel.org)
2086 * Client support for unpatched kernels:
2087 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
2088 2.6.16 - 2.6.22 vanilla (kernel.org)
2089 * Due to recently discovered recovery problems, we do not recommend
2090 using patchless RHEL 4 clients with this or any earlier release.
2091 * Recommended e2fsprogs version: 1.39.cfs8
2092 * Note that reiserfs quotas are disabled on SLES 10 in this kernel.
2093 * Starting with this release, the ldiskfs backing filesystem required
2094 by Lustre is now in its own package, lustre-ldiskfs. This package
2095 should be installed. It is versioned separately from Lustre and
2096 may be released separately in future.
2100 Description: block reactivating mgc import until all deactivates complete
2101 Details : Fix race when failing back MDT/MGS to itself (testing)
2104 Frequency : at statup only
2106 Description: mds_lov_synchronize race leads to various problems
2107 Details : simultaneous MDT->OST connections at startup can cause the
2108 sync to abort, leaving the OSC in a bad state.
2110 Severity : enhancement
2112 Description: add optional extra BUILD_VERSION info
2113 Details : add a new environment variable (namely LUSTRE_VERS) which allows
2114 to override the lustre version.
2117 Frequency : 2.6.18 servers only
2119 Description: ll_kern_mount() doesn't release the module reference
2120 Details : The ldiskfs module reference count never drops down to 0
2121 because ll_kern_mount() doesn't release the module reference.
2126 Description: server LBUG when using old ost_num_threads parameter
2127 Details : Accept the old ost_num_threads parameter but warn that it
2128 is deprecated, and fix an off-by-one error that caused an LBUG.
2133 Description: Transient SCSI error results in persistent IO issue
2134 Details : iobuf->dr_error is not reinitialized to 0 between two
2138 Frequency : sometimes when underlying device returns I/O errors
2140 Description: OSTs not going read-only during write failures
2141 Details : OSTs are not remounted read-only when the journal commit threads
2142 get I/O errors because fsfilt_ext3 calls journal_start/stop()
2143 instead of the ext3 wrappers.
2147 Description: poor connect scaling with increasing client count
2148 Details : Don't run filter_grant_sanity_check for more than 100 exports
2149 to improve scaling for large numbers of clients.
2152 Frequency : SLES10 only
2154 Description: sanity-quota.sh quotacheck failed: rc = -22
2155 Details : Quotas cannot be enabled on SLES10.
2158 Frequency : liblustre clients only
2160 Description: getdirentries does not give error when run on compute nodes
2161 Details : getdirentries does not fail when the size specified as an argument
2162 is too small to contain at least one entry
2164 Severity : enhancement
2166 Description: Add LNET router traceability for debug purposes
2167 Details : If a checksum failure occurs with a router as part of the
2168 IO path, the NID of the last router that forwarded the bulk data
2169 is printed so it can be identified.
2174 Description: OST "spontaneously" evicts client; client has imp_pingable == 0
2175 Details : Due to a race condition, liblustre clients were occasionally
2176 evicted incorrectly.
2178 Severity : enhancement
2180 Description: lfs setstripe use optional parameters instead of postional
2183 Severity : enhancement
2185 Description: Nanosecond timestamp support for ldiskfs
2186 Details : The on-disk ldiskfs filesystem has added support for nanosecond
2187 resolution timestamps. There is not yet support for this at
2188 the Lustre filesystem level.
2191 Frequency : during server recovery
2193 Description: MDS failing to send precreate requests due to OSCC_FLAG_RECOVERING
2194 Details : request with rq_no_resend flag not awake l_wait_event if they get
2198 Frequency : nfs export on patchless client
2200 Description: connectathon hang when test nfs export over patchless client
2201 Details : Disconnected dentry cannot be found with lookup, so we do not need
2202 to unhash it or make it invalid
2205 Description: fix llapi_lov_get_uuids() to allow many OSTs to be returned
2206 Details: : Change llapi_lov_get_uuids() to read the UUIDs from /proc instead
2207 of using an ioctl. This allows lfsck for > 160 OSTs to succeed.
2212 Description: open req refcounting wrong on reconnect
2213 Details : If reconnect happened between getting open reply from server and
2214 call to mdc_set_replay_data in ll_file_open, we will schedule
2215 replay for unreferenced request that we are about to free.
2216 Subsequent close will crash in variety of ways.
2217 Check that request is still eligible for replay in
2218 mdc_set_replay_data().
2223 Description: disable writes to filesystem when reading health_check file
2224 Details : the default for reading the health_check proc file has changed
2225 to NOT do a journal transaction and write to disk, because this
2226 can cause reads of the /proc file to hang and block HA state
2227 checking on a healthy but otherwise heavily loaded system. It
2228 is possible to return to the previous behaviour during configure
2229 with --enable-health-write.
2231 Severity : enhancement
2233 Description: 64-bit inode version
2234 Details: : Add a on-disk 64-bit inode version for ext3 to track changes made
2235 to the inode. This will be required for version-based recovery.
2240 Description: MDS fails to start if a duplicate client export is detected
2241 Details : in some rare cases it was possible for a client to connect to
2242 an MDS multiple times. Upon recovery the MDS would detect this
2243 and fail during startup. Handle this more gracefully.
2245 Severity : enhancement
2247 Description: Add -o localflock option to simulate old noflock
2249 Details : This will achieve local-only flock/fcntl locks
2255 Description: log_commit_thread vs filter_destroy race leads to crash
2256 Details : Take import reference before releasing llog record semaphore
2261 Description: Wrong request locking in request set processing
2262 Details : ptlrpc_check_set wrongly uses req->rq_lock for proctect add to
2263 imp_delayed_list, in this place should be used imp_lock.
2266 Frequency : when reconnection
2268 Description: Grant Leak when osc reconnect to OST
2269 Details : When osc reconnect ost, OST(filter) should check whether it
2270 should grant more space to client by comparing fed_grant and
2271 cl_avail_grant, and return the granted space to client instead
2272 of "new granted" space, because client will call osc_init_grant
2273 to update the client grant space info.
2276 Frequency : when client reconnect to OST
2278 Description: Grant Leak when osc do resend and replay bulk write
2279 Details : When osc reconnect to OST, OST(filter)should clear grant info of
2280 bulk write request, because the grant info while be sync between
2281 OSC and OST when reconnect, and we should ignore the grant info
2282 these of resend/replay write req.
2287 Description: Grant space more than avaiable left space sometimes.
2288 Details : When then OST is about to be full, if two bulk writing from
2289 different clients came to OST. Accord the avaliable space of the
2290 OST, the first req should be permitted, and the second one
2291 should be denied by ENOSPC. But if the seconde arrived before
2292 the first one is commited. The OST might wrongly permit second
2293 writing, which will cause grant space > avaiable space.
2296 Frequency : when client is evicted
2298 Description: Grant might be wrongly erased when osc is evicted by OST
2299 Details : when the import is evicted by server, it will fork another
2300 thread ptlrpc_invalidate_import_thread to invalidate the
2301 import, where the grant will be set to 0. While the original
2302 thread will update the grant it got when connecting. So if
2303 the former happened latter, the grant will be wrongly errased
2304 because of this race.
2309 Description: Checking Stale with correct fid
2310 Details : ll_revalidate_it should uses de_inode instead of op_data.fid2
2311 to check whether it is stale, because sometimes, we want the
2312 enqueue happened anyway, and op_data.fid2 will not be initialized.
2314 Severity : enhancement
2316 Description: update patchless client
2317 Details : Add support for patchless client with 2.6.20, 2.6.21 and RHEL 5
2320 Frequency : only with 2.4 kernel
2322 Description: random memory corruption
2323 Details : size of struct ll_inode_info is to big for union inode.u and this
2324 can be cause of random memory corruption.
2329 Description: Memory leak in recovery
2330 Details : Lov_mds_md was not free in an error handler in mds_create_object.
2331 It should also check obd_fail before fsfilt_start, otherwise if
2332 fsfilt_start return -EROFS,(failover mds during mds recovery).
2333 then the req will return with repmsg->transno = 0 and rc = EROFS.
2334 and we met hit the assert LASSERT(req->rq_reqmsg->transno ==
2335 req->rq_repmsg->transno) in ptlrpc_replay_interpret. Fcc should
2336 be freed no matter whether fsfilt_commit success or not.
2339 Frequency : only with huge count clients
2341 Description: Prevents from taking the superblock lock in llap_from_page for
2343 Details : using LL_ORIGIN_REMOVEPAGE origin flag instead of LL_ORIGIN_UNKNOW
2344 for llap_from_page call in ll_removepage prevents from taking the
2345 superblock lock for a soon died page.
2350 Description: Not check open intent error before release open handle
2351 Details : in some rare cases, the open intent error is not checked before
2352 release open handle, which may cause
2353 ASSERTION(open_req->rq_transno != 0), because it tries to release
2354 the failed open handle.
2359 Description: Set cat log bitmap only after create log success.
2360 Details : in some rare cases, the cat log bitmap is set too early. and it
2361 should be set only after create log success.
2365 Description: Accessing a block bevice can re-enable I/O when Lustre is
2366 tearing down a device.
2367 Details : dev_clear_rdonly(bdev) must be called in kill_bdev() instead of
2372 Description: service threads may hog cpus when there are a lot of requests
2373 Details : Insert cond_resched to give other threads a chance to use some CPU
2378 Description: the cat log was not initialized in recovery
2379 Details : When mds(mgs) do recovery, the tgt_count might be zero, so the
2380 unlink log on mds will not be initialized until mds post
2381 recovery. And also in mds post recovery, the unlink log will
2382 initialization will be done asynchronausly, so there will be race
2383 between add unlink log and unlink log initialization.
2387 Description: brw_stats were being printed incorrectly
2388 Details : brw_stats were being printed as log2 but all of them were not
2389 recorded as log2. Also remove some code duplication arising from
2390 filter_tally_{read,write}.
2394 Frequency : rare, only in recovery.
2395 Description: ASSERTION(req->rq_type != LI_POISON) failed
2396 Details : imp_lock should be held while iterating over imp_sending_list for
2397 prevent destroy request after get timeout in ptlrpc_queue_wait.
2401 Description: replay-single.sh test 52 fails
2402 Details : A lock's skiplist need to be cleanup when it being unlinked
2403 from its resource list.
2407 Description: Short directio read returns full requested size rather than
2409 Details : Direct I/O operations should return actual amount of bytes
2410 transferred rather than requested size.
2412 Severity : enhancement
2414 Description: metadata RPC reduction (e.g. for rm performance)
2415 Details : decrease the amount of synchronous RPC between clients and servers
2416 by canceling conflicing lock before the operation on the client
2417 and packing thier handles into the main operation RPC to server.
2419 Severity : enhancement
2421 Description: Async OSC create to avoid the blocking unnecessarily.
2422 Details : If a OST has no remain object, system will block on the creating
2423 when need to create a new object on this OST. Now, ways use
2424 pre-created objects when available, instead of blocking on an
2425 empty osc while others are not empty. If we must block, we block
2426 for the shortest possible period of time.
2430 Description: improve handling recoverable errors
2431 Details : if request processig with error which can be recoverable on server
2432 request should be resend, otherwise page released from cache and
2435 Severity : enhancement
2437 Description: refine locking for avoid write wrong info into lov_objid file
2438 Details : fix possible races with add new target and write/update data in
2441 --------------------------------------------------------------------------------
2443 2007-05-03 Cluster File Systems, Inc. <info@clusterfs.com>
2448 Frequency : on some architectures
2450 Description: 1.6 client sometimes fails to mount from a 1.4 MDT
2451 Details : Uninitialized flags sometimes cause configuration commands to
2455 Frequency : patchless clients only
2457 Description: missing __iget() symbol export
2458 Details : The __iget() symbol export is missing. To avoid the need for
2459 this on patchless clients the deathrow inode reaper is turned
2460 off, and we depend on the VM to clean up old inodes. This
2461 dependency was during via the fix for bug 12181.
2463 --------------------------------------------------------------------------------
2465 2007-04-19 Cluster File Systems, Inc. <info@clusterfs.com>
2467 * CONFIGURATION CHANGE. This version of Lustre WILL NOT
2468 INTEROPERATE with older versions automatically. In many cases a
2469 special upgrade step is needed. Please read the
2470 user documentation before upgrading any part of a live system.
2471 * WIRE PROTOCOL CHANGE from previous 1.6 beta versions. This
2472 version will not interoperate with 1.6 betas before beta5 (1.5.95).
2473 * WARNING: Lustre configuration and startup changes are required with
2474 this release. See https://mail.clusterfs.com/wikis/lustre/MountConf
2478 Severity : enhancement
2480 Description: MountConf
2481 Details : Lustre configuration is now managed via mkfs and mount
2482 commands instead of lmc and lconf. New obd types (MGS, MGC)
2483 are added for dynamic configuration management. See
2484 https://mail.clusterfs.com/wikis/lustre/MountConf for
2487 Severity : enhancement
2489 Description: dynamic OST addition
2490 Details : OSTs can now be added to a live filesystem
2492 Severity : enhancement
2494 Description: startup order invariance
2495 Details : MDTs and OSTs can be started in any order. Clients only
2496 require the MDT to complete startup.
2498 Severity : enhancement
2500 Description: parallel, asynchronous orphan cleanup
2501 Details : orphan cleanup is now performed in separate threads for each
2502 OST, allowing parallel non-blocking operation.
2504 Severity : enhancement
2506 Description: optimized stripe assignment
2507 Details : stripe assignments are now made based on ost space available,
2508 ost previous usage, and OSS previous usage, in order to try
2509 to optimize storage space and networking resources.
2511 Severity : enhancement
2513 Description: Permanently set tunables
2514 Details : All writable /proc/fs/lustre tunables can now be permanently
2515 set on a per-server basis, at mkfs time or on a live system.
2517 Severity : enhancement
2519 Description: Lustre message v2
2520 Details : Add lustre message format v2.
2522 Severity : enhancement
2524 Description: client OST exclusion list
2525 Details : Clients can be started with a list of OSTs that should be
2526 declared "inactive" for known non-responsive OSTs.
2530 Description: SPEC SFS validation failure on NFS v2 over lustre.
2531 Details : Changes the blocksize for regular files to be 2x RPC size,
2532 and not depend on stripe size.
2534 Severity : enhancement
2536 Description: Multiple MD RPCs in flight.
2537 Details : Further unserialise some read-only MDS RPCs - learn about intents.
2538 To avoid overly-overloading MDS, introduce a limit on number of
2539 MDS RPCs in flight for a single client and add /proc controls
2540 to adjust this limit.
2542 Severity : enhancement
2544 Description: client read/write statistics
2545 Details : Add client read/write call usage stats for performance
2546 analysis of user processes.
2547 /proc/fs/lustre/llite/*/offset_stats shows non-sequential
2548 file access. extents_stats shows chunk size distribution.
2549 extents_stats_per_process show chunk size distribution per
2552 Severity : enhancement
2554 Description: mds statistics
2555 Details : Add detailed mds operations statistics in
2556 /proc/fs/lustre/mds/*/stats.
2560 Description: Failure of copying files with lustre special EAs.
2561 Details : Client side always return success for setxattr call for lustre
2562 special xattr (currently only "trusted.lov").
2567 Description: Refcount LNET uuids
2568 Details : The global LNET uuid list grew linearly with every startup;
2569 refcount repeated list entries instead of always adding to
2572 Severity : enhancement
2574 Description: Dynamic service threads
2575 Details : Within a small range, start extra service threads
2576 automatically when the request queue builds up.
2579 Frequency : mixed-endian client/server environments
2581 Description: mixed-endian crashes
2582 Details : The new msg_v2 system had some failures in mixed-endian
2585 Severity : enhancement
2587 Description: Easy OST removal
2588 Details : OSTs can be permanently deactivated with e.g. 'lctl
2589 conf_param lustre-OST0001.osc.active=0'
2591 Severity : enhancement
2593 Description: MGS proc entries
2594 Details : Added basic proc entries for the MGS showing what filesystems
2597 Severity : enhancement
2599 Description: provide MGS failover
2600 Details : Added config lock reacquisition after MGS server failover.
2602 Severity : enhancement
2604 Description: add Linux 2.4 support
2605 Details : Added support for RHEL 2.4.21 kernel for 1.6 servers and clients
2609 Description: a large application tries to do I/O to the same resource and dies
2610 in the middle of it.
2611 Details : Check the req->rq_arrival time after the call to
2612 ost_brw_lock_get(), but before we do anything about
2613 processing it & sending the BULK transfer request. This
2614 should help move old stale pending locks off the queue as
2615 quickly as obd_timeout.
2618 Frequency : when an incorrect nid is specified during startup
2620 Description: ptlrpc connect to non-existant node causes kernel crash
2621 Details : LNET can't be re-entered from an event callback, which
2622 happened when we expire a message after the export has been
2623 cleaned up. Instead, hand the zombie cleanup off to another
2626 Severity : enhancement
2628 Description: plain/inodebits lock performance improvement
2629 Details : Grouping plain/inodebits in granted list by their request modes
2630 and bits policy, thus improving the performance of search through
2634 Frequency : only if OST filesystem is corrupted
2636 Description: client incorrectly hits assertion in ptlrpc_replay_req()
2637 Details : for a short time RPCs with bulk IO are in the replay list,
2638 but replay of bulk IOs is unimplemented. If the OST filesystem
2639 is corrupted due to disk cache incoherency and then replay is
2640 started it is possible to trip an assertion. Avoid putting
2641 committed RPCs into the replay list at all to avoid this issue.
2644 Frequency : liblustre (e.g. catamount) on a large cluster with >= 8 OSTs/OSS
2646 Description: System hang on startup
2647 Details : This bug allowed the liblustre (e.g. catamount) client to
2648 return to the app before handling all startup RPCs. This
2649 could leave the node unresponsive to lustre network traffic
2650 and manifested as a server ptllnd timeout.
2652 Severity : enhancement
2654 Description: Add "/proc/sys/lustre/debug_peer_on_timeout"
2655 Details : liblustre envirable: LIBLUSTRE_DEBUG_PEER_ON_TIMEOUT
2656 boolean to control whether to print peer debug info when a
2657 client's RPC times out.
2660 Frequency : only for kernels with patches from Lustre below 1.4.3
2662 Description: Remove old rdonly API
2663 Details : Remove old rdonly API which unused from at least lustre 1.4.3
2666 Frequency : only for devices with external journals
2668 Description: Set external device read-only also
2669 Details : During a commanded failover stop, we set the disk device
2670 read-only while the server shuts down. We now also set any
2671 external journal device read-only at the same time.
2674 Frequency : when upgrading from 1.4 while trying to change parameters
2676 Description: The wrong (new) MDC name was used when setting parameters for
2677 upgraded MDT's. Also allows changing of OSC (and MDC)
2678 parameters if --writeconf is specified at tunefs upgrade time.
