1 tbd Sun Microsystems, Inc.
4 2.6.16.60-0.33 (SLES 10),
5 2.6.18-128.1.1.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: add lazystatfs mount option to allow statfs(2) to skip down OSTs
19 Details : allow skip disconnected ost for send statfs request and hide error
25 Description: fix racy locking of mballoc block bitmaps causing BUG
26 Details : The locking of the mballoc buddy bitmap and the in-memory
27 block bitmap was using two different spin locks in some
28 cases. This made it possible to incorrectly access the
29 mballoc bitmap while another process was modifying it,
30 causing a sanity assertion to fail. While no on-disk corruption
31 was reported, there was some risk of this happening.
34 Frequency : rare, on llog test 6
36 Descriptoin: don't allow connect to already connected import
37 Details : allowing connect to already connected import is hide connecting problem.
42 Descriptoin: allow kill process which wait statahead result
43 Details : for some reasons 'ls' can stick in waiting result from statahead,
44 in this case need way for kill this process.
46 Severity : enhancement
48 Description: Add state history info file, enhance import info file
49 Details : Track import connection state changes in a new osc/mdc proc file;
50 add overview-type data to the osc/mdc import proc file.
52 Severity : enhancement
54 Description: MDS create should not wait for statfs RPC while holding DLM lock.
56 Severity : enhancement
58 Description: Update to RHEL5U3 kernel-2.6.18-128.1.1.el5.
63 Descriptoin: OST grant too much space to client even there are not enough space.
64 Details : Client will shrink its grant cache to OST if there are no write
65 activity over 6 mins (GRANT_SHRINK_INTERVAL), and OST will retrieve
66 this grant cache if there are already not enough avaible space
67 (left_space < total_clients * 32M).
70 Frequency : start MDS on uncleanly shutdowned MDS device
72 Descriptoin: ll_sync thread stay in waiting mds<>ost recovery finished
73 Details : stay in waiting mds<>ost recovery finished produce random bugs
74 due race between two ll_sync thread for one lov target. send
75 ACTIVATE event only if connect realy finished and import have
79 Frequency : rare, connect and disconnect target at same time
81 Descriptoin: ASSERTION(atomic_read(&imp->imp_inflight) == 0
82 Details : don't call obd_disconnect under lov_lock. this long time
83 operation and can block ptlrpcd which answer to connect request.
88 Descriptoin: don't lose wakeup for imp_recovery_waitq
89 Details : recover_import_no_retry or invalidate_import and import_close can
90 both sleep on imp_recovery_waitq, but we was send only one wakeup
94 Frequency : always with long access acl
96 Descriptoin: mds can't pack reply with long acl.
97 Details : mds don't control size of acl but they limited by reint/getattr
100 Severity : enhancement
102 Description: Update to SLES10 kernel-2.6.16.60-0.33.
104 Severity : enhancement
106 Description: Update to RHEL5 kernel-2.6.18-92.1.22.el5.
109 Frequency : start MDS on uncleanly shutdowned MDS device
111 Descriptoin: aborting recovery hang on MDS
112 Details : don't throttle destroy RPCs for the MDT.
115 Frequency : on remount
117 Description: external journal device not working after the remount
118 Details : clear dev_rdonly flag for external journal devices in
124 Description: shutdown vs evict race
125 Details : client_disconnect_export vs connect request race.
126 if client will evicted at this time - we start invalidate
127 thread without referece to import and import can be freed
131 Frequency : rare, need acl's on inode.
133 Description: client can't handle ost additional correctly
134 Details : if ost was added after client connected to mds client can have
135 hit lnet_try_match_md ... to big messages to wide striped files.
136 in this case need teach client to handle config events about add
137 lov target and update client max ea size at that event.
139 Severity : enhancement
141 Description: Changelogs
142 Details : Changelogs are a lightweight mechanism to track filesystem
143 metadata and namespace changes. The changelog is recorded
144 permanently on the MDTs, and is periodically "consumed" / purged
145 when records are no longer needed.
147 Severity : enhancement
149 Description: compact fld format with extents
150 Details : Store range of seq rather than every seq in FLD. Seq
151 controller update FLD rather than clients. In Case of CMD, mdt0
152 has FLD, all other metadata server act as non persistent proxy
153 for FLD queries and cache fld entries in fld cache.
158 Description: don't skip ost target if they assigned to file
159 Details : Drop slow OSCs if we can, but not for requested start idx.
160 This means "if OSC is slow and it is not the requested
161 start OST, then it can be skipped, otherwise skip it only
162 if it is inactive/recovering/out-of-space.
166 Description: more cleanup in mds_lov
167 Details : not send LOV EA under replay, we can't know about they size at this
168 time. Don't allow client connect to mds before any ost connected,
169 for avoid problems with LOV EA size and returning EIO to client.
171 Severity : enhancement
173 Description: Interoperability at server side (Disk interoperability)
175 Severity : enhancement
177 Description: Update to RHEL5 kernel-2.6.18-92.1.17.el5.
179 Severity : enhancement
181 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.31.
183 Severity : enhancement
185 Description: New client IO stack (CLIO).
187 Severity : enhancement
189 Description: Commit on sharing. Eliminate inter-client dependencies between
190 uncommitted transactions by doing transaction commits.
191 Thereby clients may recovery independently.
194 Frequency : Create a symlink file with a very long name
196 Description: ldlm_cancel_pack()) ASSERTION(max >= dlm->lock_count + count)
197 Details : If there is no extra space in the request for early cancels,
198 ldlm_req_handles_avail() returns 0 instead of a negative value.
200 Severity : enhancement
202 Description: Add /proc entry for import status
203 Details : The mdc, osc, and mgc import directories now have
204 an import directory that contains useful import data for debugging
207 Severity : enhancement
209 Description: Re-disable certain /proc logging
210 Details : Enable and disable client's offset_stats, extents_stats and
211 extents_stats_per_process stats logging on the fly.
214 Frequency : Only on FC kernels 2.6.22+
216 Description: oops in statahead
217 Details : Do not drop reference count for the dentry from VFS when lookup,
218 VFS will do that by itself.
220 Severity : enhancement
222 Description: Generic /proc file permissions
223 Details : Set /Proc file permissions in a more generic way to enable non-
224 root users operate on some /proc files.
228 Description: Hitting mdc_commit_close() ASSERTION
229 Details : Properly handle request reference release in
230 ll_release_openhandle().
234 Description: quota recovery deadlock during mds failover
235 Details : This patch includes att18982, att18236, att18237 in bz14840.
237 1. fix osts hang when mds does failover with quotaon
238 2. prevent watchdog storm when osts threads wait for the
243 Frequency : only patchless client
244 Description: add workaround for race between add/remove dentry from hash
246 Severity : enhancement
248 Description: Allow OST glimpses to return PW locks
252 Description: LBUG when llog conf file is full
253 Details : When llog bitmap is full, ENOSPC should be returned for plain log.
257 Description: Prevent import from entering FULL state when server in recovery
261 Description: service mount cannot take device name with ":"
262 Details : Only when device name contains ":/" will mount treat it as
268 Description: replace ptlrpcd with the statahead thread to interpret the async
269 statahead RPC callback
273 Frequency : on recovery
274 Description: I/O failures after umount during fail back
275 Details : if client reconnected to restarted server we need join to recovery
276 instead of find server handler is changed and process self
277 eviction with cancel all locks.
279 Severity : enhancement
281 Description: Update to RHEL5 kernel-2.6.18-92.1.10.el5.
283 Severity : enhancement
285 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.27.
287 Severity : enhancement
289 Description: Upcall on Lustre log has been dumped
290 Details : Allow for a user mode script to be called once a Lustre log has
291 been dumped. It passes the filename of the dumped log to the
292 script, the location of the script can be specified via
293 /proc/sys/lnet/debug_log_upcall.
298 Description: avoid idr_remove called for id which is not allocated.
299 Details : Move assigment s_dev for clustered nfs to end of initialization,
300 to avoid problem with error handling.
305 Description: avoid Already found the key in hash [CONN_UNUSED_HASH] messages
306 Details : When connection is reused this not moved from CONN_UNUSED_HASH
307 into CONN_USED_HASH and this prodice warning when put connection
308 again in unused hash.
311 Severity : enhancement
313 Description: File striping can now be set to use an arbitrary pool of OSTs.
315 Severity : enhancement
317 Description: Export bytes_read/bytes_write count on OSC/OST.
321 Description: Early reply size mismatch, MGC loses connection
322 Details : Apply the MGS_CONNECT_SUPPORTED mask at reconnect time so
323 the connect flags are properly negotiated.
328 Description: quotas are not honored with O_DIRECT
329 Details : all writes with the flag O_DIRECT will use grants which leads to
330 this problem. Now using OBD_BRW_SYNC to guard this.
334 Description: add quota statistics
335 Details : 1. sort out quota proc entries and proc code.
336 2. add quota statistics
338 Severity : enhancement
340 Description: enable quota support for HEAD.
344 Description: Properly propagate oinfo flags from lov to osc for statfs
345 Details : restore missing copy oi_flags to lov requests.
347 Severity : enhancement
349 Description: Add man pages for llobdstat(8), llstat(8), plot-llstat(8),
350 : l_getgroups(8), lst(8), routerstat(8)
351 Details : included man pages for llobdstat(8), llstat(8),
352 : plot-llstat(8), l_getgroups(8), lst(8), routerstat(8)
354 Severity : enhancement
356 Description: configure's --enable-quota should check the
357 : kernel .config for CONFIG_QUOTA
358 Details : configure is terminated if --enable-quota is passed but
359 : no quota support is in kernel
363 Description: Remove portals compatibility
364 Details : Remove portals compatibility, not interoperable with releases
369 Description: Resolve device initialization race
370 Details : Prevent proc handler from accessing devices added to the
371 obd_devs array but yet be intialized.