2681 Frequency : when setting specific ost indicies
2683 Description: QOS code breaks on skipped indicies
2684 Details : Add checks for missing OST indicies in the QOS code, so OSTs
2685 created with --index need not be sequential.
2687 Severity : enhancement
2689 Description: Add uninit_groups feature to ldiskfs2 to speed up e2fsck
2690 Details : The uninit_groups feature works in conjunction with the kernel
2691 filesystem code (ldiskfs2 only) and e2fsprogs-1.39-cfs6 to speed
2692 up the pass1 processing of e2fsck. This is a read-only feature
2693 in ldiskfs2 only, so older kernels and current ldiskfs cannot
2694 mount filesystems that have had this feature enabled.
2696 Severity : enhancement
2698 Description: Improve multi-block allocation algorithm to avoid fragmentation
2699 Details : The mballoc3 code (ldiskfs2 only) adds new mechanisms to improve
2700 allocation locality and avoid filesystem fragmentation.
2702 ------------------------------------------------------------------------------
2704 2007-04-01 Cluster File Systems, Inc. <info@clusterfs.com>
2706 * Support for kernels:
2707 2.4.21-47.0.1.EL (RHEL 3)
2708 2.6.5-7.283 (SLES 9)
2709 2.6.9-42.0.10.EL (RHEL 4)
2710 2.6.12.6 vanilla (kernel.org)
2711 2.6.16.27-0.9 (SLES 10)
2712 * Recommended e2fsprogs version: 1.39.cfs5
2714 * Note that reiserfs quotas are disabled on SLES 10 in this kernel
2718 Frequency : occasional, depends on client load and configuration
2719 Bugzilla : 12181, 12203
2720 Description: data loss for recently-modified files
2722 Details : In some cases it is possible that recently written or created
2723 files may not be written to disk in a timely manner (this should
2724 normally be within 30s unless client IO load is very high).
2725 The problem appears as zero-length files or files that are a
2726 multiple of 1MB in size after a client crash or client eviction
2727 that are missing data at the end of the file.
2729 This problem is more likely to be hit on clients where files are
2730 repeatedly created and unlinked in the same directory, clients
2731 have a large amount of RAM, have many CPUs, the filesystem has
2732 many OSTs, the clients are rebooted frequently, and/or the files
2733 are not accessed by other nodes after being written.
2735 The presence of the problem can be detected by looking at
2736 /proc/sys/fs/inode-state. If the first number (nr_inodes) is
2737 smaller than the second (nr_unused) then dirty files will not
2738 be flushed automatically to disk. "sync; sleep 10" should be
2739 run several times on the node before unmounting it to update
2740 Lustre (this is also safe to run on nodes without this problem).
2742 There is also a related kernel bug in the RHEL4 4 2.6.9 kernel
2743 that can cause this same problem, so customers using that kernel
2744 also need to update the kernel in addition to Lustre. In order
2745 to properly fix this bug, the RHEL3 2.4.21 kernel is also updated.
2747 It is normal that files written just before a client crash (less
2748 than 30s) may not yet have been flushed to disk, even for local
2752 Frequency : frequent on thin XT3 nodes
2754 Description: UUID collision on thin XT3 Linux nodes
2755 Details : UUIDs on Compute Node Linux XT3 nodes were not generated
2756 randomly, since we relied on an insufficiently-seeded PRNG.
2761 Description: OSS hangs after "All ost request buffers busy"
2762 Details : A deadlock between quota and journal operations caused OSS
2763 hangs after printing "All ost request buffers busy."
2766 Frequency : always on liblustre builds
2768 Description: Cleanup compiler warnings on liblustre
2771 Frequency : always on liblustre builds on XT3
2773 Description: LC_CONFIG_CDEBUG don't run while build liblustre on XT3.
2777 Description: Addition of EXT3_FEATURE_RO_COMPAT_DIR_NLINKS flag for
2778 > 32000 subdirectories
2779 Details : Add EXT3_FEATURE_RO_COMPAT_DIR_NLINK flag to
2780 EXT3_FEATURE_RO_COMPAT_SUPP. This flag will be set whenever
2781 subdirectory count crosses 32000. This will aid e2fsck to
2782 correctly handle more than 32000 subdirectories.
2785 Frequency : liblustre (e.g. catamount) on a large cluster with >= 8 OSTs/OSS
2787 Description: System hang on startup
2788 Details : This bug allowed the liblustre (e.g. catamount) client to
2789 return to the app before handling all startup RPCs. This
2790 could leave the node unresponsive to lustre network traffic
2791 and manifested as a server ptllnd timeout.
2793 Severity : enhancement
2795 Description: Add "/proc/sys/lustre/debug_peer_on_timeout"
2796 (liblustre envirable: LIBLUSTRE_DEBUG_PEER_ON_TIMEOUT)
2797 boolean to control whether to print peer debug info when a
2798 client's RPC times out.
2803 Description: make O_SYNC working on 2.6 kernels
2804 Details : 2.6 kernels use different method for mark pages for write,
2805 so need add a code to lustre for O_SYNC work.
2810 Description: Failure to close file and release space on NFS
2811 Details : Put inode details into lock acquired in ll_intent_file_open.
2812 Use mdc_intent_lock in ll_intent_open to properly
2813 detect all kind of errors unhandled by mdc_enqueue.
2818 Description: proc file read during shutdown sometimes raced obd removal,
2820 Details : Add lock to prevent obd access after proc file removal.
2823 Frequency : Only for files larger than 4GB on 32-bit clients.
2825 Description: improperly doing page alignment of locks
2826 Details : Modify lustre core code to use CFS_PAGE_* defines instead of
2827 PAGE_*. Make CFS_PAGE_MASK a 64-bit mask.
2832 Description: RPCs being resent when they shouldn't be
2833 Details : Some RPCs that should not be resent are being resent. This
2834 can cause inconsistencies in the RPC state machine. Do not
2835 resend such requests.
2838 Frequency : rare, only with NFS export
2840 Description: Crash on NFS re-export node
2841 Details : under very unusual load conditions an assertion is hit in
2842 ll_intent_file_open()
2845 Frequency : only if OST filesystem is corrupted
2847 Description: client incorrectly hits assertion in ptlrpc_replay_req()
2848 Details : for a short time RPCs with bulk IO are in the replay list,
2849 but replay of bulk IOs is unimplemented. If the OST filesystem
2850 is corrupted due to disk cache incoherency and then replay is
2851 started it is possible to trip an assertion. Avoid putting
2852 committed RPCs into the replay list at all to avoid this issue.
2857 Description: large O_DIRECT requests fail under memory pressure/fragmentation
2858 Details : Large single O_DIRECT read and write calls can fail to allocate
2859 a sufficiently large buffer to process the request. In case of
2860 allocation failure the allocation is retried with a smaller
2861 buffer and broken into smaller requests.
2863 Severity : enhancement
2865 Description: Add -o localflock option to simulate old noflock behaviour.
2866 Details : This will achieve local-only flock/fcntl locks coherentness.
2871 Description: versioning check is incomplete
2872 Details : Checking the version difference of client vs. server, report
2873 error if the gap is too big.
2878 Description: add support PG_writeback bit
2879 Details : add support for PG_writeback bit for Lustre, for more carefull
2880 work with page cache in 2.6 kernel. This also fix some deadlocks
2881 and remove hack for work O_SYNC with 2.6 kernel.
2883 Severity : enhancement
2885 Description: Add uninit_groups feature to ldiskfs2 to speed up e2fsck
2886 Details : The uninit_groups feature works in conjunction with the kernel
2887 filesystem code (ldiskfs2 only) and e2fsprogs-1.39-cfs6 to speed
2888 up the pass1 processing of e2fsck. This is a read-only feature
2889 in ldiskfs2 only, so older kernels and current ldiskfs cannot
2890 mount filesystems that have had this feature enabled.
2892 Severity : enhancement
2894 Description: Improve multi-block allocation algorithm to avoid fragmentation
2895 Details : The mballoc3 code (ldiskfs2 only) adds new mechanisms to improve
2896 allocation locality and avoid filesystem fragmentation.
2898 ------------------------------------------------------------------------------
2900 2007-02-09 Cluster File Systems, Inc. <info@clusterfs.com>
2902 * Support for kernels:
2903 2.6.9-42.0.3.EL (RHEL 4)
2904 2.6.5-7.276 (SLES 9)
2905 2.4.21-47.0.1.EL (RHEL 3)
2906 2.6.12.6 vanilla (kernel.org)
2907 2.6.16.21-0.8 (SLES10)
2908 * Recommended e2fsprogs version: 1.39.cfs2-0
2910 * The backwards-compatible /proc/sys/portals symlink has been removed
2911 in this release. Before upgrading, please ensure that you change
2912 any configuration scripts or /etc/sysctl.conf files that access
2913 /proc/sys/portals/* or sysctl portals.* to use the corresponding
2914 entry in /proc/sys/lnet or sysctl lnet.*. This change can be made
2915 in advance of the upgrade on any system running Lustre 1.4.6 or
2916 newer, since /proc/sys/lnet was added in that version.
2917 * Note that reiserfs quotas are disabled on SLES 10 in this kernel
2921 Frequency : only when quota is used
2923 Description: avoid scanning export list for quota master
2924 Details : Change the algorithms to avoid scanning export list in order
2925 to improve the efficiency.
2928 Frequency : MDS failover only, very rarely
2930 Description: "went back in time" messages on mds failover
2931 Details : The greatest transno may be lost when the current operation
2932 finishes with an error (transno==0) and the client's last_rcvd
2933 record is over-written. Save the greatest transno in the
2934 mds_last_transno for this case.
2937 Frequency : always for specific kernels and striping counts
2939 Description: client may get "Matching packet too big" without ACL support
2940 Details : Clients compiled without CONFIG_FS_POSIX_ACL get an error message
2941 when trying to access files in certain configurations. The
2942 clients should in fact be denied when mounting because they do
2943 not understand ACLs.
2946 Frequency : Cray XT3 with more than 4000 clients and multiple jobs
2948 Description: many clients connecting with IO in progress causes connect timeouts
2949 Details : Avoid synchronous journal commits to avoid delays caused by many
2950 clients connecting/disconnecting when bulk IO is in progress.
2951 Queue liblustre connect requests on OST_REQUEST_PORTAL instead of
2952 OST_IO_PORTAL to avoid delays behind potentially many pending
2956 Frequency : occasionally with multiple writers to a single file
2958 Description: shared writes to file may result in wrong size reported by stat()
2959 Details : Allow growing of kms when extent lock is cancelled
2962 Frequency : always with random mmap IO to multi-striped file
2964 Description: mmap write might be lost if we are writing to a 'hole' in stripe
2965 Details : Only if the hole is at the end of OST object so that kms is too
2966 small. Fix is to increase kms accordingly in ll_nopage.
2969 Frequency : rare, only if OST filesystem is inconsistent with MDS filesystem
2971 Description: writes to a missing object would leak memory on the OST
2972 Details : If there is an inconsistency between the MDS and OST filesystems,
2973 such that the MDS references an object that doesn't exist, writes
2974 to that object will leak memory due to incorrect cleanup in the
2975 error handling path, eventually running out of memory on the OST.
2980 Description: Creating too long symlink causes lustre errors
2981 Details : Check symlink and name lengths before sending requests to MDS.
2984 Frequency : only if flock is enabled (not on by default)
2986 Description: posix locks not released on fd closure on 2.6.9+
2987 Details : We failed to add posix locks to list of inode locks on 2.6.9+
2988 kernels, this caused such locks not to be released on fd close and
2989 then assertions on fs unmount about still used locks.
2992 Frequency : MDS failover only, very rarely
2994 Description: clients may get ASSERTION(granted_lock != NULL)
2995 Details : When request was taking a long time, and a client was resending
2996 a getattr by name lock request. The were multiple lock requests
2997 with the same client lock handle and
2998 mds_getattr_name->fixup_handle_for_resent_request found one of the
2999 lock handles but later failed with ASSERTION(granted_lock != NULL).
3004 Description: handle->h_buffer_credits > 0, assertion failure
3005 Details : h_buffer_credits is zero after truncate, causing assertion
3006 failure. This patch extends the transaction or creates a new
3010 Frequency : NFS re-export or patchless client
3011 Bugzilla : 11179, 10796
3012 Description: Crash on NFS re-export node (__d_move)
3013 Details : We do not want to hash the dentry if we don't have a lock.
3014 But if this dentry is later used in d_move, we'd hit uninitialised
3015 list head d_hash, so we just do this to init d_hash field but
3016 leave dentry unhashed.
3019 Frequency : NFS re-export or patchless client
3021 Description: NFS exports has problem with symbolic link
3022 Details : lustre client didn't properly install dentry when re-exported
3023 to NFS or running patchless client.
3026 Frequency : NFS re-export or patchless client
3028 Description: Various nfs/patchless fixes.
3029 Details : fixes reuse disconected alias for lookup process - this fixes
3030 warning "find_exported_dentry: npd != pd",
3031 fix permission error with open files at nfs.
3032 fix apply umask when do revalidate.
3035 Frequency : occasional
3037 Description: Crash on NFS re-export node
3038 Details : calling clear_page() on the wrong pointer triggered oops in
3039 generic_mapping_read().
3042 Frequency : rarely, using O_DIRECT IO
3044 Description: unaligned directio crashes client with LASSERT
3045 Details : check for unaligned buffers before trying any requests.
3048 Frequency : rarely, using CFS RAID5 patches in non-standard kernel series
3050 Description: stale data returned from RAID cache
3051 Details : If only a small amount of IO is done to the RAID device before
3052 reading it again it is possible to get stale data from the RAID
3053 cache instead of reading it from disk.
3056 Frequency : always for sles10 kernel
3058 Description: sles10 support
3059 Details : ll_follow_link: compile fixes and using of nd_set_link
3060 under newer kernels.
3063 Frequency : depends on arch, kernel and compiler version, always on sles10
3066 Description: recursive or deep enough symlinks cause stack overflow
3067 Details : getting rid of large stack-allocated variable in
3071 Frequency : depends on hardware
3073 Description: lustre write performance loss in the SLES10 kernel
3074 Details : the performance loss is caused by using of write barriers in the
3075 ext3 code. The SLES10 kernel turns barrier support on by
3076 default. The fix is to undo that change for ldiskfs.
3078 ------------------------------------------------------------------------------
3080 2006-12-09 Cluster File Systems, Inc. <info@clusterfs.com>
3082 * Support for kernels:
3083 2.6.9-42.0.3EL (RHEL 4)
3084 2.6.5-7.276 (SLES 9)
3085 2.4.21-47.0.1.EL (RHEL 3)
3086 2.6.12.6 vanilla (kernel.org)
3090 Frequency : quota enabled and large files being deleted
3092 Description: releasing more than 4GB of quota at once hangs OST
3093 Details : If a user deletes more than 4GB of files on a single OST it
3094 will cause the OST to spin in an infinite loop. Release
3095 quota in < 4GB chunks, or use a 64-bit value for 1.4.7.1+.
3100 Description: statfs data retrieved from /proc may be stale or zero
3101 Details : When reading per-device statfs data from /proc, in the
3102 {kbytes,files}_{total,free,avail} files, it may appear
3103 as zero or be out of date.
3106 Frequency : systems with MD RAID1 external journal devices
3108 Description: lconf's call to blkid is confused by RAID1 journal devices
3109 Details : Use the "blkid -l" flag to locate the MD RAID device instead
3110 of returning all block devices that match the journal UUID.
3113 Frequency : always, for aggregate stripe size over 4GB
3115 Description: "lfs setstripe" fails assertion when setting 4GB+ stripe width
3116 Details : Using "lfs setstripe" to set stripe size * stripe count over 4GB
3117 will fail the kernel with "ASSERTION(lsm->lsm_xfersize != 0)"
3120 Frequency : always if "lfs find" used on a local file/directory
3122 Description: "lfs find" segfaults if used on a local file/directory
3123 Details : The case where a directory component was not specified wasn't
3124 handled correctly. Handle this properly.
3127 Frequency : always on ppc64
3129 Description: the write to an ext3 filesystem mounted with mballoc got stuck
3130 Details : ext3_mb_generate_buddy() uses find_next_bit() which does not
3131 perform endianness conversion.
3134 Frequency : rarely (truncate to non-zero file size after write under load)
3135 Bugzilla : 10730, 10687
3136 Description: Files padded with zeros to next 4K multiple
3137 Details : With filesystems mounted using the "extents" option (2.6 kernels)
3138 it is possible that files that are truncated to a non-zero size
3139 immediately after being written are filled with zero bytes beyond
3140 the truncated size. No file data is lost.
3142 Severity : enhancement
3144 Description: Allow recovery/failover for liblustre clients.
3145 Details : liblustre clients were unaware of failover configurations until
3148 Severity : enhancement
3150 Description: user file locks should fail when not mounting with flock option
3151 Details : Set up an error-returning stub in ll_file_operations.lock field
3152 to prevent incorrect behaviour when client is mounted without
3153 flock option. Also, set up properly f_op->flock field for
3157 Frequency : always on ia64
3159 Description: "lfs df" loops on printing out MDS statfs information
3160 Details : The obd_ioctl_data was not initialized and in some systems
3161 this caused a failure during the ioctl that did not return
3162 an error. Initialize the struct and return an error on failure.
3165 Frequency : SLES 9 only
3167 Description: Error of copying files with lustre special EAs as root
3168 Details : Client side always return success for setxattr call for lustre
3169 special xattr (currently only "trusted.lov").
3172 Frequency : rarely on clusters with both ia64+i386 clients
3174 Description: ia64+i686 clients doing shared IO on the same file may LBUG
3175 Details : In rare cases when both ia64+i686 (or other mixed-PAGE_SIZE)
3176 clients are doing concurrent writes to the same file it is
3177 possible that the ia64 clients may LASSERT because the OST
3178 extent locks are not PAGE_SIZE aligned. Ensure that grown
3179 locks are always aligned on the request boundary.
3182 Frequency : specific use, occasional
3184 Description: Overwriting in use executable truncates on-disk binary image
3185 Details : If one node attempts to overwrite an executable in use by
3186 another node, we now correctly return ETXTBSY instead of
3187 truncating the file.
3189 Severity : enhancement
3191 Description: Async OSC create to avoid the blocking unnecessarily.
3192 Details : If a OST has no remain object, system will block on the creating
3193 when need to create a new object on this OST. Now, ways use
3194 pre-created objects when available, instead of blocking on an
3195 empty osc while others are not empty. If we must block, we block
3196 for the shortest possible period of time.
3201 Description: chmod on Lustre root is propagated to other clients
3202 Details : Re-validate root's dentry in ll_lookup_it to avoid having it
3203 invalid by the follow_mount time.
3206 Frequency : liblustre clients only
3208 Description: Race in 'instant cancel' lock handling could lead to such locks
3209 never to be granted in case of SMP MDS
3210 Details : Do not destroy not yet granted but cbpending locks in
3214 Frequency : replay/resend of open
3216 Description: non null lock assetion failure in mds_intent_policy
3217 Details : Trying to replay/resend lockless open requests resulted in
3218 mds_open() returning 0 with no lock. Now it sets a flag if
3219 a lock is going to be returned.
3221 Severity : enhancement
3223 Description: Checksum enhancements
3224 Details : New checksum enhancements allow for resending RPCs that failed
3227 Severity : enhancement
3229 Description: Tunables on number of dirty pages in cacche
3230 Details : Allow to set limit on number of dirty pages cached.
3235 Description: client crash on unmount - lock still has references
3236 Details : In some error handling cases it was possible to leak a lock
3237 reference on a client while accessing a file. This was not
3238 harmful to the client during operation, but would cause the
3239 client to crash when the filesystem is unmounted.
3242 Frequency : specific case, rare
3244 Description: ETXTBSY on mds though file not in use by client
3245 Details : ETXTBSY is no longer incorrectly returned when attempting to
3246 chmod or chown a directory that the user previously tried to
3247 execute or a currently-executing binary.
3250 Frequency : extremely rare except on liblustre-based clients
3252 Description: Lustre space not freed when files are deleted
3253 Details : Clean up open-unlinked files after client eviction. Previously
3254 the unlink was skipped and the files remained as orphans.
3259 Description: OST failure "would be an LBUG" in waiting_locks_callback()
3260 Details : In some cases it was possible to send a blocking callback to
3261 a client doing a glimpse, even though that client didn't get
3262 a lock granted. When the glimpse lock is cancelled on the OST
3263 the freed lock is left on the waiting list and corrupted the list.
3266 Frequency : all core dumps
3268 Description: Broke core dumps to lustre
3269 Details : Negative dentry may be unhashed if parent does not have UPDATE
3270 lock, but some callers, e.g. do_coredump, expect dentry to be
3271 hashed after successful create, hash it in ll_create_it.
3273 ------------------------------------------------------------------------------
3275 2006-09-13 Cluster File Systems, Inc. <info@clusterfs.com>
3277 * Support for kernels:
3278 2.6.9-42.0.2.EL (RHEL 4)
3279 2.6.5-7.276 (SLES 9)
3280 2.4.21-40.EL (RHEL 3)
3281 2.6.12.6 vanilla (kernel.org)
3285 Frequency : always on RHEL 3
3287 Description: Number of open files grows over time
3288 Details : The number of open files grows over time, whether or not
3289 Lustre is started. This was due to a filp leak introduced
3290 by one of our kernel patches.
3292 ------------------------------------------------------------------------------
3294 08-20-2006 Cluster File Systems, Inc. <info@clusterfs.com>
3296 * Support for kernels:
3297 2.6.9-42.EL (RHEL 4)
3298 2.6.5-7.276 (SLES 9)
3299 2.4.21-40.EL (RHEL 3)
3300 2.6.12.6 vanilla (kernel.org)
3305 Bugzilla : 5719, 9635, 9792, 9684
3306 Description: OST (or MDS) trips assertions in (re)connection under heavy load
3307 Details : If a server is under heavy load and cannot reply to new
3308 connection requests before the client resends the (re)connect,
3309 the connection handling code can behave badly if two service
3310 threads are concurrently handing separate (re)connections from
3311 the same client. Add better locking to the connection handling
3312 code, and ensure that only a single connection will be processed
3313 for a given client UUID, even if the lock is dropped.
3315 Severity : enhancement
3317 Description: add TCP zero-copy support to kernel
3318 Details : Add support to the kernel TCP stack to allow zero-copy bulk
3319 sends if the hardware supports scatter-gather and checksumming.
3320 This allows socklnd to do client-write and server-read more
3321 efficiently and reduce CPU utilization from skbuf copying.
3324 Frequency : only if NFS exporting from client
3326 Description: NULL pointer deref in ll_iocontrol() if chattr mknod file
3327 Details : If setting attributes on a file created under NFS that had
3328 never been opened it would be possible to oops the client
3329 if the file had no objects.
3332 Frequency : always for liblustre
3334 Description: liblustre client does MDS+OSTs setattr RPC for each write
3335 Details : When doing a write from a liblustre client, the client
3336 incorrectly issued an RPC to the MDS and each OST the file was
3337 striped over in order to update the timestamps. When writing
3338 with small chunks and many clients this could overwhelm the MDS
3339 with RPCs. In all cases it would slow down the write because
3340 these RPCs are unnecessary.
3342 Severity : enhancement
3344 Description: allow number of MDS service threads to be changed at module load
3345 Details : It is now possible to change the number of MDS service threads
3346 running. Adding "options mds mds_num_threads={N}" to the MDS's
3347 /etc/modprobe.conf will set the number of threads for the next
3348 time Lustre is restarted (assuming the "mds" module is also
3349 reloaded at that time). The default number of threads will
3350 stay the same, 32 for most systems.
3355 Description: OST crash if filesystem is unformatted or corrupt
3356 Details : If an OST is started on a device that has never been formatted
3357 or if the filesystem is corrupt and cannot even mount then the
3358 error handling cleanup routines would dereference a NULL pointer.
3363 Description: NULL pointer deref in llap_from_page.
3364 Details : get_cache_page_nowait can return a page with NULL (or otherwise
3365 incorrect) mapping if the page was truncated/reclaimed while it was
3366 searched for. Check for this condition and skip such pages when
3367 doing readahead. Introduce extra check to llap_from_page() to
3368 verify page->mapping->host is non-NULL (so page is not anonymous).
3371 Frequency : Sometimes when using sys_sendfile
3373 Description: "page not covered by a lock" warnings from ll_readpage
3374 Details : sendfile called ll_readpage without right page locks present.
3375 Now we introduced ll_file_sendfile that does necessary locking
3376 around call to generic_file_sendfile() much like we do in
3380 Frequency : with certain MDS communication failures at client mount time
3382 Description: NULL pointer deref after failed client mount
3383 Details : a client connection request may delayed by the network layer
3384 and not be sent until after the PTLRPC layer has timed out the
3385 request. If the client fails the mount immediately it will try
3386 to clean up before the network times out the request. Add a
3387 reference from the request import to the obd device and delay
3388 the cleanup until the network drops the request.
3391 Frequency : occasionally during client (re)connect
3393 Description: assertion failure during client (re)connect
3394 Details : processing a client connection request may be delayed by the
3395 client or server longer than the client connect timeout. This
3396 causes the client to resend the connection request. If the
3397 original connection request is replied in this interval, the
3398 client may trip an assertion failure in ptlrpc_connect_interpret()
3399 which thought it would be the only running connect process.
3402 Frequency : only with obd_echo servers and clients that are rebooted
3404 Description: kernel BUG accessing uninitialized data structure
3405 Details : When running an obd_echo server it did not start the ping_evictor
3406 thread, and when a client was evicted an uninitialized data
3407 structure was accessed. Start the ping_evictor in the RPC
3408 service startup instead of the OBD startup.
3410 Severity : enhancement
3411 Bugzilla : 10193 (patchless)
3412 Description: Remove dependency on various unexported kernel interfaces.
3413 Details : No longer need reparent_to_init, exit_mm, exit_files,
3414 sock_getsockopt, filemap_populate, FMODE_EXEC, put_filp.
3417 Frequency : rare (only users of deprecated and unsupported LDAP config)
3419 Description: write_conf for zeroconf mount queried LDAP incorrectly for client
3420 Details : LDAP apparently contains 'lustreName' attributes instead of
3421 'name'. A simple remapping of the name is sufficient.
3424 Frequency : rare (only with non-default dump_on_timeout debug enabled)
3426 Description: waiting_locks_callback trips kernel BUG if client is evicted
3427 Details : Running with the dump_on_timeout debug flag turned on makes
3428 it possible that the waiting_locks_callback() can try to dump
3429 the Lustre kernel debug logs from an interrupt handler. Defer
3430 this log dumping to the expired_lock_main() thread.
3432 Severity : enhancement
3434 Description: Support NFS exporting on 2.6 kernels.
3435 Details : Implement non-rawops metadata methods for NFS server to use without
3436 changing NFS server code.
3439 Frequency : very rare (synthetic metadata workload only)
3441 Description: two racing renames might cause an MDS thread to deadlock
3442 Details : Running the "racer" program may cause one MDS thread to rename
3443 a file from being the source of a rename to being the target of
3444 a rename at exactly the same time that another thread is doing
3445 so, and the second thread has already enqueued these locks after
3446 doing a lookup of the target and is trying to relock them in
3447 order. Ensure that we don't try to re-lock the same resource.
3450 Frequency : only very large systems with liblustre clients
3452 Description: slow eviction of liblustre clients with the "evict_by_nid" RPC
3453 Details : Use asynchronous set_info RPCs to send the "evict_by_nid" to
3454 all OSTs in parallel. This allows the eviction of stale liblustre
3455 clients to proceed much faster than if they were done in series,
3456 and also offers similar improvements for other set_info RPCs.
3461 Description: excessive CPU usage during initial read phase on client
3462 Details : During the initial read phase on a client, it would agressively
3463 retry readahead on the file, consuming too much CPU and impacting
3464 performance (since 1.4.5.8). Improve the readahead algorithm
3465 to avoid this, and also improve some other common cases (read
3466 of small files in particular, where "small" is files smaller than
3467 /proc/fs/lustre/llite/*/max_read_ahead_whole_mb, 2MB by default).
3472 Description: MDS crash when receiving packet with unknown intent.
3473 Details : Do not LBUG in unknown intent case, just return -EFAULT
3475 Severity : enhancement
3476 Bugzilla : 9293, 9385
3477 Description: MDS RPCs are serialised on client. This is unnecessary for some.
3478 Details : Do not serialize getattr (non-intent version) and statfs.
3481 Frequency : occasional, when OST network is overloaded/intermittent
3483 Description: client evicted by OST after bulk IO timeout
3484 Details : If a client sends a bulk IO request (read or write) the OST
3485 may evict the client if it is unresposive to its data GET/PUT
3486 request. This is incorrect if the network is overloaded (takes
3487 too long to transfer the RPC data) or dropped the OST GET/PUT
3488 request. There is no need to evict the client at all, since
3489 the pinger and/or lock callbacks will handle this, and the
3490 client can restart the bulk request.
3493 Frequency : Always when mmapping file with no objects
3495 Description: client crashes when mmapping file with no objects
3496 Details : Check that we actually have objects in a file before doing any
3497 operations on objects in ll_vm_open, ll_vm_close and
3503 Description: Request leak when working with deleted CWD
3504 Details : Introduce advanced request refcount tracking for requests
3505 referenced from lustre intent.
3507 Severity : Enhancement
3509 Description: Cache open file handles on client.
3510 Details : MDS now will return special lock along with openhandle, if
3511 requested and client is allowed to hold openhandle, even if unused,
3512 until such a lock is revoked. Helps NFS a lot, since NFS is opening
3513 closing files for every read/write openration.
3515 Severity : Enhancement
3517 Description: Cache open negative dentries on client when possible.
3518 Details : Guard negative dentries with UPDATE lock on parent dir, drop
3519 negative dentries on lock revocation.
3524 Description: Remounting a client read-only wasn't possible with a zconf mount
3525 Details : It wasn't possible to remount a client read-only with llmount.
3527 Severity : enhancement
3528 Description: Include MPICH 1.2.6 Lustre ADIO interface patch
3529 Details : In lustre/contrib/ or /usr/share/lustre in RPM a patch for
3530 MPICH is included to add Lustre-specific ADIO interfaces.
3531 This is based closely on the UFS ADIO layer and only differs
3532 in file creation, in order to allow the OST striping to be set.
3533 This is user-contributed code and not supported by CFS.
3538 Description: extended inode attributes (immutable, append-only) work improperly
3539 when 2.4 and 2.6 kernels are used on client/server or vice versa
3540 Details : Introduce kernel-independent values for these flags.
3542 Severity : enhancement
3545 Description: Allow fractional MB tunings for lustre in /proc/ filesystem.
3546 Details : Many of the /proc/ tunables can only be tuned at a megabyte
3547 granularity. Now, Fractional MB granularity is be supported,
3548 this is very useful for low memory system.
3550 Severity : enhancement
3552 Description: Getattr by fid
3553 Details : Getting a file attributes by its fid, obtaining UPDATE|LOOKUP
3554 locks, avoids extra getattr rpc requests to MDS, allows '/' to
3555 have locks and avoids getattr rpc requests for it on every stat.
3558 Frequency : Always, for filesystems larger than 2TB
3560 Description: ldiskfs crash at mount for filesystem larger than 2TB with mballoc
3561 Details : Kenrel kmalloc limits allocations to 128kB and this prevents
3562 filesystems larger than 2TB to be mounted with mballoc enabled.
3565 Frequency : Always, for 32-bit kernel without CONFIG_LBD and filesystem > 2TB
3567 Description: ldiskfs crash at mount for filesystem larger than 2TB with mballoc
3568 Details : If a 32-bit kernel is compiled without CONFIG_LBD enabled and a
3569 filesystems larger than 2TB is mounted then the kernel will
3570 silently corrupt the start of the filesystem. CONFIG_LBD is
3571 enabled for all CFS-supported kernels, but the possibility of
3572 this happening with a modified kernel config exists.
3574 Severity : enhancement
3576 Description: add client O_DIRECT support for 2.6 kernels
3577 Details : It is now possible to do O_DIRECT reads and writes to files
3578 in the Lustre client mountpoint on 2.6 kernel clients.
3580 Severity : enhancement
3582 Description: parallel glimpse, setattr, statfs, punch, destroy requests
3583 Details : Sends glimpse, setattr, statfs, punch, destroy requests to OSTs in
3584 parallel, not waiting for response from every OST before sending
3585 a rpc to the next OST.
3590 Description: setattr vs write race when updating file timestamps
3591 Details : Client processes that update a file timestamp into the past
3592 right after writing to the file (e.g. tar) it is possible that
3593 the updated file modification time can be reset to the current
3594 time due to a race between processing the setattr and write RPC.
3596 Severity : enhancement
3598 Description: Bring 'lfs find' closer in line with regular Linux find.
3599 Details : lfs find util supports -atime, -mtime, -ctime, -maxdepth, -print,
3600 -print0 options and obtains all the needed info through the lustre
3603 Severity : enhancement
3605 Description: support up to 1024 configured devices on one node
3606 Details : change obd_dev array from statically allocated to dynamically
3607 allocated structs as they are first used to reduce memory usage
3612 Description: Flush dirty partially truncated pages during truncate
3613 Details : Immediatelly flush partially truncated pages in filter_setattr,
3614 this way we completely avoid having any pages in page cache on OST
3615 and can retire ugly workarounds during writes to flush such pages.
3620 Description: i_sem vs transaction deadlock in mds_obd_destroy during unlink.
3621 Details : protect inode from truncation within vfs_unlink() context
3622 just take a reference before calling vfs_unlink() and release it
3623 when parent's i_sem is free.
3628 Description: last_id value checked outside lock on OST caused LASSERT failure
3629 Details : If there were multiple MDS->OST object precreate requests in
3630 flight, it was possible that the OST's last object id was checked
3631 outside a lock and incorrectly tripped an assertion. Move checks
3632 inside locks, and discard old precreate requests.
3635 Frequency : always, if extents are used on OSTs
3637 Description: index ei_leaf_hi (48-bit extension) is not zeroed in extent index
3638 Details : OSTs using the extents format would not zero the high 16 bits of
3639 the index physical block number. This is not a problem for any
3640 OST filesystems smaller than 16TB, and no kernels support ext3
3641 filesystems larger than 16TB yet. This is fixed in 1.4.7 (all
3642 new/modified files) and can be fixed for existing filesystems
3643 with e2fsprogs-1.39-cfs1.
3648 Description: import connection selection may be incorrect if timer wraps
3649 Details : Using a 32-bit jiffies timer with HZ=1000 may cause backup
3650 import connections to be ignored if the 32-bit jiffies counter
3651 wraps. Use a 64-bit jiffies counter.
3654 Frequency : very large clusters immediately after boot
3656 Description: LNET request buffers exhausted under heavy short-term load
3657 Details : If a large number of client requests are generated on a service
3658 that has previously never seen so many requests it is possible
3659 that the request buffer growth cannot keep up with the spike in
3660 demand. Instead of dropping incoming requests, they are held in
3661 the LND until the RPC service can accept more requests.
3664 Frequency : Sometimes during replay
3666 Description: Assertion failure in ll_local_open after replay.
3667 Details : If replay happened on an open request reply before we were able
3668 to set replay handler, reply will become not swabbed tripping the
3669 assertion in ll_local_open. Now we set the handler right after
3670 recognising of open request
3673 Frequency : very rare
3675 Description: kernel reports "badness in vsnprintf"
3676 Details : Reading from the "recovery_status" /proc file in small chunks
3677 may cause a negative length in lprocfs_obd_rd_recovery_status()
3678 call to vsnprintf() (which is otherwise harmless). Exit early
3679 if there is no more space in the output buffer.
3681 Severity : enhancement
3683 Description: clear OBD RPC statistics by writing to them
3684 Details : It is now possible to clear the OBD RPC statistics by writing
3685 to the "stats" file.
3690 Description: Client mtime is not the same on different clients after utimes
3691 Details : In some cases, the client was using the utimes() syscall on
3692 a file cached on another node. The clients now validate the
3693 ctime from the MDS + OSTs to determine which one is right.
3698 Description: Inability to activate failout mode
3699 Details : lconf script incorrectly assumed that in pythong string's numeric
3700 value is used in comparisons.
3703 Frequency : always with multiple stripes per file
3705 Description: Inefficient object allocation for mutli-stripe files
3706 Details : When selecting which OSTs to stripe files over, for files with
3707 a stripe count that divides evenly into the number of OSTs,
3708 the MDS is always picking the same starting OST for each file.
3709 Return the OST selection heuristic to the original design.
3714 Description: mount failures may take full timeout to return an error
3715 Details : Under some heavy load conditions it is possible that a
3716 failed mount can wait for the full obd_timeout interval,
3717 possibly several minutes, before reporting an error.
3718 Instead return an error as soon as the status is known.
3720 Frequency : quota enabled and large files being deleted
3722 Description: releasing more than 4GB of quota at once hangs OST
3723 Details : If a user deletes more than 4GB of files on a single OST it
3724 will cause the OST to spin in an infinite loop. Release
3725 quota in < 4GB chunks, or use a 64-bit value for 1.4.7.1+.
3730 Description: statfs data retrieved from /proc may be stale or zero
3731 Details : When reading per-device statfs data from /proc, in the
3732 {kbytes,files}_{total,free,avail} files, it may appear
3733 as zero or be out of date.
3736 Frequency : systems with MD RAID1 external journal devices
3738 Description: lconf's call to blkid is confused by RAID1 journal devices
3739 Details : Use the "blkid -l" flag to locate the MD RAID device instead
3740 of returning all block devices that match the journal UUID.