373 Severity : enhancement
375 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.23.
377 Severity : enhancement
379 Description: Update to RHEL5 kernel-2.6.18-92.1.6.el5.
384 Description: Using wrong pointer in osc_brw_prep_request
385 Details : Access to array[-1] can produce panic if kernel compiled with
386 CONFIG_PAGE_ALLOC enabled
390 Description: Client runs out of low memory
391 Details : Consider only lowmem when counting initial number of llap pages
395 Description: *optional* service tags registration
396 Details : if the "service tags" package is installed on a Lustre node
397 When the filesystem is mounted, a local-node service tag will
398 be created. See http://inventory.sun.com/ for more information
399 about the Service Tags asset management system.
403 Description: Kernel BUG tries to release flock
404 Details : Lustre does not destroy flock lock before last reference goes
405 away. So always drop flock locks when client is evicted and
406 perform unlock regardless of successfulness of speaking to MDS.
410 Description: add refcount protection for osc callbacks, avoid panic on shutdown
414 Description: sanity test 65a fails if stripecount of -1 is set
415 Details : handle -1 striping on filesystem in ll_dirstripe_verify
420 Description: ASSERTION(CheckWriteback(page,cmd)) failed
421 Details : badly clear PG_Writeback bit in ll_ap_completion can produce false
424 Severity : enhancement
426 Description: Update to RHEL5 kernel-2.6.18-53.1.21.el5.
430 Description: do not process already freed flock
431 Details : flock can possibly be freed by another thread before it reaches
432 to ldlm_flock_completion_ast.
436 Description: LBUG during stress test
437 Details : Need properly lock accesses the flock deadlock detection list.
441 Description: oops in page fault handler
442 Details : kernel page fault handler can return two special 'pages' in error
443 case, don't try dereference NOPAGE_SIGBUS and NOPAGE_OMM.
447 Description: timeout with invalidate import.
448 Details : ptlrpcd_check call obd_zombie_impexp_cull and wait request which
449 should be handled by ptlrpcd. This produce long age waiting and
450 -ETIMEOUT ptlrpc_invalidate_import and as result LASSERT.
452 Severity : enhancement
454 Description: Update to RHEL5 kernel-2.6.18-53.1.19.el5.
458 Description: enable MGS and MDT services start separately
459 Details : add a 'nomgs' option in mount.lustre to enable start a MDT with
460 a co-located MGS without starting the MGS, which is a complement
461 to 'nosvc' mount option.
465 Frequency : after recovery
466 Description: precreate to many object's after del orphan.
467 Details : del orphan st in oscc last_id == next_id and this triger growing
468 count of precreated objects. Set flag LOW to skip increase count
469 of precreated objects.
473 Frequency : rare, on clear nid stats
474 Description: ASSERTION(client_stat->nid_exp_ref_count == 0)
475 Details : when clean nid stats sometimes try destroy live entry,
476 and this produce panic in free.
480 Description: Stack overflow during MDS log replay
481 ease stack pressure by using a thread dealing llog_process.
485 Description: wait until IO finished before start new when do lock cancel.
486 Details : VM protocol want old IO finished before start new, in this case
487 need wait until PG_writeback is cleared until check dirty flag and
488 call writepages in lock cancel callback.
490 Severity : enhancement
492 Description: using special macro for print time and cleanup in includes.
496 Description: mds_mfd_close() ASSERTION(rc == 0)
497 Details : In mds_mfd_close(), we need protect inode's writecount change
498 within its orphan write semaphore to prevent possible races.
502 Description: Obsolete CURRENT_SECONDS and use cfs_time_current_sec() instead.
506 Frequency : rare, on shutdown ost
507 Description: don't hit live lock with umount ost.
508 Details : shrink_dcache_parent can be in long loop with destroy dentries,
509 use shrink_dcache_sb instead.
513 Description: don't panic with use echo client
514 Details : echo client pass NULL as client nid pointer and this produce null
519 Description: fix build on ppc32
520 Details : compile code with -m64 flag produce wrong object file for ppc32.
524 Description: add message levels for liblustreapi
528 Description: fix for occasional failure case of -ENOSPC in recovery-small tests
529 Details : Move the 'good_osts' check before the 'total_bavail' check. This
530 will result in an -EAGAIN and in the exit call path we call
531 alloc_rr() which will with increasing aggressiveness attempt to
532 aquire precreated objects on the minimum number of required OSCs.
536 Description: Use old size assignment to avoid deadlock
537 Details : Reverts the changes in bugs 2369 and bug 14138 that introduced
538 the scheduling while holding a spinlock. We do not need locking
539 for size in ll_update_inode() because size is only updated from
540 the MDS for directories or files without objects, so there is no
541 other place to do the update, and concurrent access to such inodes
542 are protected by the inode lock.
546 Description: resolve "_IOWR redefined" build error on SLES10
550 Description: dump the memory debugging after all modules are unloaded to
551 suppress false negative in conf_sanity test 39
553 Severity : enhancement
555 Description: build kernel-ib packages for OFED 1.3 in our release cycle
560 Description: fix SLES kernel versioning
561 Details : the kernel version for our SLES 10 kernel did not include a "-"
562 before the "smp" at the end. while this was not a problem in
563 general, it did mean that software trying to use the kernel
564 version to try to detect a vendor specific kernel would fail.
565 this was most evident by the OFED build scripts.
569 Description: Don't update lov_desc members until making sure they are valid
570 Details : When updating lov_desc members via proc fs, need fix their
571 validities before doing the real update.
575 Description: don't put request into delay list while invalidate in flight.
576 Details : ptlrpc_delay_request sometimes put in delay list while invalidate
577 import in flight. this produce timeout for invalidate and sometimes
578 can cause stale data.
582 Frequency : on ppc only
583 Description: not convert ost objects for directory because it's not exist.
584 Details : ll_dir_getstripe assume dirrectory has ost objects but this wrong.
588 Description: Add FMODE_EXEC file flag for SLES10 SP1 kernel.
590 Severity : enhancement
592 Description: Update to support 2.6.22.14 vanilla kernel.
596 Frequency : rare, on recovery
597 Description: read procfs can produce deadlock in some situation
598 Details : Holding lprocfs lock which send rpc can produce block for destroy
599 obd objects and this also block reconnect with -EALREADY. This
600 isn't fix all lprocfs bugs - but make it rare.
602 Severity : enhancement
604 Description: Update kernel to RHEL5 2.6.18-53.1.14.el5.
607 Frequency : frequent on X2 node
609 Description: mdc_set_open_replay_data LBUG
610 Details : Set replay data for requests that are eligible for replay.
614 Description: lustre_mgs: operation 101 on unconnected MGS
615 Details : When MGC is disconnected from MGS long enough, MGS will evict the
616 MGC, and late on MGC cannot successfully connect to MGS and a lot
617 of the error messages complaining that MGS is not connected.
621 Frequency : on network error
622 Description: panic with double free request if network error
623 Details : mdc_finish_enqueue is finish request if any network error ocuring,
624 but it's true only for synchronus enqueue, for async enqueue
625 (via ptlrpcd) this incorrect and ptlrpcd want finish request
628 Severity : enhancement
630 Description: client-side metadata stat-ahead during readdir(directory readahead)
631 Details : perform client-side metadata stat-ahead when the client detects
632 readdir and sequential stat of dir entries therein
635 Frequency : on start mds
637 Description: Implement get_info(last_id) in obdfilter.
640 Frequency : occasional
642 Description: Correctly check stale fid, not start epoch if ost not support SOM
643 Details : open with flag O_CREATE need set old fid in op_fid3 because
644 op_fid2 was overwritten with new generated fid, but mds can answer
645 with one of these two fids and both is not stale. Setattr
646 incorrectly started an epoch and assume will be called
647 done_writing, but without SOM done_writing ever being called.
650 Frequency : rare, depends on device drivers and load
652 Description: MDS or OSS nodes crash due to stack overflow
653 Details : Code changes in 1.8.0 increased the stack usage of some functions.
654 In some cases, in conjunction with device drivers that use a lot
655 of stack the MDS (or possibly OSS) service threads could overflow
656 the stack. One change which was identified to consume additional
657 stack has been reworked to avoid the extra stack usage.
660 Frequency : occasional
662 Description: Do not fail import if osc_interpret_create gets -EAGAIN
663 Details : If osc_interpret_create got -EAGAIN it immediately exits and
664 wakeup oscc_waitq. After wakeup oscc_wait_for_objects call
665 oscc_has_objects and see OSC has no objests and call
666 oscc_internal_create to resend create request.
668 Severity : enhancement
670 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.5.