3743 Frequency : always, for aggregate stripe size over 4GB
3745 Description: assertion fails when trying to use 4GB stripe size
3746 Details : Use "setstripe" to set stripe size over 4GB will fail the kernel,
3747 complaining "ASSERTION(lsm->lsm_xfersize != 0)"
3750 Frequency : always on ppc64
3752 Description: the first write on an ext3 filesystem with mballoc got stuck
3753 Details : ext3_mb_generate_buddy() uses find_next_bit() which does not
3754 perform endianness conversion.
3756 ------------------------------------------------------------------------------
3758 02-14-2006 Cluster File Systems, Inc. <info@clusterfs.com>
3760 * WIRE PROTOCOL CHANGE. This version of Lustre networking WILL NOT
3761 INTEROPERATE with older versions automatically. Please read the
3762 user documentation before upgrading any part of a live system.
3763 * WARNING: Lustre networking configuration changes are required with
3764 this release. See https://bugzilla.clusterfs.com/show_bug.cgi?id=10052
3767 * Support for kernels:
3768 2.6.9-22.0.2.EL (RHEL 4)
3769 2.6.5-7.244 (SLES 9)
3770 2.6.12.6 vanilla (kernel.org)
3773 Severity : enhancement
3774 Bugzilla : 7981/8208
3775 Description: Introduced Lustre Networking (LNET)
3776 Details : LNET is new networking infrastructure for Lustre, it includes
3777 a reorganized network configuration mode (see the user
3778 documentation for full details) as well as support for routing
3779 between different network fabrics. Lustre Networking Devices
3780 (LNDs) for the supported network fabrics have also been
3781 created for this new infrastructure.
3783 Severity : enhancement
3784 Description: Introduced Access control lists
3785 Details : clients can set ACLs on files and directories in order to have
3786 more fine-grained permissions than the standard Unix UGO+RWX.
3787 The MDS must be started with the "-o acl" mount option.
3789 Severity : enhancement
3790 Description: Introduced filesystem quotas
3791 Details : Administrators may now establish per-user quotas on the
3794 Severity : enhancement
3796 Description: Configuration change for the XT3
3797 The PTLLND is now used to run Lustre over Portals on the XT3
3798 The configure option(s) --with-cray-portals are no longer used.
3799 Rather --with-portals=<path-to-portals-includes> is used to
3800 enable building on the XT3. In addition to enable XT3 specific
3801 features the option --enable-cray-xt3 must be used.
3806 Description: Running on many-way SMP OSTs can trigger oops in llcd_send()
3807 Details : A race between allocating a new llcd and re-getting the llcd_lock
3808 allowed another thread to grab newly-allocated llcd.
3810 Severity : enhancement
3812 Description: 2.6 OST async journal commit and locking fix to improve performance
3813 Details : The filter_direct_io()+filter_commitrw_write() journal commits for
3814 2.6 kernels are now async as they already were in 2.4 kernels so
3815 that they can commit concurrently with the network bulk transfer.
3816 For block-allocated files the filter allocation semaphore is held
3817 to avoid filesystem fragmentation during allocation. BKL lock
3818 removed for 2.6 xattr operations where it is no longer needed.
3823 Description: lconf incorrectly determined whether two IP networks could talk
3824 Details : In some more complicated routing and multiple-network
3825 configurations, lconf will avoid trying to make a network
3826 connection to a disjoint part of the IP space. It was doing the
3827 math incorrectly for one set of cases.
3832 Description: Fix for potential infinite loop processing records in an llog.
3833 Details : If an llog record is corrupted/zeroed, it is possible to loop
3834 forever in llog_process(). Validate the llog record length
3835 and skip the remainder of the block on error.
3838 Frequency : occasional (liblustre only)
3840 Description: liblustre could not open files whose last component is a symlink
3841 Details : sysio_path_walk() would incorrectly pass the open intent to
3842 intermediate path components.
3845 Frequency : rare (liblustre only with non-standard tuning)
3846 Bugzilla : 7201 (7350)
3847 Description: Tuning the MDC DLM LRU size to zero triggers client LASSERT
3848 Details : llu_lookup_finish_locks() tries to set lock data on a lock
3849 after it has been released, only do this for referenced locks
3851 Severity : enhancement
3853 Description: specifying an (invalid) directory default stripe_size of -1
3854 would reset the directory default striping
3855 Details : stripe_size -1 was used internally to signal directory stripe
3856 removal, now use "all default" to signal dir stripe removal
3857 as a directory striping of "all default" is not useful
3860 Frequency : common for large clusters running liblustre clients
3862 Description: doing an ls when liblustre clients are running is slow
3863 Details : sending a glimpse AST to a liblustre client waits for every AST
3864 to time out, as liblustre clients will not respond. Since they
3865 cannot cache data we refresh the OST lock LVB from disk instead.
3867 Severity : enhancement
3869 Description: doing an ls at the same time as file IO can be slow
3870 Details : enqueue and other "small" requests can be blocked behind many
3871 large IO requests. Create a new OST IO portal for non-IO
3872 requests so they can be processed faster.
3875 Frequency : rare (only HPUX clients mounting unsupported re-exported NFS vol)
3877 Description: an HPUX NFS client would get -EACCESS when ftruncate()ing a newly
3878 created file with mode 000
3879 Details : the Linux NFS server relies on an MDS_OPEN_OWNEROVERRIDE hack to
3880 allow an ftruncate() as a non-root user to a file with mode 000.
3881 Lustre now respects this flag to disable mode checks when
3882 truncating a file owned by the user
3885 Frequency : liblustre-only, when liblustre client dies unexpectedly or becomes
3888 Description: Revoking locks from clients that went dead or catatonic might take
3890 Details : New lock flags FL_CANCEL_ON_BLOCK used by liblustre makes
3891 cancellation of such locks instant on servers without waiting for
3892 any reply from clients. Clients drops these locks when cancel
3893 notification from server is received without replying.
3896 Frequency : liblustre-only, when liblustre client dies or becomes busy
3898 Description: Doing ls on Linux clients can take a long time with active
3900 Details : Liblustre client cannot handle ASTs in timely manner, so avoid
3901 granting such locks to it in the first place if possible. Locks
3902 are taken by proxy on the OST during the read or write and
3903 dropped immediately afterward. Add connect flags handling, do
3904 not grant locks to liblustre clients for glimpse ASTs.
3906 Severity : enhancement
3908 Description: Improve read-ahead algorithm to avoid excessive IO for random reads
3909 Details : Existing read-ahead algorithm is tuned for the case of streamlined
3910 sequential reads and behaves badly with applications doing random
3911 reads. Improve it by reading ahead at least read region, and
3912 avoiding excessive large RPC for small reads.
3914 Severity : enhancement
3916 Description: Creating more than 1000 files for a single job may cause a load
3917 imbalance on the OSTs if there are also a large number of OSTs.
3918 Details : qos_prep_create() uses an OST index reseed value that is an
3919 even multiple of the number of available OSTs so that if the
3920 reseed happens in the middle of the object allocation it will
3921 still utilize the OSTs as uniformly as possible.
3926 Description: OST or MDS may oops in ping_evictor_main()
3927 Details : ping_evictor_main() drops obd_dev_lock if deleting a stale export
3928 but doesn't restart at beginning of obd_exports_timed list
3931 Severity : enhancement
3933 Description: improve by-nid export eviction on the MDS and OST
3934 Details : allow multiple exports with the same NID to be evicted at one
3935 time without re-searching the exports list.
3938 Frequency : rare, only with supplementary groups enabled on SMP 2.6 kernels
3940 Description: MDS may oops in groups_free()
3941 Details : in rare race conditions a newly allocated group_info struct is
3942 freed again, and this can be NULL. The 2.4 compatibility code
3943 for groups_free() checked for a NULL pointer, but 2.6 did not.
3946 Frequency : common for liblustre clients doing little filesystem IO
3947 Bugzilla : 9352, 7313
3948 Description: server may evict liblustre clients accessing contended locks
3949 Details : if a client is granted a lock or receives a completion AST
3950 with a blocking AST already set it would not reply to the AST
3951 for LDLM_FL_CANCEL_ON_BLOCK locks. It now replies to such ASTs.
3954 Frequency : lfs setstripe, only systems with more than 160 OSTs
3956 Description: unable to set striping with a starting offset beyond OST 160
3957 Details : llapi_create_file() incorrectly limited the starting stripe
3958 index to the maximum single-file stripe count.
3961 Frequency : LDAP users only
3963 Description: lconf did not handle in-kernel recovery with LDAP properly
3964 Details : lconf/LustreDB get_refs() is searching the wrong namespace
3966 Severity : enhancement
3968 Description: bind OST threads to NUMA nodes to improve performance
3969 Details : all OST threads are uniformly bound to CPUs on a single NUMA
3970 node and do their allocations there to localize memory access
3972 Severity : enhancement
3974 Description: llmount can determine client NID directly from Myrinet (GM)
3975 Details : the client NID code from gmnalnid was moved directly into
3976 llmount, removing the need to use this or specifying the
3977 client NID explicitly when mounting GM clients with zeroconf
3980 Frequency : if client is started with down MDS
3982 Description: if client is started with down MDS mount hangs in ptlrpc_queue_wait
3983 Details : Having an LWI_INTR() wait event (interruptible, but no timeout)
3984 will wait indefinitely in ptlrpc_queue_wait->l_wait_event() after
3985 ptlrpc_import_delayed_req() because we didn't check if the
3986 request was interrupted, and we also didn't break out of the
3987 event loop if there was no timeout
3992 Description: data loss during non-page-aligned writes to a single file from
3993 both multiple nodes and multiple threads on one node at same time
3994 Details : updates to KMS and lsm weren't protected by common lock. Resulting
3995 inconsistency led to false short-reads, that were cached and later
3996 used by ->prepare_write() to fill in partially written page,
3997 leading to data loss.
4000 Frequency : always, if lconf --abort_recovery used
4002 Description: lconf --abort_recovery fails with 'Operation not supported'
4003 Details : lconf was attempting to abort recovery on the MDT device and not
4006 Severity : enhancement
4008 Description: remove cleanup logs
4009 Details : replace lconf-generated cleanup logs with lustre internal
4010 cleanup routines. Eliminates the need for client-cleanup and
4013 Severity : enhancement
4015 Description: add support for EAs (user and system) on lustre filesystems
4016 Details : it is now possible to store extended attributes in the Lustre
4017 client filesystem, and with the user_xattr mount option it
4018 is possible to allow users to store EAs on their files also
4020 Severity : enhancement
4022 Description: Add possibility (config option) to show minimal available OST free
4024 Details : When compiled with --enable-mindf configure option, statfs(2)
4025 (and so, df) will return least minimal free space available from
4026 all OSTs as amount of free space on FS, instead of summary of
4027 free spaces of all OSTs.
4029 Severity : enhancement
4031 Description: do not expand extent locks acquired on OST-side
4032 Details : Modify ldlm_extent_policy() to not expand local locks, acquired
4033 by server: they are not cached anyway.
4036 Frequency : when mmap is used/binaries executed from Lustre
4038 Description: Unmmap pages before throwing them away from read cache.
4039 Details : llap_shrink cache now attempts to unmap pages before discarding
4040 them (if unmapping failed - do not discard). SLES9 kernel has
4041 extra checks that trigger if this unmapping is not done first.
4046 Description: lconf didn't resolve symlinks before checking to see whether a
4047 given mountpoint was already in use
4050 Frequency : when migrating failover services
4051 Bugzilla : 6395, 9514
4052 Description: When migrating a subset of services from a node (e.g. failback
4053 from a failover service node) the remaining services would
4054 time out and evict clients.
4055 Details : lconf --force (implied by --failover) sets the global obd_timeout
4056 to 5 seconds in order to quickly disconnect, but this caused
4057 other RPCs to time out too quickly. Do not change the global
4058 obd_timeout for force cleanup, only set it for DISCONNECT RPCs.
4060 Severity : enhancement
4061 Frequency : if MDS is started with down OST
4062 Bugzilla : 9439,5706
4063 Description: Allow startup/shutdown of an MDS without depending on the
4064 availability of the OSTs.
4065 Details : Asynchronously call mds_lov_synchronize during MDS startup.
4066 Add appropriate locking and lov-osc refcounts for safe
4067 cleaning. Add osc abort_inflight calls in case the
4068 synchronize never started.
4071 Frequency : occasional (Cray XT3 only)
4073 Description: root not authorized to access files in CRAY_PORTALS environment
4074 Details : The client process capabilities were not honoured on the MDS in
4075 a CRAY_PORTALS/CRAY_XT3 environment. If the file had previously
4076 been accessed by an authorized user then root was able to access
4077 the file on the local client also. The root user capabilities
4078 are now allowed on the MDS, as this environment has secure UID.
4081 Frequency : occasional
4083 Description: ldiskfs "too long searching" message happens too often
4084 Details : A debugging message (otherwise harmless) prints too often on
4085 the OST console. This has been reduced to only happen when
4086 there are fragmentation problems on the filesystem.
4091 Description: Division by zero in statfs when all OSCs are inactive
4092 Details : lov_get_stripecnt() returns zero due to incorrect order of checks,
4093 lov_statfs divides by value returned by lov_get_stripecnt().
4097 Bugzilla : 9489, 3273
4098 Description: First write from each client to each OST was only 4kB in size,
4099 to initialize client writeback cache, which caused sub-optimal
4100 RPCs and poor layout on disk for the first writen file.
4101 Details : Clients now request an initial cache grant at (re)connect time
4102 and so that they can start streaming writes to the cache right
4103 away and always do full-sized RPCs if there is enough data.
4104 If the OST is rebooted the client also re-establishes its grant
4105 so that client cached writes will be honoured under the grant.
4110 Description: Slow ls (and stat(2) syscall) on files residing on IO-loaded OSTs
4111 Details : Now I/O RPCs go to different portal number and (presumably) fast
4112 lock requests (and glimses) and other RPCs get their own service
4113 threads pool that should be able to service those RPCs
4116 Severity : enhancement
4118 Description: Ability to exchange lustre version between client and servers and
4119 issue warnings at client side if client is too old. Also for
4120 liblustre clients there is ability to refuse connection of too old
4122 Details : New 'version' field is added to connect data structure that is
4123 filled with version info. That info is later checked by server and
4127 Frequency : rare, liblustre only.
4128 Bugzilla : 9296, 9581
4129 Description: Two simultaneous writes from liblustre at offset within same page
4130 might proceed at the same time overwriting eachother with stale
4132 Details : I/O lock withing llu_file_prwv was released too early, before data
4133 actually was hitting the wire. Extended lock-holding time until
4134 server acknowledges receiving data.
4137 Frequency : extremely rare. Never observed in practice.
4139 Description: avoid generating lustre_handle cookie of 0.
4140 Details : class_handle_hash() generates handle cookies by incrementing
4141 global counter, and can hit 0 occasionaly (this is unlikely, but
4142 not impossible, because initial value of cookie counter is
4143 selected randonly). Value of 0 is used as a sentinel meaning
4144 "unassigned handle" --- avoid it. Also coalesce two critical
4145 sections in this function into one.
4147 Severity : enhancement
4149 Description: allow liblustre clients to delegate truncate locking to OST
4150 Details : To avoid overhead of locking, liblustre client instructs OST to
4151 take extent lock in ost_punch() on client's behalf. New connection
4152 flag is added to handle backward compatibility.
4154 Severity : enhancement
4155 Bugzilla : 4928, 7341, 9758
4156 Description: allow number of OST service threads to be specified
4157 Details : a module parameter allows the number of OST service threads
4158 to be specified via "options ost ost_num_threads={N}" in the
4159 OSS's /etc/modules.conf or /etc/modprobe.conf.
4163 Bugzilla : 6146, 9635, 9895
4164 Description: servers crash with bad pointer in target_handle_connect()
4165 Details : In rare cases when a client is reconnecting it was possible that
4166 the connection request was the last reference for that export.
4167 We would temporarily drop the export reference and get a new
4168 one, but this may have been the last reference and the export
4169 was just destroyed. Get new reference before dropping old one.
4171 Severity : enhancement
4172 Frequency : if client is started with failover MDS
4174 Description: Allow multiple MDS hostnames in the mount command
4175 Details : Try to read the configuration from all specified MDS
4176 hostnames during a client mount in case the "primary"
4179 Severity : enhancement
4181 Description: Stop sending data to evicted clients as soon as possible.
4182 Details : Check if the client we are about to send or are sending data to
4183 was evicted already. (Check is done every second of waiting,
4184 for which l_wait_event interface was extended to allow checking
4185 of exit condition at specified intervals).
4188 Frequency : rare, normally only when NFS exporting is done from client
4190 Description: 'bad disk LOV MAGIC: 0x00000000' error when chown'ing files
4192 Details : Make mds_get_md() recognise empty md case and set lmm size to 0.
4195 Frequency : always, if srand() is called before liblustre initialization
4197 Description: Liblustre uses system PRNG disturbing its usage by user application
4198 Details : Introduce internal to lustre fast and high-quality PRNG for
4199 lustre usage and make liblustre and some other places in generic
4200 lustre code to use it.
4202 Severity : enhancement
4203 Bugzilla : 9477, 9557, 9870
4204 Description: Verify that the MDS configuration logs are updated when xml is
4205 Details : Check if the .xml configuration logs are newer than the config
4206 logs stored on the MDS and report an error if this is the case.
4207 Request --write-conf, or allow starting with --old_conf.
4209 Severity : enhancement
4211 Description: Handle symlinks in the path when checking if Lustre is mounted.
4212 Details : Resolve intermediate symlinks when checking if a client has
4213 mounted a filesystem to avoid duplicate client mounts.
4218 Description: lconf can hit an error exception but still return success.
4219 Details : The lconf command catches the Command error exception at the top
4220 level script context and will exit with the associated exit
4221 status, but doesn't ensure that this exit status is non-zero.
4226 Description: failure of ptlrpc thread startup can cause oops
4227 Details : Starting a ptlrpc service thread can fail if there are a large
4228 number of threads or the server memory is very fragmented.
4229 Handle this without oopsing.
4232 Frequency : always, only if liblustre and non-default acceptor port was used
4234 Description: liblustre cannot connect to servers with non-default acceptor port
4235 Details : tcpnal_set_default_params() was not called and was therefore
4236 ignoring the environment varaible TCPNAL_PORT, as well as other
4237 TCPNAL_ environment variables
4242 Description: two objects could be created on the same OST for a single file
4243 Details : If an OST is down, in some cases it was possible to create two
4244 objects on a single OST for a single file. No problems other
4245 than potential performance impact and spurious error messages.
4249 Bugzilla : 5681, 9562
4250 Description: Client may oops in ll_unhash_aliases
4251 Details : Client dcache may become inconsistent in race condition.
4252 In some cases "getcwd" can fail if the current directory is
4258 Description: Inode refcounting problems in NFS export code
4259 Details : link_raw functions used to call d_instantiate without obtaining
4260 extra inode reference first.