672 Severity : enhancement
674 Description: Update to RHEL5 latest kernel-2.6.18-53.1.13.el5.
677 Frequency : very rare
679 Description: Fix replay if there is an un-replied request and open
680 Details : In some cases, older replay request will revert the
681 mcd->mcd_last_xid on MDS which is used to record the client's
684 Severity : enhancement
686 Description: Update to RHEL5 latest kernel-2.6.18-53.1.6.el5.
688 Severity : enhancement
690 Description: Add rhel5 support to HEAD.
692 Serverity : enhancement
694 Description: Update RHEL4 kernel to 2.6.9-67.0.4.
699 Description: Don't allow skipping OSTs if index has been specified.
700 Details : Don't allow skipping OSTs if index has been specified, make
701 locking in internal create lots better.
705 Description: LBUG in ptlrpc_check_set() bad phase ebc0de00
706 Details : access to bitfield in structure is always rounded to long
707 and this produce problem with not atomic change any bit.
711 Description: Lustre make rpms failed.
712 Details : Remove ldiskfs spec file to avoids rpmbuild be confused when
713 builds Lustre rpms from tarball.
716 Frequency : rare on shutdown ost
718 Description: If llog cancel was not send before clean_exports phase, this can
719 produce deadlock in llog code.
720 Details : If llog thread has last reference to obd and call class_import_put
721 this produce deadlock because llog_cleanup_commit_master wait when
722 last llog_commit_thread exited, but this never success because was
723 called from llog_commit_thread.
727 Description: allow userland application know is lost one of stripes.
728 Details : fill lvb_blocks with error code on ost and return it to
729 application if error flag found.
733 Description: NULL lov_tgts causing MDS oops
734 Details : more safe checks for NULL lov_tgts for avoid oops.
736 Severity : enhancement
738 Description: Update to RHEL4 latest kernel-2.6.9-67.0.1.EL.
742 Descriptoin: make lov_create() will not stuck in obd_statfs_rqset()
743 Details : If an OST is down the MDS will hang indefinitely in
744 obd_statfs_rqset() waiting for the statfs data. While for
745 MDS QOS usage of statfs, it should not stuck in waiting.
747 Severity : enhancement
749 Description: remote_acl support
750 Details : Support ACL-based permission check for remote user.
751 Support setfacl/getfacl for remote user with the utils
752 "lfs {l,r}{s,g}etfacl" which follow the same parameter format as
753 the system "{s,g}etfacl" utils.
755 Severity : enhancement
757 Description: Update to RHEL4 U6 kernel-2.6.9-67.EL.
759 Severity : enhancement
761 Description: Update to RHEL5 latest kernel-2.6.18-53.1.4.el5.
765 Description: make mgs_setparam() handle fsname containing dash
766 Details : fsname containing a dash does not work with lctl conf_param
768 Severity : enhancement
770 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.3.
772 Severity : enhancement
774 Description: Update to RHEL5 Update-1 kernel 2.6.18-53.el5.
778 Frequency : rare, at shutdown
779 Description: access already free / zero obd_namespace.
780 Details : if client_disconnect_export was called without force flag set,
781 and exist connect request in flight, this can produce access to
782 NULL pointer (or already free pointer) when connect_interpret
783 store ocd flags in obd_namespace.
787 Frequency : only at startup
788 Description: not alloc memory with spinlock held.
789 Details : allocation memory with GFP_KERNEL can produce sleep deadlock,
790 if any spinlock held.
792 Severity : enhancement
794 Description: make lustre randomly fail allocating memory
795 Details : Make lustre randomly failed allocating memory for testing purpose.
797 Severity : enhancement
799 Description: lost problems with lov objid file
800 Details : Fixes some scability and access to not inited memory problems
801 in work with lov objdid file.
806 Description: lfs find does not continue on file error
807 Details : Continue other files processing when a file/dir is absent.
811 Description: Inconsistent usage of lustre_pack_reply()
812 Details : Standardize the usage of lustre_pack_reply() such that it
813 always generate a CERROR on failure.
816 Frequency : occasional
818 Description: MDS hang or stay in waiting lock
819 Details : If client receive lock with CBPENDING flag ldlm need send lock
820 cancel as separate rpc, to avoid situation when cancel request
821 can't processed due all i/o threads stay in wait lock.
825 Description: Update to RHEL5 kernel 2.6.18-8.1.15.el5.
829 Description: Update to SLES10 SP1 kernel 2.6.16.53-0.16
833 Description: Update to SLES9 kernel-2.6.5-7.287.3.
837 Description: Update to RHEL4 latest kernel.
839 Severity : enhancement
841 Description: Build SLES10 patchless client fails
842 Details : The configure was broken by run ./configure with
843 --with-linux-obj=.... argument for patchless client. When the
844 configure use --with-linux-obj, the LINUXINCLUDE= -Iinclude
845 can't search header adequately. Use absolute path such as
846 -I($LINUX)/include instead.
850 Description: interrupt oig_wait produce painc on resend.
851 Details : brw_redo_request can be used for resend requests from ptlrpcd and
852 private set, and this produce situation when rq_ptlrpcd_data not
853 copyed to new allocated request and triggered LBUG on assert
854 req->rq_ptlrpcd_data != NULL. But this member used only for wakeup
855 ptlrpcd set if request is changed and can be safety changed to use
860 Description: LASSERT_{REQ,REP}SWAB macros are buggy
861 Details : If SWAB_PARANOIA is disabled, the LASSERT_REQSWAB and
862 LASSERT_REPSWAB macros become no-ops, which is incorrect. Drop
863 these macros and replace them with their difinitions instead.
867 Description: Update kernel patches for SLES10 2.6.16.53-0.8.
868 Details : Update which_patch & target file for SLES10 latest kernel.
871 Description: Remove client patches from SLES 10 kernel.
872 Details : This causes SLES 10 clients to behave as patchless clients
873 even on a Lustre-patched (server) kernel.
875 Severity : enhancement
877 Description: self-adjustable client's lru lists
878 Details : use adaptive algorithm for managing client cached locks lru
879 lists according to current server load, other client's work
880 pattern, memory activities, etc. Both, server and client
881 side namespaces provide number of proc tunables for controlling
884 Severity : enhancement
886 Description: light-weight GSS support
887 Details : Support krb5n and krb5a mode, which keep Kerberos 5 authentication
888 and reduce performance overhead.
890 Severity : enhancement
892 Description: Linux keyring support
893 Details : Support using service of Linux keyring for Lustre GSS internal
894 context refresh/cache mechanism.
898 Description: Fix errors in lfs documentation
899 Details : Fixes man pages
903 Description: don't use GFP_* in generic Lustre code.
904 Details : Use cfs_alloc_* functions and CFS_* flags for code portability.
908 Description: obdclass is limited by single OBD_ALLOC(idarray)
909 Details : replace OBD_ALLOC/OBD_FREE with OBD_VMALLOC/OBD_VFREE
913 Description: warnings with build patchless client with vanila 2.6.19 and up
914 Details : change old ctl_table style and replace ctl_table/ctl_table_header
915 with cfs_sysctl_table_t/cfs_sysctl_table_header_t
919 Frequency : Only for SLES
920 Description: sanity_quota fail test_1
921 Details : There are multiple occurences of $TSTUSR in SLES's /etc/group
922 file, which makes TSTID[2] inunique.
926 Frequency : Only for SLES9
927 Description: Kernel patches for SLES9 2.6.5-7.286 kernel
928 Details : Update target/ChangeLog/which_patch .
932 Frequency : Only for test_10 in sanity-quota.sh
933 Description: a bug in quota test code
934 Details : checking a test flag in a wrong place
938 Frequency : Only for test_13 in sanity-quota.sh
939 Description: a bug in quota test script
940 Details : shouldn't check overfull allocated quota
944 Description: jbd statistics
945 Details : Port older jbd statistics patch for sles10
947 Severity : enhancement
949 Description: Add printing inode info into message about error in writepage.
954 Description: reply_lock_interpret crash due to race with it and lock cancel.
955 Details : Do not replay locks that are being cancelled. Do not reference
956 locks by their address during replay, just by their handle.
960 Frequency : When flocks are used.
961 Description: assertion failure in ldlm_cli_enquque_fini for non NULL lock.
962 Details : Flock locks might destroy just granted lock if it could be merged
963 with another existing flock, this is done in completion handler,
964 so teach ldlm_cli_enquque_fini that this is a valid case for
970 Description: Oops in read and write path when failing to allocate lock.
971 Details : Check if lock allocation failed and return error back.
975 Description: lstripe command fails for valid OST index
976 Details : The stripe offset is compared to lov->desc.ld_tgt_count
977 instead of lov->desc.ld_active_tgt_count.
981 Description: sanity.sh failed test 103
982 Details : RHEL mis-interpret setfacl "-X" param, so we won't test setfacl
987 Description: df doesn't work properly if diskfs blocksize != 4K
988 Details : Choose biggest blocksize of OST's as the LOV's blocksize.
992 Description: lfs find on -1 stripe looping in lsm_lmm_verify_common()
993 Details : Avoid lov_verify_lmm_common() on directory with -1 stripe count.
995 Severity : enhancement
997 Description: Adaptive timeouts
998 Details : RPC timeouts adapt to changing server load and network
999 conditions to reduce resend attempts and improve recovery time.
1003 Description: llapi_file_create() does not allow some changes
1004 Details : add llapi_file_open() that allows specifying the mode and
1005 open flags, and also returns an open file handle.