4264 Bugzilla : 9942, 9903
4265 Description: Referencing freed requests leading to crash, memleaks with NFS.
4266 Details : We used to require that call to ll_revalidate_it was always
4267 followed by ll_lookup_it. Also with revalidate_special() it is
4268 possible to call ll_revalidate_it() twice for the same dentry
4269 even if first occurence returned success. This fix changes semantic
4270 between DISP_ENQ_COMPLETE disposition flag to mean there is extra
4271 reference on a request referred from the intent.
4272 ll_intent_release() then releases such a request.
4275 Frequency : rare, normally benchmark loads only
4277 Description: unlinked inodes were kept in memory on the client
4278 Details : If a client is repeatedly creating and unlinking files it
4279 can accumulate a lot of stale inodes in the inode slab cache.
4280 If there is no other client load running this can cause the
4281 client node to run out of memory. Instead flush old inodes
4282 from client cache that have the same inode number as a new inode.
4285 Frequency : SLES9 2.6.5 kernel and long filenames only
4286 Bugzilla : 9969, 10379
4287 Description: utime reports stale NFS file handle
4288 Details : SLES9 uses out-of-dentry names in some cases, which confused
4289 the lustre dentry revalidation. Change it to always use the
4293 Frequency : rare, unless heavy write-truncate concurrency is continuous
4294 Bugzilla : 4180, 6984, 7171, 9963, 9331
4295 Description: OST becomes very slow and/or deadlocked during object unlink
4296 Details : filter_destroy() was holding onto the parent directory lock
4297 while truncating+unlinking objects. For very large objects this
4298 may block other threads for a long time and slow overall OST
4299 responsiveness. It may also be possible to get a lock ordering
4300 deadlock in this case, or run out of journal credits because of
4301 the combined truncate+unlink. Solution is to do object truncate
4302 first in one transaction without parent lock, and then do the
4303 final unlink in a new transaction with the parent lock. This
4304 reduces the lock hold time dramatically.
4307 Frequency : rare, 2.4 kernels only
4309 Description: MDS or OST cleanup may trip kernel BUG when dropping kernel lock
4310 Details : mds_cleanup() and filter_cleanup() need to drop the kernel lock
4311 before unmounting their filesystem in order to avoid deadlock.
4312 The kernel_locked() function in 2.4 kernels only checks whether
4313 the kernel lock is held, not whether it is this process that is
4314 holding it as 2.6 kernels do.
4319 Description: MDS or OST may oops/LBUG if a client is connecting multiple times
4320 Details : The client ptlrpc code may be trying to reconnect to a down
4321 server before a previous connection attempt has timed out.
4322 Increase the reconnect interval to be longer than the connection
4323 timeout interval to avoid sending duplicate connections to
4327 Frequency : echo_client brw_test command
4329 Description: fix echo_client to work with OST preallocated code
4330 Details : OST preallocation code (5137) didn't take echo_client IO path
4331 into account: echo_client calls filter methods outside of any
4332 OST thread and, hence, there is no per-thread preallocated
4333 pages and buffers to use. Solution: hijack pga pages for IO. As
4334 a byproduct, this avoids unnecessary data copying.
4338 Bugzilla : 3555, 5962, 6025, 6155, 6296, 9574
4339 Description: Client can oops in mdc_commit_close() after open replay
4340 Details : It was possible for the MDS to return an open request with no
4341 transaction number in mds_finish_transno() if the client was
4342 evicted, but without actually returning an error. Clients
4343 would later try to replay that open and may trip an assertion
4344 Simplify the client close codepath, and always return an error
4345 from the MDS in case the open is not successful.
4348 Frequency : rare, 2.6 OSTs only
4350 Description: OST may deadlock under high load on fragmented files
4351 Details : If there was a heavy load and highly-fragmented OST filesystems
4352 it was possible to have all the OST threads deadlock waiting on
4353 allocation of biovecs, because the biovecs were not released
4354 until the entire RPC IO was completed. Instead, release biovecs
4355 as soon as they are complete to ensure forward IO progress.
4357 Severity : enhancement
4359 Description: Support for specifying external journal device at mount
4360 Details : If an OST or MDS device is formatted with an external journal
4361 device, this device major/minor is stored in the ext3 superblock
4362 and may not be valid for failover. Allow detecting and
4363 specifying the external journal at mount time.
4368 Description: Mounting an MDS with pending unlinked files may cause oops
4369 Details : target_finish_recovery() calls mds_postrecov() which returned
4370 the number of orphans unlinked. mds_lov_connect->mds_postsetup()
4371 considers this an error and immediately begins cleaning up the
4372 lov, just after starting the mds_lov process
4374 Severity : enhancement
4376 Description: Implement 'lfs df' to report actual free space on per-OST basis
4377 Details : Add sub-command 'df' on 'lfs' to report the disk space usage of
4378 MDS/OSDs. Usage: lfs df [-i][-h]. Command Options: '-i' to report
4379 usage of objects; '-h' to report in human readable format.
4381 ------------------------------------------------------------------------------
4383 08-26-2005 Cluster File Systems, Inc. <info@clusterfs.com>
4390 Description: Mounting an ldiskfs file system with mballoc may crash OST node.
4391 Details : ldiskfs mballoc code may reference an uninitialized buddy struct
4392 at startup during orphan unlinking. Instead, skip buddy update
4393 before setup, as it will be regenerated after recovery is complete.
4398 Description: If an OST is inactive, its locks might reference stale inodes.
4399 Details : lov_change_cbdata() must iterate over all namespaces, even if
4400 they are inactive to clear inode references from the lock.
4402 Severity : enhancement
4403 Frequency : occasional, if non-standard max_dirty_mb used
4405 Description: Client will block write RPCs if not enough grant
4406 Details : If a client has max_dirty_mb smaller than max_rpcs_in_flight,
4407 then the client will block writes while waiting for another RPC
4408 to complete instead of consuming its dirty limit. With change
4409 we get improved performance when max_dirty_mb is small.
4411 Severity : enhancement
4412 Bugzilla : 3389, 6253
4413 Description: Add support for supplementary groups on the MDS.
4414 Details : The MDS has an upcall /proc/fs/lustre/mds/{mds}/group_upcall
4415 (set to /usr/sbin/l_getgroups if enabled) which will do MDS-side
4416 lookups for user supplementary groups into a cache.
4420 Description: O_CREAT|O_EXCL open flags in liblustre always return -EEXIST
4421 Details : Make libsysio to not enforce O_EXCL by clearing the flag,
4422 for liblustre O_EXCL is enforced by MDS.
4426 Description: readdir never returns NULL in liblustre.
4427 Details : Corrected llu_iop_getdirentries logic, to return offset of next
4428 dentry in struct dirent.
4432 Frequency : liblustre only, depends on application IO pattern
4433 Description: liblustre clients evicted if not contacting servers
4434 Details : Don't put liblustre clients into the ping_evictor list, so
4435 they will not be evicted by the pinger ever.
4437 Severity : enhancement
4439 Description: Add ability to evict clients by NID from MDS.
4440 Details : By echoing "nid:$NID" string into
4441 /proc/fs/lustre/mds/.../evict_client client with nid that equals to
4442 $NID would be instantly evicted from this MDS and from all active
4443 OSTs connected to it.
4447 Description: Do not query file size twice, somewhat slowing stat(2) calls.
4448 Details : lookup_it_finish() used to query file size from OSTs that was not
4453 Description: service threads change working directory to that of init
4454 Details : Starting lustre service threads may pin the working directory
4455 of the parent thread, making that filesystem busy. Threads
4456 now change to the working directory of init to avoid this.
4460 Frequency : during shutdown only
4461 Description: shutdown with a failed MDS or OST can cause unmount to hang
4462 Details : Don't resend DISCONNECT messages in ptlrpc_disconnect_import()
4467 Frequency : 2.6 only
4468 Description: chmod/chown may include an extra supplementary group
4469 Details : ll{,u}_mdc_pack_op_data() does not properly initialize the
4470 supplementary group and if none is specified this is used.
4473 Bugzilla : 5479 (6816)
4475 Description: Racing open + rm can assert client in mdc_set_open_replay_data()
4476 Details : If lookup is in progress on a file that is unlinked we might try
4477 to revalidate the inode and fail in revalidate after lookup is
4478 complete and ll_file_open() enqueues the open again but
4479 it_open_error() was not checking DISP_OPEN_OPEN errors correctly.
4482 Frequency : always, if lconf --abort_recovery used
4484 Description: lconf --abort_recovery fails with 'Operation not supported'
4485 Details : lconf was attempting to abort recovery on the MDT device and not
4488 ------------------------------------------------------------------------------
4490 2005-08-08 Cluster File Systems, Inc. <info@clusterfs.com>
4495 Frequency : rare (only unsupported configurations with a node running as an
4497 Bugzilla : 6514, 5137
4498 Description: Mounting a Lustre file system on a node running as an OST could
4500 Details : OSTs now preallocates memory needed to write out data at
4501 startup, instead of when needed, to avoid having to
4502 allocate memory in possibly low memory situations.
4503 Specifically, if the file system is mounted on on OST,
4504 memory pressure could force it to try to write out data,
4505 which it needed to allocate memory to do. Due to the low
4506 memory, it would be unable to do so and the node would
4507 become unresponsive.
4509 Severity : enhancement
4511 Description: Addition of lconf --service command line option
4512 Details : lconf now accepts a '--service <arg>' option, which is
4513 shorthand for 'lconf --group <arg> --select <arg>=<hostname>'
4515 Severity : enhancement
4517 Description: Failover mode is now the default for OSTs.
4518 Details : By default, OSTs will now run in failover mode. To return to
4519 the old behaviour, add '--failout' to the lmc line for OSTs.
4521 Severity : enhancement
4523 Description: Health checks are now provided for MDS and OSTs
4524 Details : Additional detailed health check information on MSD and OSTs
4525 is now provided through the procfs health_check value.
4528 Frequency : occasional, depends on IO load
4530 Description: Disk fragmentation on the OSTs could eventually cause slowdowns
4531 after numerous create/delete cycles
4532 Details : The ext3 inode allocation policy would not allocate new inodes
4533 very well on the OSTs because there are no new directories
4534 being created. Instead we look for groups with free space if
4535 the parent directories are nearly full.
4540 Description: Network or server problems during mount may cause partially
4541 mounted clients instead of returning an error.
4542 Details : The config llog parsing code may overwrite the error return
4543 code during mount error handling, returning success instead
4549 Description: MDS can fail to allocate large reply buffers
4550 Details : After long uptimes the MDS can fail to allocate large reply
4551 buffers (e.g. zconf client mount config records) due to memory
4552 fragmentation or consumption by the buffer cache. Preallocate
4553 some large reply buffers so that these replies can be sent even
4554 under memory pressure.
4558 Frequency : rare (liblustre)
4559 Description: fsx running with liblustre complained that using truncate() to
4560 extend the file doesn't work. This patch corrects that issue.
4561 Details : This is the liblustre equivalent of the fix for bug 6196. Fixes
4562 ATTR_SIZE and lsm use in llu_setattr_raw.
4566 Frequency : rare, only 2.6 kernels
4567 Description: Unusual file access patterns on the MDS may result in inode
4568 data being lost in very rare circumstances.
4569 Details : Bad interaction between the ea-in-inode patch and the "no-read"
4570 code in the 2.6 kernel caused the inode and/or EA data not to
4571 be read from disk, causing single-file corruption.
4575 Frequency : rare, only 2.6 filesystems using extents
4576 Description: Heavy concurrent write and delete load may cause data corruption.
4577 Details : It was possible under high-load situations to have an extent
4578 metadata block in the block device cache from a just-unlinked
4579 file overwrite a newly-allocated data block. We now unmap any
4580 metadata buffers that alias just-allocated data blocks.
4584 Frequency : filesystems with default stripe_count larger than 77
4585 Description: lconf+mke2fs fail when formatting filesystem with > 77 stripes
4586 Details : lconf specifies an inode size of 4096 bytes when the default
4587 stripe_count is larger than 77. This conflicts with the default
4588 inode density of 1 per 4096 bytes. Allocate smaller inodes in
4589 this case to avoid pinning too much memory for large EAs.
4591 ------------------------------------------------------------------------------
4593 2005-07-07 Cluster File Systems, Inc. <info@clusterfs.com>
4598 Frequency : rare (extremely heavy IO load with hundreds of clients)
4600 Description: Client is evicted, gets IO error writing to file
4601 Details : lock ordering changes for bug 5492 reintroduced bug 3267 and
4602 caused clients to be evicted for AST timeouts. The fixes in
4603 bug 5192 mean we no longer need to have such short AST timeouts
4604 so ldlm_timeout has been increased.
4607 Frequency : occasional during --force or --failover shutdown under load
4608 Bugzilla : 5949, 4834
4609 Description: Server oops/LBUG if stopped with --force or --failover under load
4610 Details : a collection of import/export refcount and cleanup ordering
4611 issues fixed for safer force cleanup
4614 Frequency : only filesystems larger than 120 OSTs
4615 Bugzilla : 5990, 6223
4616 Description: lfs getstripe would oops on a very large filesystem
4617 Details : lov_getconfig used kfree on vmalloc'd memory
4620 Frequency : only filesystems exporting via NFS to Solaris 10 clients
4621 Bugzilla : 6242, 6243
4622 Description: reading from files that had been truncated to a non-zero size
4623 but never opened returned no data
4624 Details : ll_file_read() reads zeros from no-object files to EOF
4629 Description: A bug in MDS/OSS recovery could cause the OSS to fail an assertion
4630 Details : There's little harm in aborting MDS/OSS recovery and letting it
4631 try again, so I removed the LASSERT and return an error instead.
4633 Severity : enhancement
4635 Description: New debugging infrastructure for tracking down data corruption
4636 Details : The I/O checksum code was replaced to: (a) control it at runtime,
4637 (b) cover more of the client-side code path, and (c) try to narrow
4638 down where problems occurred
4642 Bugzilla : 3819, 4364, 4397, 6313
4643 Description: Racing close and eviction MDS could cause assertion in mds_close
4644 Details : It was possible to get multiple mfd references during close and
4645 client eviction, leading to one thread referencing a freed mfd.
4647 Severity: : enhancement
4648 Bugzilla : 3262, 6359
4649 Description: Attempts to reconnect to servers are now more aggressive.
4650 Details : This builds on the enhanced upcall-less recovery that was added
4651 in 1.4.2. When trying to reconnect to servers, clients will
4652 now try each server in the failover group every 10 seconds. By
4653 default, clients would previously try one server every 25 seconds.
4658 Description: After recovery, certain operations trigger a failed
4659 assertion on a client.
4660 Details : Failing over an mds, using lconf -d --failover, while a
4661 client was doing a readdir() call would cause the client to
4662 LBUG after recovery completed and the readdir() was resent.
4664 Severity : enhancement
4666 Description: Default groups are now added by lconf
4667 Details : You can now run lconf --group <servicename> without having to
4668 manually add groups with lmc.
4671 Frequency : occasional
4673 Description: Nodes with an elan id of 0 trigger a failed assertion
4676 Frequency : always when accessing e.g. tty/console device nodes
4678 Description: tty and some other devices nodes cannot be used on lustre
4679 Details : file's private_data field is used by device data and lustre
4680 values in there got lost. New field was added to struct file to
4681 store fs-specific private data.
4684 Frequency : when exporting Lustre via NFS
4686 Description: NFSD failed occasionally when looking up a path component
4687 Details : NFSD is looking up ".." which was broken in ext3 directories
4688 that had grown large enough to become hashed.
4691 Frequency : Clusters with multiple interfaces not on the same subnet
4693 Description: Nodes will repeatedly try to reconnect to an interface which it
4694 cannot reach and report an error to the log.
4695 Details : Extra peer list entries will be created by lconf with some peers
4696 unreachable. lconf now validates the peer before adding it.
4699 Frequency : Only if a default stripe is set on the filesystem root.
4701 Description: Setting a default stripe on the filesystem root prevented the
4702 filesystem from being remounted.
4703 Details : The client was sending extra request flags in the root getattr
4704 request and did not allocate a reply buffer for the dir EA.
4707 Frequency : occasional, higher if lots of files are accessed by one client
4708 Bugzilla : 6159, 6097
4709 Description: Client trips assertion regarding lsm mismatch/magic
4710 Details : While revalidating inodes the VFS looks up inodes with ifind()
4711 and in rare cases can find an inode that is being freed.
4712 The ll_test_inode() code will free the lsm during ifind()
4713 when it finds an existing inode and then the VFS later attaches
4714 this free lsm to a new inode.
4718 Bugzilla : 6422, 7030
4719 Description: MDS deadlock between mkdir and client eviction
4720 Details : Creating a new file via mkdir or mknod (starting a transaction
4721 and getting the ns lock) can deadlock with client eviction
4722 (gets ns lock and trying to finish a synchronous transaction).
4725 Frequency : occasional
4726 Description: While starting a server, the fsfilt_ext3 module could not be
4728 Details : CFS's improved ext3 filesystem is named ldiskfs for 2.6
4729 kernels. Previously, lconf would still use the ext3 name
4730 when trying to load modules. Now, it will correctly use
4731 ext3 on 2.4 and ldiskfs on 2.6.
4733 Severity : enhancement
4734 Description: The default stripe count has been changed to 1
4735 Details : The interpretation of the default stripe count (0, to lfs
4736 or lmc) has been changed to mean striping across a single
4737 OST, rather than all available. For general usage we have
4738 found a stripe count of 1 or 2 works best.
4740 Severity : enhancement
4741 Description: Add support for compiling against Cray portals.
4742 Details : Conditional compiling for some areas that are different
4746 Frequency : occasional
4747 Bugzilla : 6409, 6834
4748 Description: Creating files with an explicit stripe count may lead to
4749 a failed assertion on the MDS
4750 Details : If some OSTs are full or unavailable, creating files may
4751 trigger a failed assertion on the MDS. Now, Lustre will
4752 try to use other servers or return an error to the
4756 Frequency : occasional
4758 Description: Multiple concurrent overlapping read+write on multiple SMP nodes
4759 caused lock timeout during readahead (since 1.4.2).
4760 Details : Processes doing readahead might match a lock that hasn't been
4761 granted yet if there are overlapping and conflicting lock
4762 requests. The readahead process waits on ungranted lock
4763 (original lock is CBPENDING), while OST waits for that process
4764 to cancel CBPENDING read lock and eventually evicts client.
4766 Severity : enhancement
4768 Description: Initial enabling of flock support for clients
4769 Details : Implements fcntl advisory locking and file status functions.
4770 This feature is provided as an optional mount flag (default
4771 off), and is NOT CURRENTLY SUPPORTED. Not all types of record
4772 locking are implemented yet, and those that are are not guaranteed
4773 to be completely correct in production environments.