1009 Description: merge and cleanup kernel patches.
1010 Details : Remove mnt_lustre_list in vfs_intent-2.6-rhel4.patch.
1014 Description: Add journal checksum support.(Kernel part)
1015 Details : The journal checksum feature adds two new flags i.e
1016 JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT and
1017 JBD2_FEATURE_COMPAT_CHECKSUM. JBD2_FEATURE_CHECKSUM flag
1018 indicates that the commit block contains the checksum for
1019 the blocks described by the descriptor blocks. Now commit
1020 record can be sent to disk without waiting for descriptor
1021 blocks to be written to disk. This behavior is controlled
1022 using JBD2_FEATURE_ASYNC_COMMIT flag.
1026 Description: OSS needs mutliple precreate threads
1027 Details : Add ability to start more than one create thread per OSS.
1031 Description: Sanity.sh test_65k failed: setstripe should have succeeded
1032 Details : execute lfs setstripe on client
1036 Description: mds_obd_create error creating tmp object
1037 Details : When the user sets quota on root, llog will be affected and can't
1038 create files and write files.
1041 Frequency : Always on ia64 patchless client, and possibly others.
1043 Description: Add EXPORT_SYMBOL check for node_to_cpumask symbol.
1044 Details : This allows the patchless client to be loaded on architectures
1045 without this export.
1049 Description: RedHat Update kernel for RHEL5
1050 Details : Add the kernel config file for RHEL5.
1054 Description: RedHat Update kernel for RHEL5
1055 Details : Modify the kernel config file more closer RHEL5.
1059 Description: Build failure against Centos5 (RHEL5)
1060 Details : Define PAGE_SIZE when it isn't present.
1064 Description: test_7 on sanity-quota.sh can't be used on separate machines
1065 Details : add facet to handle it
1069 Description: "ll_intent_file_open()) lock enqueue: err: -13" with nfs
1070 Details : with NFS, the anon dentry's parent was set to itself in
1071 d_alloc_anon(), so in MDS, we use rec->ur_fid1 to find the
1072 corresponding dentry other than use rec->ur_name.
1074 Severity : enhancement
1076 Description: lfs setstripe enhancement
1077 Details : Make lfs setstripe understand 'k', 'm' and 'g' for stripe size.
1081 Description: enable data checksumming by default
1082 Details : enable checksum by default, allow --disable-checksum
1083 configure option and "-o nochecksum" mount option. Checksums
1084 can also be disabled at runtime via $LPROC/osc/*/checksum_pages.
1088 Description: lustre support for RHEL5
1089 Details : Add support for RHEL5.
1093 Description: Client eviction due to failover config
1094 Details : after a connection loss, the lustre client should attempt to
1095 reconnect to the last active server first before trying the
1096 other potential connections.
1100 Description: when mds and osts use different quota unit(32bit and 64bit),
1101 quota will be released repeatly.
1102 Details : void sending multiple quota reqs to mds, which will keep the
1103 status between the reqs.
1107 Description: osts not allocated evenly to files
1108 Details : change the condition to increase offset_idx
1111 Frequency : Always for filesystems larger than 2TB on 32-bit systems.
1113 Description: Data corruption for OSTs that are formatted larger than 2TB
1115 Details : When generating the bio request for lustre file writes the
1116 sector number would overflow a temporary variable before being
1117 used for the IO. The data reads correctly from Lustre (which
1118 will overflow in a similar manner) but other file data or
1119 filesystem metadata may be corrupted in some cases.
1123 Description: Tune the kernel for good SCSI performance.
1124 Details : Set the value of /sys/block/{dev}/queue/max_sectors_kb
1125 to the value of /sys/block/{dev}/queue/max_hw_sectors_kb
1130 Description: rewrite ext2-derived code in obdclass/uuid.c
1131 Details : rewrite inherited code (uuid parsing code from ext2 utils)
1132 from scratch preserving functionality.
1136 Description: "lfs find -obd UUID" prints directories
1137 Details : "lfs find -obd UUID" will return all directory names instead
1138 of just file names. It is incorrect because the directories
1139 do not reside on the OSTs.
1143 Description: use i_size_read and i_size_write in 2.6 port
1144 Details : replace inode->i_size access with i_size_read/write()
1147 Frequency : when removing large files
1149 Description: scheduling issue during removal of large Lustre files
1150 Details : Don't take the BKL in fsfilt_ext3_setattr() for 2.6 kernels.
1151 It causes scheduling issues when removing large files (17TB in the
1155 Frequency : only with liblustre clients on XT3
1157 Description: evictions taking too long
1158 Details : allow llrd to evict clients directly on OSTs
1161 Frequency : only on ppc
1163 Description: /proc/fs/lustre/devices broken on ppc
1164 Details : The patch as applied to 1.6.2 doesn't look correct for all arches.
1165 We should make sure the type of 'index' is loff_t and then cast
1166 explicitly as needed below. Do not assign an explicitly cast
1170 Frequency : only for rhel5
1172 Description: Kernel patches update for RHEL5 2.6.18-8.1.10.el5.
1173 Details : Modify the target file & which_kernel.
1175 Severity : enhancement
1177 Description: omit set fsid for export NFS
1178 Details : fix set/restore device id for avoid EMFILE error and mark lustre
1179 fs as FS_REQUIRES_DEV for avoid problems with generate fsid.