4774 mount -t lustre -o [flock|noflock] ...
4777 Frequency : occasional
4779 Description: OSTs running 2.4 kernels but with extents enabled might trip an
4780 assertion in the ext3 JBD (journaling) layer.
4781 Details : The b_committed_data struct is protected by the big kernel lock
4782 in 2.4 kernels, serializing journal_commit_transaction() and
4783 ext3_get_block_handle->ext3_new_block->find_next_usable_block()
4784 access to this struct. In 2.6 kernels there is finer grained
4785 locking to improve SMP performance of the JBD layer.
4789 Description: Changes the "SCSI I/O Stats" kernel patch to default to "enabled"
4794 Description: merge and cleanup kernel patches.
4797 -----------------------------------------------------------------------------
4799 2005-05-05 Cluster File Systems, Inc. <info@clusterfs.com>
4801 NOTE: Lustre 1.4.2 uses an incompatible network protocol than previous
4802 versions of Lustre. Please update all servers and clients to
4803 version 1.4.2 or later at the same time. You must also run
4804 "lconf --write-conf {config}.xml" on the MDS while it is stopped
4805 to update the configuration logs.
4807 - fix for HPUX NFS client breakage when NFS exporting Lustre (5781)
4808 - mdc_enqueue does not need max_mds_easize request buffer on send (5707)
4809 - swab llog records of type '0' so we get proper header size/idx (5861)
4810 - send llog cancel req to DLM cancel portal instead of cb portal (5515)
4811 - fix rename of one directory over another leaking an inode (5953)
4812 - avoid SetPageDirty on 2.6 (5981)
4813 - don't re-add just-being-destroyed locks to the waiting list (5653)
4814 - when creating new directories, inherit the parent's custom
4815 striping settings if present parent (3048)
4816 - flush buffers from cache before direct IO in 2.6 obdfilter (4982)
4817 - don't hold i_size_sem in ll_nopage() and ll_ap_refresh_count (6077)
4818 - don't hold client locks on temporary worklist from l_lru (5666)
4819 - handle IO errors in 2.6 obdfilter bio completion routine (6046)
4820 - automatically evict dead clients (5921)
4821 - Update file size properly in create+truncate+fstat case (6196)
4822 - Do not unhash mountpoint dentries, do not allow removal of
4824 - Avoid lock ordering deadlock issue with write/truncate (6203,5654)
4825 - reserve enough journal credits in fsfilt_start_log for setattr (4554)
4826 - ldlm_enqueue freed-export error path would always LBUG (6149,6184)
4827 - don't reference lr_lvb_data until after we hold lr_lvb_sem (6170)
4828 - don't overwrite last_rcvd if there is a *_client_add() error (6086)
4829 - Correctly handle reads of files with no objects (6243)
4830 - lctl recover will also mark a device active if deactivate used (5933)
4832 - by default create 1 inode per 4kB space on MDS, per 16kB on OSTs
4833 - allow --write-conf on an MDS with different nettype than client (5619)
4834 - don't write config llogs to MDS for mounts not from that MDS (5617)
4835 - lconf should create multiple TCP connections from a client (5201)
4836 - init scripts are now turned off by default; run chkconfig --on
4837 lustre and chkconfig --on lustrefs to use them
4838 - upcalls are no longer needed for clients to recover to failover
4840 - add --abort-recovery option to lconf to abort recovery on device
4842 - add support for an arbitrary number of OSTs (3026)
4843 - Quota support protocol changes.
4844 - forward compatibility changes to wire structs (6007)
4845 - rmmod NALs that might be loaded because of /etc/modules.conf (6133)
4846 - support for mountfsoptions and clientoptions to the Lustre LDAP (5873)
4847 - improved "lustre status" script
4848 - initialize blocksize for non-regular files (6062)
4849 - added --disable-server and --disable-client configure options (5782)
4850 - introduce a lookup cache for lconf to avoid repeated DB scans (6204)
4851 - Vanilla 2.4.29 support
4852 - increase maximum number of obd devices to 520 (6242)
4853 - remove the tcp-zero-copy patch from the suse-2.4 series (5902)
4854 - Quadrics Elan drivers are now included for the RHEL 3 2.4.21 and
4855 SLES 9 2.6.5 kernels
4856 - limit stripes per file to 160 (the maximum EA size) (6093)
4858 2005-03-22 Cluster File Systems, Inc. <info@clusterfs.com>
4861 - don't LASSERT in ll_release on NULL lld with NFS export (4655, 5760)
4862 - hold NS lock when calling handle_ast_error->del_waiting_lock (5746)
4863 - fix setattr mtime regression from lovcleanup merge (4829, 5669)
4864 - workaround for 2.6 crash in ll_unhash_aliases (5687, 5210)
4865 - small ext3 extents cleanups and fixes (5733)
4866 - improved mballoc code, several small races and bugs fixed (5733, 5638)
4867 - kernel version 43 - fix remove_suid bugs in both 2.4 and 2.6 (5695)
4868 - avoid needless client->OST connect, fix handle mismatch (5317)
4869 - fix DLM error path that led to out-of-sync client, long delays (5779)
4870 - support common vfs-enforced mount options (nodev,nosuid,noexec) (5637)
4871 - fix several locking issues related to i_size (5492,5624,5654,5672)
4872 - don't move pending lock onto export if it is already evicted (5683)
4873 - fix kernel oops when creating .foo in unlinked directory (5548)
4874 - fix deadlock in obdfilter statistics vs. object create (5811)
4875 - use time_{before,after} to avoid timer jiffies wrap (5882)
4876 - shutdown --force/--failover stability (3607,3651,4797,5203,4834)
4877 - Do not leak request if server was not able to process it (5154)
4878 - If mds_open unable to find parent dir, make that negative lookup(5154)
4879 - don't create new directories with extent-mapping (5909, 5936)
4881 - fix lustre/lustrefs init scripts for SuSE (patch from Scali, 5702)
4882 - don't hold the pinger_sem in ptlrpc_pinger_sending_on_import
4883 - change obd_increase_kms to obd_adjust_kms (up or down) (5654)
4884 - lconf, lmc search both /usr/lib and /usr/lib64 for Python libs (5800)
4885 - support for RHEL4 kernel on i686 (5773)
4886 - provide error messages when incompatible logs are encountered (5898)
4888 2005-02-18 Cluster File Systems, Inc. <info@clusterfs.com>
4889 * version 1.4.0.10 (1.4.1 release candidate 1)
4891 - don't keep a lock reference when lock is not granted (4238)
4892 - unsafe list practices (rarely) led to infinite eviction loop (4908)
4893 - add per-fs limit of Lustre pages in page cache, avoid OOM (4699)
4894 - drop import inflight refcount on signal_completed_replay error (5255)
4895 - unlock page after async write error during send (3677)
4896 - handle missing objects in filter_preprw_read properly (5265)
4897 - no transno return for symlink open, don't save no-trasno open (3440)
4898 - don't try to complete elan receive that already failed (4012)
4899 - free RPC server reply state on error (5406)
4900 - clean up thread from ptlrpc_start_thread() on error (5160)
4901 - readahead could read extra page into cache that wasn't ejected (5388)
4902 - prevent races in class_attach/setup/cleanup/detach (5260)
4903 - don't dereference de->d_inode after l_dput of de (5458)
4904 - use "int" for stripe value returned from lock_to_stripe (5544)
4905 - mballoc allocation and error-checking fixes in 2.6 (5504)
4906 - block device patches to fix I/O request sizes in 2.6 (5482)
4907 - look up hostnames for IB nals (5602)
4908 - 2.6 changed lock ordering of 2 semaphores, caused deadlock (5654)
4909 - don't start multiple acceptors for the same port (5277)
4910 - fix incorrect LASSERT in mds_getattr_name (5635)
4911 - export a proc file for general "ping" checking (5628)
4912 - fix "lfs check" to not block when the MDS is down (5628)
4914 - service request history (4965)
4915 - put {ll,lov,osc}_async_page structs in a single slab (4699)
4916 - create an "evict_client" /proc entry on OSTs, like the MDS has
4917 - fix mount usage message, return errors per mount(8) (5168)
4918 - change grep [] to grep "[]" in tests so they work in more UMLs
4919 - fix ppc64/x86_64 spec to use %{_libdir} instead of /usr/lib (5389)
4920 - remove ancient LOV_MAGIC_V0 EA support (5047)
4921 - add "disk I/Os in flight" and "I/O req time" stats in obdfilter
4922 - align r/w RPCs to PTLRPC_MAX_BRW_SIZE boundary for performance (3451)
4923 - allow readahead allocations to fail when low on memory (5383)
4924 - mmap locking landed again, after considerable improvement (2828)
4925 - add get_hostaddr() to lustreDB.py for LDAP support (5459)
4927 2004-11-23 Cluster File Systems, Inc. <info@clusterfs.com>
4930 - send OST transaction number in read/write reply to free req (4966)
4931 - don't ASSERT in ptl_send_rpc() if we run out of memory (5119)
4932 - lock /proc/sys/portals/routes internal state, avoiding oops (4827)
4933 - the watchdog thread now runs as interruptible (5246)
4934 - flock/lockf fixes (but it's still disabled, pending 5135)
4935 - don't use EXT3 constants in llite code (5094)
4936 - memory shortage at startup could cause assertion (5176)
4938 - reorganization of lov code
4939 - single portals codebase
4941 - add extents/mballoc support (5025)
4942 - direct I/O reads in the obdfilter (4048)
4943 - kernel patches from LNXI for 2.6 (bluesmoke, perfctr, mtd, kexec)
4945 tbd Cluster File Systems, Inc. <info@clusterfs.com>
4948 - send OST transaction number in read/write reply to free req (4966)
4949 - don't ASSERT in ptl_send_rpc() if we run out of memory (5119)
4950 - lock /proc/sys/portals/routes internal state, avoiding oops (4827)
4951 - the watchdog thread now runs as interruptible (5246)
4952 - handle missing objects in filter_preprw_read properly (5265)
4953 - unsafe list practices (rarely) led to infinite eviction loop (4908)
4954 - drop import inflight refcount on signal_completed_replay error (5255)
4955 - unlock page after async write error during send (3677)
4956 - return original error code on reconstructed replies (3761)
4957 - no transno return for symlink open, don't save no-trasno open (3440)
4959 - add pid to ldlm debugging output (4922)
4960 - bump the watchdog timeouts -- we can't handle 30sec yet
4961 - extra debugging for orphan dentry/inode bug (5259)
4963 2004-11-16 Cluster File Systems, Inc. <info@clusterfs.com>
4966 - fix TCP_NODELAY bug, which caused extreme perf regression (5134)
4967 - allocate qswnal tx descriptors singly to avoid fragmentation (4504)
4968 - don't LBUG on obdo_alloc() failure, use OBD_SLAB_ALLOC() (4800)
4969 - fix NULL dereference in /proc/sys/portals/routes (4827)
4970 - allow failed mdc_close() operations to be interrupted (4561)
4971 - stop precreate on OST before MDS would time out on it (4778)
4972 - don't send partial-page writes before EOF from client (4410)
4973 - discard client grant for sub-page writes on large-page clients (4520)
4974 - don't free dentries not owned by NFS code, check generation (4806)
4975 - fix lsm leak if mds_create_objects() fails (4801)
4976 - limit debug_daemon file size, always print CERROR messages (4789)
4977 - use transno after validating reply (3892)
4978 - process timed out requests if import state changes (3754)
4979 - update mtime on OST during writes, return in glimpse (4829)
4980 - add mkfsoptions to LDAP (4679)
4981 - use ->max_readahead method instead of zapping global ra (5039)
4982 - don't interrupt __l_wait_event() during strace
4984 - add software watchdogs to catch hung threads quickly (4941)
4985 - make lustrefs init script start after nfs is mounted
4986 - fix CWARN/ERROR duplication (4930)
4987 - return async write errors to application if possible (2248)
4988 - add /proc/sys/portal/memused (bytes allocated by PORTALS_ALLOC)
4989 - print NAL number in %x format (4645)
4990 - update barely-supported suse-2.4.21-171 series (4842)
4991 - support for sles 9 %post scripts
4992 - support for building 2.6 kernel-source packages
4993 - support for sles km_* packages
4995 2004-10-07 Cluster File Systems, Inc. <info@clusterfs.com>
4998 - ignore -ENOENT errors in osc_destroy (3639)
4999 - notify osc create thread that OSC is being cleaned up (4600)
5000 - add nettype argument for llmount in #5d in conf-sanity.sh (3936)
5001 - reconstruct ost_handle() like mds_handle() (4657)
5002 - create a new thread to do import eviction to avoid deadlock (3969)
5003 - let lconf resolve symlinked-to devices (4629)
5004 - don't unlink "objects" from directory with default EA (4554)
5005 - hold socknal file ref over connect in case target is down (4394)
5006 - allow more than 32000 subdirectories in a single directory (3244)
5007 - fix blocks count for O_DIRECT writes (3751)
5008 - OST returns ENOSPC from object create when no space left (4539)
5009 - don't send truncate RPC if file size isn't changing (4410)
5010 - limit OSC precreate to 1/2 of value OST considers bogus (4778)
5011 - bind to privileged port in socknal and tcpnal (3689)
5013 - rate limit CERROR/CWARN console message to avoid overload (4519)
5014 - GETFILEINFO dir ioctl returns LOV EA + MDS stat in 1 call (3327)
5015 - basic mmap support (3918)
5016 - kernel patch series update from b1_4 (4711)
5018 2004-09-16 Cluster File Systems, Inc. <info@clusterfs.com>
5021 - avoid crash during MDS cleanup with OST shut down (2775)
5022 - fix loi_list_lock/oig_lock inversion on interrupted IO (4136)
5023 - don't use bad inodes on the MDS (3744)
5024 - dynamic object preallocation to improve recovery speed (4236)
5025 - don't hold spinlock over lock dumping or change debug flags (4401)
5026 - don't zero obd_dev when it is force cleaned (3651)
5027 - print grants to console if they go negative (4431)
5028 - "lctl deactivate" will stop automatic recovery attempts (3406)
5029 - look for existing locks in ldlm_handle_enqueue() (3764)
5030 - don't resolve lock handle twice in recovery avoiding race (4401)
5031 - revalidate should check working dir is a directory (4134)
5033 - don't always mark "slow" obdfilter messages as errors (4418)
5035 2004-08-24 Cluster File Systems, Inc. <info@clusterfs.com>
5038 - don't close LustreDB during write_conf until it is done (3860)
5039 - fix typo in lconf for_each_profile (3821)
5040 - allow dumping logs from multiple threads at one time (3820)
5041 - don't allow multiple threads in OSC recovery (3812)
5042 - fix debug_size parameters (3864)
5043 - fix mds_postrecov to initialize import for llog ctxt (3121)
5044 - replace config semaphore with spinlock (3306)
5045 - be sure to send a reply for a CANCEL rpc with bad export (3863)
5046 - don't allow enqueue to complete on a destroyed export (3822)
5047 - down write_lock before checking llog header bitmap (3825)
5048 - recover from lock replay timeout (3764)
5049 - up llog sem before sending rpc (3652)
5050 - reduce ns lock hold times when setting kms (3267)
5051 - change a dlm LBUG to LASSERTF, to maybe learn something (4228)
5052 - fix NULL deref and obd_dev leak on setup error (3312)
5053 - replace some LBUG about llog ops with error handling (3841)
5054 - don't match INVALID dentries from d_lookup and spin (3784)
5055 - hold dcache_lock while marking dentries INVALID and hashing (4255)
5056 - fix invalid assertion in ptlrpc_set_wait (3880)
5058 - add libwrap support for the TCP acceptor (3996)
5059 - add /proc/sys/portals/routes for non-root route listing (3994)
5060 - allow setting MDS UUID in .xml (2580)
5061 - print the stack of a process that LBUGs (4228)
5063 2004-07-14 Cluster File Systems, Inc. <info@clusterfs.com>
5066 - don't cleanup request in ll_file_open() on failed MDS open (3430)
5067 - make sure to unset replay flag from failed open requests (3440)
5068 - if default stripe count is 0, use OST count for inode size (3636)
5069 - update parent mtime/ctime on client for create/unlink (2611)
5070 - drop dentry ref in ext3_add_link from open_connect_dentry (3266)
5071 - free recovery state on server during a forced cleanup (3571)
5072 - unregister_reply for resent reqs (3063)
5073 - loop back devices mounting and status check on 2.6 (3563)
5074 - fix resource-creation race that can provoke i_size == 0 (3513)
5075 - don't try to use bad inodes returned from MDS/OST fs lookup (3688)
5076 - more debugging for page-accounting assertion (3746)
5077 - return -ENOENT instead of asserting if ost getattr+unlink race (3558)
5078 - avoid deadlock after precreation failure (3758)
5079 - fix race and lock order deadlock in orphan handling (3450, 3750)
5080 - add validity checks when grabbing inodes from l_ast_data (3599)
5082 - add /proc/.../recovery_status to obdfilter (3428)
5083 - lightweight CDEBUG infrastructure, debug daemon (3668)
5084 - change default OSC RPC parameters to be better on small clusters
5085 - turn off OST read cache for files smaller than 32MB
5086 - install man pages and include them in rpms (3100)
5087 - add new init script for (un)mounting lustre filesystems (2593)
5088 - run chkconfig in %post for init scripts (3701)
5089 - drop scimac NAL (unmaintained)
5091 2004-06-17 Cluster File Systems, Inc. <info@clusterfs.com>
5094 - clean kiobufs before and after use (3485)
5095 - strip trailing '/'s before comparing paths with /proc/mounts (3486)
5096 - remove assertions to work around "in-flight rpcs" recovery bug (3063)
5097 - change init script to fail more clearly if not run as root (1528)
5098 - allow clients to reconnect during replay (1742)
5099 - fix ns_lock/i_sem lock ordering deadlock for kms update (3477)
5100 - don't do DNS lookups on NIDs too small for IP addresses (3442)
5101 - re-awaken ptlrpcd if new requests arrive during check_set (3554)
5102 - fix cond_resched (3554)
5103 - only evict unfinished clients after recovery (3515)
5104 - allow bulk resend, prevent data loss (3570)
5105 - dynamic ptlrpc request buffer allocation (2102)
5106 - don't allow unlinking open directory if it isn't empty (2904)
5107 - set MDS/OST threads to umask 0 to not clobber client modes (3359)
5108 - remove extraneous obd dereference causing LASSERT failure (3334)
5109 - don't use get_cycles() when creating temp. files on the mds (3156)
5110 - hold i_sem when setting i_size in ll_extent_lock() (3564)
5111 - handle EEXIST for set-stripe, set proper directory name (3336)
5113 - servers can dump a log evicting a client - lustre.dump_on_timeout=1
5114 - fix ksocknal_fmb_callback() error messages (2918)
5116 2004-05-27 Cluster File Systems, Inc. <info@clusterfs.com>
5119 - don't copy lvb into (possibly NULL) reply on error (2983)
5120 - don't deref dentry after dput, don't free lvb on error (2922)
5121 - use the kms to determine writeback rpc length (2947)
5122 - increment oti_logcookies when osc is inactive (2948)
5123 - update client's i_blocks count via lvb messages (2543)
5124 - handle intent open/close of special files properly (1557)
5125 - mount MDS with errors=remount-ro, like obdfilter (2009)
5126 - initialize lock handle to avoid ASSERT on error cleanup (3057)
5127 - don't use cancelling-locks' kms values (2947)
5128 - use highest lock extent for kms, not last one (2925)
5129 - don't dereference ERR_PTR() dentry in error handling path (3107)
5130 - fix thread race in portals_debug_dumplog() (3122)
5131 - create lprocfs device entries at setup instead of at attach (1519)
5132 - common AST error handler, don't evict client on completion race (3145)
5133 - zero nameidata in detach_mnt in 2.6 (3118)
5134 - verify d_inode after revalidate_special is valid in 2.6 (3116)
5135 - use lustre_put_super() to handle zconf unmounts in 2.6 (3064)
5136 - initialize RPC timeout timer earlier for 2.6 (3219)
5137 - don't dereference NULL reply buffer if mdc_close was never sent (2410)
5138 - print nal/nid for unknown nid (3258)
5139 - additional checks for oscc recovery before doing precreate (3284)
5140 - fix ll_extent_lock() error return code for 64-bit systems (3043)
5141 - don't crash in mdc_close for bad permissions on open (3285)
5142 - zero i_rdev for non-device files (3147)
5143 - clear page->private before handing to FS, better assertion (3119)
5144 - tune the read pipeline (3236)
5145 - fix incorrect decref of invalidated dentry (2350)
5146 - provide read-ahead stats and refine rpc in flight stats (3328)
5147 - don't hold journal transaction open across create RPC (3313)
5148 - update atime on MDS at close time (3265)
5149 - close LDAP connection when recovering to avoid server load (3315)
5150 - update iopen-2.6 patch with fixes from 2399,2517,2904 (3301)
5151 - don't leak open file on MDS after open resend (3325)
5152 - serialize filter_precreate and filter_destroy_precreated (3329)
5153 - loop device shouldn't call sync_dev() for nul device (3092)
5154 - clear page cache after eviction (2766)
5155 - resynchronize MDS->OST in background (2824)
5156 - refuse to mount the same filesystem twice on same mountpoint (3394)
5157 - allow llmount to create routes for mounting behind routers (3320)
5158 - push lock cancellation to blocking thread for glimpse ASTs (3409)
5159 - don't call osc_set_data_with_check() for TEST_LOCK matches (3159)
5160 - fix rare problem with rename on htree directories (3417)
5162 - allow default OST striping configuration per directory (1414)
5163 - fix compilation for qswnal for 2.6 kernels (3125)
5164 - increase maximum number of MDS request buffers for large systems
5165 - change liblustreapi to be useful for external progs like lfsck (3098)
5166 - increase local configuration timeout for slow disks (3353)
5167 - allow configuring ldlm AST timeout - lustre.ldlm_timeout=<seconds>
5169 2004-03-22 Cluster File Systems, Inc. <info@clusterfs.com>
5172 - fixes for glimpse AST timeouts / incorrectly 0-sized files (2818)
5173 - don't overwrite extent policy data in reply if lock was blocked (2901)
5174 - drop filter export grants atomically with removal from device (2663)
5175 - del obd_self_export from work_list in class_disconnect_exports (2908)
5176 - don't LBUG if MDS recovery times out during orphan cleanup (2530)
5177 - swab reply message in mdc_close, other PPC fixes (2464)
5178 - fix destroying of named logs (2325)
5179 - overwrite old logs when running lconf --write_conf (2264)
5180 - bump LLOG_CHUNKSIZE to 8k to allow for larger clusters (2306)
5181 - fix race in target_handle_connect (2898)
5182 - mds_reint_create() should take same inode create lock (2926)
5183 - correct journal credits calculated for CANCEL_UNLINK_LOG (2931)
5184 - don't close files for self_export to avoid uninitialized obd (2936)
5185 - allow MDS with the same name as client node (2939)
5186 - hold dentry reference for closed log files for unlink (2325)
5187 - reserve space for all logs during transactions (2059)
5188 - don't evict page beyond end of stripe extent (2925)
5189 - don't oops on a deleted current working directory (2399)
5190 - handle hard links to targets without a parent properly (2517)
5191 - don't dereference NULL lock when racing during eviction (2867)
5192 - don't grow lock extents when lots of conflicting locks (2919)
5194 2004-03-04 Cluster File Systems, Inc. <info@clusterfs.com>
5197 - account for cache space usage on clients to avoid data loss (974)
5198 - lfsck support in lustre kernel code (2349)
5199 - reduce journal credits needed for BRW writes (2370)
5200 - orphan handling to avoid losing space on client/server crashes
5201 - ptlrpcd can be blocked, stopping ALL progress (2477)
5202 - use lock value blocks to assist in proper KMS, faster stat (1021)
5203 - takes i_sem instead of DLM locks internally on obdfilter (2720)
5204 - recovery for initial connections (2355)
5205 - fixes for mds_cleanup_orphans (1934)
5206 - abort_recovery crashes MDS in b_eq (mds_unlink_orphan) (2584)
5207 - block all file creations until orphan recovery completes (1901)
5208 - client remove rq_connection from request struct (2423)
5209 - conf-sanity test_5, proper cleanup in umount log not availale (2640)
5210 - recovery timer race (2670)
5211 - mdc_close recovey bug (2532)
5212 - ptlrpc cleanup bug (2710)
5213 - mds timeout on local locks (2588)
5214 - namespace lock held during RPCs (2431)
5215 - handle interrupted sync write properly (2503)
5216 - don't try to handle a message that hasn't been replied to (2699)
5217 - client assert failure during cleanup after abort recovery (2701)
5218 - leak mdc device after failed mount (2712)
5219 - ptlrpc_check_set allows timedout requests to complete (2714)
5220 - wait for inflight reqs when ptlrpcd finishes (2710)
5221 - make sure unregistered services are removed from the srv_list
5222 - reset bulk XID's when resending them (caught by 1138 test)
5223 - unregister_bulk after timeout
5224 - fix lconf error (2694)
5225 - handle write after unfinished setstripe, stripe-only getstripe (2388)
5226 - readahead locks pages, leaves pending causing memory pressure (2673)
5227 - increase OST request buffers to 4096 on large machines (2729)
5228 - fix up permission of existing directories in simple_mkdir (2661)
5229 - init deleted item, add assertions ptlrpc_abort_inflight() (2725)
5230 - don't assign transno to errored transactions (2742)
5231 - don't delete objects on OST if given a bogus objid from MDS (2751)
5232 - handle large client PAGE_SIZE readdir on small PAGE_SIZE MDS (2777)
5233 - if rq_no_resend, then timeout request after recovery (2432)
5234 - fix MDS llog_logid record size, 64-bit array alignment (2733)
5235 - don't call usermode_helper from ptlrpcd, DEFAULT upcall (2773)
5236 - put magic in mount.lustre data, check for bad/NULL mount data (2529)
5237 - MDS recovery shouldn't delete objects that it has given out (2730)
5238 - if enqueue arrives after completion, don't clobber LVB (2819)
5239 - don't unlock pages twice when trigger_group_io returns error (2814)
5240 - don't deref NULL rq_repmsg if ldlm_handle_enqueue failed (2822)
5241 - don't write pages to disk if there was an error (1450)
5242 - don't ping imports that have recovery disabled (2676)
5243 - take buffered bytes into account when balancing socknal conn (2817)
5244 - hold a DLM lock over readdir always, use truncate_inode_pages (2706)
5245 - reconnect unlink llog connection after MDS reconnects to OST (2816)
5246 - remove little-endian swabbing of llog records (1987)
5247 - set/limit i_blksize to LL_MAX_BLKSIZE on client (2884)
5248 - retry reposting request buffers if they fail (1191)
5249 - grow extent at grant time to avoid granting a revoked lock (2809)
5250 - lock revoke doesn't evict page if covered by a second lock (2765)
5251 - disable VM readahead to avoid reading outside lock extents (2805)
5253 - return LL_SUPER_MAGIC from statfs for the filesystem type (1972)
5254 - updated kernel patches for hp-2.4.20 kernel (2681)
5256 2004-02-07 Cluster File Systems, Inc. <info@clusterfs.com>
5259 - fix truncated write corruption (2366)
5260 - fix for failed assertion in iopen_connect_dentry (1792,2517)
5262 - don't flag the ptlrpcd thread with PF_MEMALLOC (2636)
5263 - ensure len(uuid) < 37 in lmc (1171)
5264 - fix ia64 OOPS in llog_test (2255)
5265 - zero end of page at obdfilter for partial page writes (2648)
5266 - don't leave stale dentries around after renames (bug 2428)
5267 - fix timeouts when evicting a client with a single lock held (2642)
5268 - set deadline for the initial HELLO message to drain (2634)
5269 - print out dotted-quad IP addresses in the socknal (2302)
5271 - additional debugging for MDS client eviction problem (2443)
5272 - fix mkfsoptions support for osts (2603, 2604)
5274 2004-01-27 Cluster File Systems, Inc. <info@clusterfs.com>
5277 - add series for the vanilla 2.6.0 kernel
5278 - add series for the vanilla 2.4.24 kernel
5279 - add series for a cray x86/64 UL kernel drop
5280 - fix xattr patches for the vanilla 2.4.19 series
5282 - generate true UUIDs in lmc (1171)
5283 - have portals stack dumping break in UML (2466)
5284 - avoid bad dchild deref; avoid inum lock w/o creation (2362)
5285 - allocate with _NOFS in ldlm to avoid deadlock (1933)
5286 - wake callback waiting threads on client eviction (2460)
5287 - Add --ptldebug and --subsystem to lmc (1719)
5288 - update assertion to allow safe interrupt allocation
5289 - set rq_no_resend for cancel requests (2432)
5290 - recalculate ptlrpcd timeout after resend (2494)
5291 - call vfs_rmdir when removing pending directories (2368)
5292 - fix renaming a file to itself (2429)
5293 - lmc creates a default one-stripe lov (2454)
5294 - expand procfs space to handle large clusters (2326)
5295 - increase UML stack to avoid overflow
5296 - update lconf's list of debug and subsystem masks
5297 - fix lfs find --obd (2510)
5298 - /proc tunable for disabling filter read caching (2591)
5299 - stop rpm packages from altering slapd.conf (2301)
5300 - disable nagle in the socknal under 0conf (2578)
5301 - choose mds inode size based on stripe count (2572)
5302 - fix kernel-source rpm problems (2516)
5304 - add --disable-doc to avoid pdf generation (2421)
5305 - update documentation, tests, type-os, comments
5306 - avoid format warnings on ia64
5307 - remove the TOE NAL
5308 - tiny code cleanups by removing unused fields
5310 2004-01-07 Cluster File Systems, Inc. <info@clusterfs.com>
5313 - fix obvious semaphore misuse in as-yet-unused setattr path (2348)
5314 - remove the most blatant lies from BUILDING file (2371)
5315 - change default debug level to reasonable production setting
5316 - reduce client side cache size to reduce cache flush time
5317 - reduce max RPCs in flight to avoid unnecessary file fragmentation
5318 - make TCP zerocopy and pinger support enabled by default (2476)
5319 - sync writes completed after process exits caused crashes (2319)
5320 - maintain correct mount count on the MDS (2356)
5321 - backout 1557, because 2316 wasn't really fixed
5322 - better file I/O statistics gathering in /proc
5323 - don't take unnecessary, deadlock-inducing bug in readpage (2383)
5324 - another kernel patch to fix zero-copy TCP function export
5325 - don't take duplicate lock when processing re-sent getattr (2420)
5326 - lctl uses obd_self_export instead of creating new conn (2353)
5327 - MDS/OST recovery case which requires object creation asserted (2425)
5328 - move lfs from /usr/sbin to /usr/bin in packages
5329 - fix race between mds_client_add and mds_client_free (2417)
5330 - use kmalloc instead of slabs in portals (2430)
5331 - don't create duplicate records when a failover MDS is present (2442)
5332 - remove unnecessary mount age check (2332)
5333 - don't remove directory inodes from locks prematurely (2451)
5334 - don't break if MDS service name is the same as hostname (2103)
5335 - fix races in client write RPC generation when cache full (2482)
5337 2003-12-13 Cluster File Systems, Inc. <info@clusterfs.com>
5340 - remove now-unused request->rq_obd (278)
5341 - if an allocation fails, print out how much memory we've used (1933)
5342 - use PORTAL_SLAB_ALLOC for structures, to get GFP_MEMALLOC (1933)
5343 - add the "configurable stack size" patch to most series files (1256)
5344 - ability to write large log records, for 100+ OST configs (2306)
5345 - fix NULL deref when filter_prep fails (2314)
5346 - fix operator precedence error in filter_sync
5347 - dynamic allocation of socknal TX descriptors (2315)
5348 - fix a missed case in the GFP_MEMALLOC patch, can cause deadlock (2310)
5349 - fix gcc 2.96 compilation problem in xattr kernel patch (2294)
5350 - ensure that CWARN messages in Portals always get to the syslog
5351 - __init/__exit are not for prototype decls (ldlm_init/exit)
5352 - x86-64 compile warning fixes
5353 - fix gateway LMC keyword conflict (2318)
5354 - fix MDS lock inversions in getattr/reint paths (1844)
5355 - fix a rare lock re-ordering bug, which caused deadlock (2322)
5356 - fix i_sem/journal inversion in fsfilt_ext3_write_record (2306)
5357 - DLM race condition prevented some lock evictions (2328)
5358 - ENOMEM detection and retry on socknal sends (2230)
5359 - use GFP_NOFS throughout Lustre, to combat ENOMEM (2230)
5360 - move osc_rpcd into ptlrpc, for use in MDC and others (2329)
5361 - protect MDS inode fsdata with stronger locking; fixes assertion (2313)
5362 - better error messages when a client is rejected during recovery (1505)
5363 - avoid cancelling locks which were never granted, after failure (2330)
5364 - fix i_sem/journal inversion in mds_client_add (2333)
5365 - fix truncate/getattr lock cycle deadlock (2334)
5366 - use rpcd to send close; allows resend after timeout, avoid leak (1897)
5367 - fix two rare exit paths which could leak an l_lock() ref (2321)
5368 - fencepost error in MDS/OST orphan recovery (2226)
5369 - make log record alignment 8 bytes (1988)
5370 - lstripe now fails when requested offset > ost_count (2237)
5371 - ensure that all kernel series have a complete list.h (1607)
5372 - fix crashes in special-file operations (2316)
5373 - lctl create/brw OID mismatch, caused by obsolete filter loop (2339)
5375 - allow configurable automake binary, for testing new versions
5376 - small update to the lfs documentation
5378 2003-12-03 Cluster File Systems, Inc. <info@clusterfs.com>
5380 * fix negative export reference count in fsfilt_sync (2312)
5382 2003-12-01 Cluster File Systems, Inc. <info@clusterfs.com>
5383 * release candidate 0.9.1
5385 - orphans are moved into the PENDING directory for possible recovery
5386 - replayed opens now open by fid for orphan/rename safety (1042)
5387 - last close of an orphan inode generates a transno (683)
5388 - chdir() and mount() now pin the directory entry (1020)
5389 - avoid CERROR in normal ll_setattr_raw() error case (1500)
5390 - discard very old requests without processing them (1502)
5391 - remove some common, well-understood CERRORs (1505)
5392 - require O_DIRECT I/O to be page-sized to workaround IA64 crash (1609)
5393 - clear "grant" flags in OST replies until OST grant code lands (1644)
5394 - fix read performance by not clobbering i_blksize on client (1598)
5395 - fix __ldlm_handle2lock oops by not dereferencing lock after PUT (1625)
5396 - make LRU size a /proc tunable, clears locks when reduced (707)
5397 - fix some lprocfs rot that prevented ptlbd from loading (1732)
5398 - server locks take references on exports now (1558)
5399 - build fixes for 2.4.20-rh trees (1663)
5400 - return an error from lov_create if all OSCs are inactive (1751)
5401 - fix import levels when a reconnect happens without a timeout (1597)
5402 - exit early from mds_open if we get a lookup error (1749)
5403 - partial page read at EOF wouldn't wait for disk before sending (1642)
5404 - avoid NULL deref in obdfilter when reading page past EOF (1592)
5405 - avoid LASSERT in ll_intent_lock if server failed very early (1090)
5406 - fix LBUG in ll_it_open_error with rc = -2 (1861)
5407 - write/truncate lock inversion (1639)
5408 - Don't auto-load obdclass, portals modules during cleanup (1495)
5409 - fix timestamps from jumping to "now" (1763)
5410 - extra journal assertions (1648)
5411 - add an extra multiunlink test (1771)
5412 - fix read_record/write_record API (1776)
5413 - fix leak of offset_extent, possible incorrect i_size later (1772)
5414 - fix lasserts in mis-matched transnos during open-unlink testing (1541)
5415 - Debugging for the kqswnal_get_idle_tx problems (1820)
5416 - Allow recovery to be attempted multiple times (1536)
5417 - Write out MDS last_rcvd file after it is first created (1600)
5418 - Fix tx_descriptor leak in failed transmit situations (1827)
5419 - ext3 journaling fixes for assertion failure after IO error (1871)
5420 - class_export_put() on freed export after completion AST error (1896)
5421 - Fix revalidate looping in VFS (1322)
5422 - Don't access a freed export during MDS_REINT timeout (1521)
5423 - Add open-unlink recovery support on the MDS (1673,1764)
5424 - Return an error if no MDS data was read from last_rcvd (1946)
5425 - Fix for lookup "." or ".." crash on error (1932,1931,1935)
5426 - Don't setup a disk device that doesn't match exported UUID (317)
5427 - Reduce bulk RPC timeout to avoid cascading client/OST failures (1845)
5428 - avoid committing NULL handle in force close
5429 - local.sh is now a one-stripe LOV configuration
5430 - POSIX utime.4 -EPERM on FIFO not owned by user (56)
5431 - fix ext3 htree duplicate directory entry corruption (1516)
5432 - POSIX creat.13, fstat.1, open.18, stat.3 new file atime/mtime (2020)
5433 - update to new LOV EA format (2097)
5434 - interoperability for different PAGE_SIZE/wordsize (686,1821,1343,2042)
5436 2003-06-15 Phil Schwan <phil@clusterfs.com>
5439 - imports and exports cleanup too early, need refcounts (349, 879, 1045)
5440 - per-import/export recovery handling (958, 931, 959)
5441 - multiple last-rcvd slots, for serving multiple FSes (949)
5442 - connections are again shared between multiple imp/exports (963, 964)
5443 - "umount -f" would hang if any requests needed to be sent (393, 978)
5444 - avoid pinning large req buffer by copying for queued messages (989)
5445 - add "uuid" to "lctl device" command to help upcalls (991)
5446 - "open" RPCs with transnos would confuse recovery counters (1037)
5447 - do proper endian conversion of all wire messages (288, 340, 891)
5448 - remove OST bulk get LBUGs, fix ost_brw_write cleanup (1126)
5449 - call waiting locks callback from LDLM recovery thread (1127, 1151)
5450 - fix ptlrpc_connection leak in target_handle_connect (1174)
5451 - fix import refcounting bug in OST and MDS cleanup (1134)
5452 - if an invalid-at-open-time OSC returned before close(), LBUG (1150)
5453 - fix very unlikely obd_types race condition (501)
5454 - remove osc_open hack for echo_client (1187)
5455 - we leaked exports/dlmimps for forcibly disconnected clients (1143)
5456 - a failure in read_inode2 leads to deadlock (1139)
5457 - cancel ack-locks as soon as transaction is committed (1072)
5458 - fix major leaks and crashes in the bulk I/O path (937, 1057)
5459 - make sure to commitrw after any preprw to avoid deadlock (1162)
5460 - failing to execute a file in a lustre FS would lock inode (1203)
5461 - small DEBUG_REQ fix to avoid dereferencing a NULL (1227)
5462 - don't ASSERT while cleaning up an incompletely-setup obd (1248)
5463 - obd_uuid2tgt would walk off the end of the list (1255)
5464 - on IA64 the osc would give portals incorrect bulk size (1258)
5465 - fix debug daemon ioctl interface; allows daemon on ia64 (1274)
5466 - fix lock inversion caused by new llite matching code (1282)
5467 - limit the number of dirty pages on a client to 10MB (1286)
5468 - timed out locks were not being corrected cancelled (1289)
5469 - fix O_DIRECT above 4GB on IA-32 (1292)
5470 * major user-visible changes
5471 - fail out/fail over policy now controlled by the upcall (993)
5473 - add OBD_PING to check server availability and failure (954)
5474 - lustre messages are now sent in sending host order (288, 340, 891)
5475 - add eadatalen to MDS getattr reply (340)
5476 - OST read replies may contain second buffer, with per-page status (593)
5478 2003-03-11 Phil Schwan <phil@clusterfs.com>
5481 - LDLM_DEBUG macro fix, for gcc 3.2 (850)
5482 - failed open()s could cause deadlock; fixed (867, 869)
5483 - stop cancelling OST locks when files are closed (481)
5484 - overlapping XID spaces caused network corruption (851, 853)
5485 - fix unsafe fsfilt counter arithmetic; change to atomic_t
5486 - setattr_raw added, to do single-RPC, server-side setattrs
5487 - lmc/lconf syntax change for OST UUIDs
5488 - fix crashy race condition between ptlrpc_free_req and osc_close
5489 - don't use request in mdc_enqueue if we hit a timeout (889)
5490 - don't set the inode i_size for regular files from the MDS (896)
5491 - handle out of order completion AST (842)
5492 - don't LBUG if a lock request times out after receiving AST (913)
5493 - avoid d_rehash race in ll_find_alias by rehashing inside dcache_lock
5494 - if a bad lock AST arrives, send an error instead of dropping entirely
5495 - return 0 from revalidate2 if ll_intent_lock returns -EINTR (912)
5496 - fix leak in bulk IO when only partially completed (899, 900, 926)
5497 - fix O_DIRECT for ia64 (55)
5498 - (almost) eliminate Lustre-kernel-thread effects on load average (722)
5499 - C-z after timeout could hang a process forever; fixed (977)
5501 - client-side I/O cache (678, 924, 929, 941, 970)
5503 - READPAGE and SETATTRs which don't take server-side locks get
5506 2003-02-11 Phil Schwan <phil@clusterfs.com>
5509 - Fix ldlm_lock_match on the MDS to avoid matching remote locks (592)
5510 - Fix fsfilt_extN_readpage() to read a full page of directory
5511 entries, or fake the remainder if PAGE_SIZE != blocksize (500)
5512 - Avoid extra mdc_getattr() in ll_intent_lock when possible (534, 604)
5513 - Fix imbalanced LOV object allocation and out-of-bound access (469)
5514 - Most intent operations were removed, in favour of a new RPC mode
5515 that does a single RPC to the server and bypasses most of the VFS
5516 - All LDLM resource ID arrays were removed in favour of ldlm_res_id
5517 - Aggressively cancel local locks on DLM servers
5518 - mds_reint_unlink sends EA to the client if it's the last nlink.
5519 client uses that EA to unlink OST objects.
5520 - mds_reint_{rename,unlink,link} were rewritten to take ordered locks
5521 - recursive symlinks were fixed (439)
5522 - fixed NULL deref in DEBUG_REQ
5523 - filter_update_lastobjid no longer calls sync, which annoyed extN
5524 - fixed multi-client small-writes to a single file problem (445)
5525 - fixed mtime updates during file writes (607)
5526 - fixed vector writes on obdfilter causing problems when ENOSPC (670)
5527 - fixed bug in obd_brw_read/write() (under guise of testing 367)
5528 - fixed Linux OST size reporting problem (444, 656)
5529 - OST now updates object mtime with writes or setattr (607, 619)
5530 - client verifies file size before zeroing page past EOF (445)
5531 - OST now writes last allocated objid to disk with allocation (108)
5532 - LOV on echo now works (409)
5534 - mds_reint_unlink sends a new buffer, with the EA included. this
5535 buffer is only valid if body->valid & OBD_MD_FLEASIZE, which is only
5536 set if a regular file was being unlinked, and it was the last link
5537 - use PtlGet from the target for bulk writes (315)
5538 - OST now updates object mtime with writes or setattr (607, 619)
5539 - LDLM now has a grant-time callback to revalidate locked items, if
5541 - Many MDS operations were reorganized to combat race conditions
5543 - Merge b_intel branch (updated lprocfs code) - now at /proc/fs/lustre
5544 - configure check to avoid gcc version 2.96 20000731-2.96-98) (606)
5546 2003-01-06 Andreas Dilger <adilger@clusterfs.com>
5549 - Fully reactivate OST imports after reconnection (512, others)
5550 - Make sure client sees our -ENOTCONN from mds_handle (513 - partial)
5551 - More graceful error handling for truncating on dead OST (515)
5552 - Don't error out unless we're actually accessing dead stripes (474)
5553 - Fix garbage sizes when stripes are missing (410)
5554 - LRU counters were broken, causing constant lock purge (433, 432)
5555 - garbage on read from stripes with failed OSTs (441)
5556 - mark OSCs as active before reconnecting during recovery (438)
5557 - lov_enqueue and lov_cancel need to handle inactive OSTs (403)
5558 - lfind did not preserve OST order in output (443)
5559 - symlinks cause hung clients, incorrect data (439)
5560 - stop dereferencing request after dropping refcount (457)
5561 - don't LASSERT(spin_is_locked) on non-SMP (455)
5562 - fixes for many rename() bugs
5563 - fstat didn't correctly synchronize attributes (399)
5564 - server must handle lock cancellation during blocking AST prep (487)
5565 - bulk descriptors were free()d too soon (511)
5566 - fix paths in lconf, which would load incorrect modules (451, 507)
5567 - fix confusing lconf 'host not found' error message (386)
5568 - fix lock order deadlock on OST (O/R i_sem before journal ops, 478)
5569 - fix race condition in mdc_blocking_ast() for inode access (526)
5570 - fix lov_unpackmd() unpacking wrong number of stripes (537)
5571 - fix lov_set_osc_active() marking wrong OSC inactive (440)
5572 - fix bad lstripe lov_unpackmd() assertion (fix layering too) (527)
5573 - fix multiple writes of stripe MD to MDS (358, maybe 519)
5574 - fix lstripe in several ways (kernel side) (527)
5575 - fix request leak in ldlm_cli_enqueue (262)
5576 - incorrect OSC was marked inactive after OST failure
5577 - call mds_fs_cleanup before unmounting filesystem (524)
5578 - fix races between taking ns_lock and ldlm_lock_change_resource
5579 - fix races updating LOV export open file list
5580 - fix lov_enqueue error path, avoid decref-ing bad lock handle
5581 - fix recovery NULL deref in ldlm_cli_cancel_unused
5582 - fix some DLM races by using new hash table for lock handles (419)
5583 - permit the client to specify desired inodes, at replay
5584 - duplicate requests when we queue them for replay reintegration
5585 - fix last_rcvd offset calculation
5586 - sync after each recovered transaction, so we always make progress
5587 - never, not always, ERESTART requests without transnos
5588 - store the lov_desc in the MDS, so we don't depend on getlovinfo to
5590 - skip replay if the MDS says that the client is already connected
5591 - don't check for a recovery-enabled export to match lctl's UUID
5592 - don't INC_USE_COUNT for phantom exports
5593 - don't crash when cleaning up phantom exports (567)
5594 - don't double-finish or set replay data for errored mdc_open requests
5595 - abort requests when they time out, so we don't get old replies
5596 - send/receive replies for AST messages again
5597 - if the client says that it doesn't have the lock, cancel it on the
5599 - if we timeout during I/O, don't try to cancel an in-use lock; instead
5600 mark it as destroyed, it will all work out when decref is called
5601 - fix module use counts (22, 581)
5603 - ASTs now expect a reply (server cancels lock on error reply)
5605 2002-12-02 Andreas Dilger <adilger@clusterfs.com>
5608 - fix many simultaneous client startup (392)
5609 - fix dentry->d_it clobbering
5610 - credentials weren't being shipped for readdir/getattr operations
5611 - remove invalid assertions triggered during some concurrent MD
5613 - proper Lustre versions added (336, 389)
5614 - fix memory leak for create error case (398)
5615 - fix LOV locking bug that would get cli/srv out of sync
5616 - fix echo client over LOV (409)
5617 - fix dbench 2, extN refcount problem (170, 258, 356, 418)
5618 - fix double-O_EXCL intent crash (424)
5619 - avoid sending multiple lock CANCELs (352)
5621 - MDS can do multi-client recovery (modulo bugs in new code)
5623 - many updates, edits, cleanups
5625 2002-11-18 Phil Schwan <phil@clusterfs.com>
5628 - fix null d_it dereference (346)
5629 - fix full OST/dbench hang (333)
5630 - fix permission problem with file removal (286)
5631 - fix removal of OSCs from LOV when they fail
5632 - fix NULL deref during bulk timeout (214)
5633 - fix problems related to multiple filesystems on one MDS (241)
5634 - fixed serious subtle metadata locking bugs
5635 - free locks on clients when inodes are removed due to memory
5637 - fix inode pointer in lock data (285)
5638 - partial support for multiple MDS on a single host (241)
5639 - data locks weren't cancelled at clear_inode time (290, 311)
5640 - intent locks could lead to unbounded lock growth (205)
5641 - added a maximum lock count, an LRU list, and a flusher
5642 - fix multiple rename (365)
5643 - properly abstracted the echo client
5644 - OSC locked 1 byte too many; fixed
5645 - rewrote brw callback code:
5646 - fixed recovery bugs related to LOVs (306)
5647 - fixed too-many-pages-in-one-write crash (191)
5648 - fixed (again) crash in sync_io_timeout (214)
5649 - probably fixed callback-related race (385)
5651 - Add capability to MDS protocol
5652 - LDLM cancellations and callbacks on different portals
5654 2002-10-28 Andreas Dilger <adilger@clusterfs.com>
5657 - limit client IOV size to PTL_MD_MAX_IOV (611336, 191)
5658 - defer open object destruction to close time (601981, 138)
5659 - open/close OST file handle in obdo (OBD_MD_FLHANDLE) (601981, 138)
5660 - move LDLM_ENQUEUE/CONVERT back to MDS portal (625069)
5661 - abstract ll_lookup2, fix ll_revalidate2 to use abstraction (256)
5662 - don't call obd_setattr in ll_file_release for destroyed objects
5663 * protocol change to lustre_msg: move |version| and add |flags|
5664 * protocol change to osc_punch: "start" in "o_size", "end" in "o_blocks"
5665 * lock replay: for LDLM_FL_REPLAY trust client to do right thing
5666 * added replay of create, unlink, link and rename operations during
5667 MDS failover; recovery should be much more robust now
5668 * remove failed OSCs from LOVs (only lov_create uses this so far)
5669 * the lustre-HOWTO was brought (more) up to date (582544)
5671 2002-10-23 Phil Schwan <phil@clusterfs.com>
5674 - in-use dentries weren't being reused properly (617851)
5675 - prevent multiple LDLM setup (599178)
5676 - fix LOV size calculations for truncate (617853)
5677 - fix client handling of MDS intent errors (POSIX)
5678 - fix permission bug in lovstripe.c test (624321)
5679 - fix MDS thread deadlock - move LDLM handler to DLM portal (625069)
5680 - truncate past end of file could corrupt data
5681 - proper cleanup after timeouts, crashes, etc (592524, 550815)
5682 - a race in recovery could return ETIMEDOUT to apps (623947)
5683 - building outside the source directory was fixed
5684 * the lustre-HOWTO was brought (more) up to date (582544)
5685 * major progress was made on recovery functionality
5687 2002-10-10 Phil Schwan <phil@clusterfs.com>
5690 - recovery deadlock fix
5691 - rm -rf causes LBUG fix (617817)
5692 - file open by multiple tasks fix (618962)
5693 - directory permissions bugs (602707 and 620007)
5694 - journal_stop fixed with locking (611313)
5695 - O_APPEND failures resolved (618273, perhaps 614459)
5696 - lconf PATH fix (619770)
5697 - IA64 build fix (621450)
5698 - RPC buffer sizes scale with amount of memory
5700 2002-10-01 Phil Schwan <phil@clusterfs.com>
5703 - locks would be cancelled without throwing away data pages,
5704 resulting in inconsistent data (605627)
5705 - inode attributes were not always being refreshed (605627, 612449)
5706 - lconf now continues to cleanup after lctl reports an error
5707 - MDS now enforces user permissions (602707)
5708 - lprocfs cleanup fixed, but not yet enabled (614157)
5709 - fixed infinite server hang, should a client not respond to an AST
5710 - avoid going into recovery if user calls readlink() with a buffer
5711 that's too small (613941)
5712 - AST RPCs no longer require replies (614867) -- this may be changed
5713 - don't crash server if client sends an IOV that's too big (611336)
5714 - fixed lock conversion deadlock (611892)
5715 - fixed the following of symlinks (614622)
5716 * recovery: the server can remove locks from a client that dies, other
5717 clients can make progress
5718 * more extN patch fixes
5719 * compile-time configurable ptlrpc buffer allocations
5721 - collaborative read cache document
5722 - Lustre Lite Performance CDR document-in-progress
5724 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
5727 - fix typo in patch-2.4.18
5729 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
5732 - clear ptlrpc request each time in handle_incoming_request()
5733 - unlink of files now destroys the object on the OST
5735 2002-09-19 Peter Braam <braam@clusterfs.com>
5737 * add hard link support
5738 * change obdfile creation method
5739 * kernel patch changed
5741 2002-09-19 Peter Braam <braam@clusterfs.com>
5744 - stack overflow bug in extN fixed
5746 2002-09-18 Andreas Dilger <adilger@clusterfs.com>
5748 * documentation updates
5749 - add man pages for config tools
5750 - update tests/README to describe testing with new config tools
5751 - finish metadata API descriptions
5752 * bug fixes and cleanups
5753 - statfs workaround for 16TB limit
5754 - LOV stripe allocation improved, can stripe on subset of OSTs
5755 - LOV file size/IO offset was wrong for files > 4GB in size
5756 - object EA data was being dropped, caused files to be unreadable
5757 - memory overflow with non-LOV OST caused memory corruption
5758 - fixed regression tests to work with new config tools, obdfilter
5759 - fixed bug when directory size became larger than 1 block
5760 - fixed bug (for single client case) when PWD was deleted
5761 - invalidate local directory pages when doing intent-based ops
5762 - avoid LDLM oops when lock callback contained bad data
5764 2002-09-09 Andreas Dilger <adilger@clusterfs.com>
5766 * documentation updates
5767 * bug fixes and cleanups
5768 - configuration tools
5771 - 64-bit compile warnings
5772 - 64-bit internal statfs data
5774 * test_brw on persistent OST devices
5776 * lprocfs (disabled)
5778 2002-09-04 Andreas Dilger <adilger@clusterfs.com>
5780 * documentation updates
5781 * bug fixes and cleanups
5782 * configuration tools
5784 2002-08-30 Peter J. Braam <braam@clusterfs.com>
5787 * many small fixes to 0_5_4
5788 * io/network handling
5789 * thinkos in MDS operations
5791 2002-08-24 Peter J. Braam <braam@clusterfs.com>
5794 * crucial basic fixes to 0.5.3
5795 * IOR, Iozone work over Elan
5798 2002-08-07 Phil Schwan <phil@clusterfs.com>
5799 * version 0_5_3, our first alpha
5800 * we use the new Portals iovs
5801 * documentation updates
5802 * bug fixes and cleanups
5803 * small changes in the DLM wire protocol
5805 2002-07-25 Peter J. Braam <braam@clusterfs.com>
5806 * version 0_5_1 with some initial stability,
5807 * locking on MD and file I/O.
5808 * documentation updates
5809 * several bug fixes since 0.5.0
5810 * small changes in wire protocol
5812 2002-07-18 Phil Schwan <phil@clusterfs.com>
5814 * delivered as Lustre Light Alpha
5815 * fixed a crash after handling invalid MDS requests
5816 * fixed directory pages for architectures with non-4k pages sizes
5818 2002-07-11 Andreas Dilger <adilger@clusterfs.com>
5819 * release version v0_4_4
5820 * Moves TCP acceptor to be on port 2432 (unused Coda port) instead
5822 * Fixes a number of interruption problems with OST operations.
5823 * Update documentation for portals header changes
5824 * Move all wire protocol structs/defines to lustre_idl.h
5825 * Fixes symlink length bug.
5826 * Add tcpdump to repository.
5828 2002-07-05 Andreas Dilger <adilger@clusterfs.com>
5829 * release version v0_4_3
5830 * Fixes statfs for inodes on extN.
5831 * Fixes bug in runtests which would delete /etc/hosts.
5832 * Use 64-bit object IDs wherever possible (not into VFS though)
5833 Remove ost_get_info, which is unused by lustre, and out of date.
5835 2002-07-03 Peter Braam <braam@clusterfs.com>
5836 * release version v0_4_2 Fixes a lookup error (type not passed)
5837 * move forward to head of Portals
5838 * move forward to latest Lustre kernel
5840 2002-06-25 Peter Braam <braam@clusterfs.com>
5841 * release version v0_4_1. Hopefully stable on single node use.