1183 Frequency : Always, for kernels after 2.6.16
1184 Description: Fix warning idr_remove called for id=.. which is not allocated.
1185 Details : Last kernels save old s_dev before kill super and not allow
1186 to restore from callback - restore it before call kill_anon_super.
1190 Description: buffer overruns could theoretically occur
1191 Details : llapi_semantic_traverse() modifies the "path" argument by
1192 appending values to the end of the origin string, and a
1193 overrun may occur. Adding buffer overrun check in liblustreapi.
1197 Description: Fix error on 'ls .' at the top of the Lustre mount.
1198 Details : Don't revalidate dentry if it is a root dentry.
1202 Description: Kernel patches update for RHEL4 2.6.9-55.0.6.
1203 Details : Modify vm-tunables-rhel4.patch.
1207 Description: Kernel config for 2.6.18-vanilla.
1208 Details : Modify targets/2.6-vanilla.target.in.
1209 Add config file kernel-2.6.18-2.6-vanilla-i686.config.
1210 Add config file kernel-2.6.18-2.6-vanilla-i686-smp.config.
1211 Add config file kernel-2.6.18-2.6-vanilla-x86_64.config.
1212 Add config file kernel-2.6.18-2.6-vanilla-x86_64-smp.config.
1216 Description: improve handling recoverable errors
1217 Details : if request processig with error which can be recoverable on server
1218 request should be resend, otherwise page released from cache and
1223 Description: Kernel patches update for RHEL5 2.6.18-8.1.14.el5.
1224 Details : Modify target file & which_patch.
1225 A flaw was found in the IA32 system call emulation provided
1226 on AMD64 and Intel 64 platforms. An improperly validated 64-bit
1227 value could be stored in the %RAX register, which could trigger an
1228 out-of-bounds system call table access. An untrusted local user
1229 could exploit this flaw to run code in the kernel
1230 (ie a root privilege escalation). (CVE-2007-4573).
1234 Description: O_DIRECT bypasses client statistics.
1235 Details : When running with O_DIRECT I/O, neither the client rpc_stats nor
1236 read_ahead_stats were updated. copied the stats section from
1237 osc_send_oap_rpc() into async_internal().
1241 Description: Add jbd statistics patch for RHEL5 and 2.6.18-vanilla
1245 Description: change order of libsysio includes
1246 Details : '#include sysio.h' should always come before '#include xtio.h'
1250 Description: handle "serious error: objid * already exists" more gracefully
1251 Details : If LAST_ID value on disk is smaller than the objects existing in
1252 the O/0/d* directories, it indicates disk corruption and causes an
1253 LBUG(). If the object is 0-length, then we should use the existing
1254 object. This will help to avoid a full fsck in most cases.
1256 Severity : enhancement
1258 Description: adapt the lustre_config script to support the upgrade case
1259 Details : Add "-u" option for lustre_config script to support upgrading 1.4
1260 server targets to 1.6 in parallel.
1264 Description: To avoid grant space > avaible space when the disk is almost
1265 full. Without this patch you might see the error "grant XXXX >
1266 available" or some LBUG about grant, when the disk is almost
1268 Details : In filter_check_grant, for non_grant cache write, we should
1269 check the left space by if (*left > ungranted + bytes), instead
1270 of (*left > ungranted), because only we are sure the left space
1271 is enough for another "bytes", then the ungrant space should be
1272 increase. In client, we should update cl_avail_grant only there
1273 is OBD_MD_FLGRANT in the reply.
1277 Description: Update RHEL 4 kernel to fix local root privilege escalation.
1278 Details : Update to the latest RHEL 4 kernel to fix the vulnerability
1279 described in CVE-2007-4573. This problem could allow untrusted
1280 local users to gain root access.
1283 Frequency : when using O_DIRECT and quotas
1285 Description: Incorrect file ownership on O_DIRECT output files
1286 Details : block usage reported by 'lfs quota' does not take into account
1287 files that have been written with O_DIRECT.
1292 Description: touch file failed when fs is not full
1293 Details : OST in recovery should not be discarded by MDS in alloc_qos(),
1294 otherwise we can get ENOSP while fs is not full.
1298 Description: parallel lock callbacks
1299 Details : Instead of sending blocking and completion callbacks as separated
1300 requests, adding them to a set and sending in parallel.
1303 Frequency : only for Cray XT3
1304 Bugzilla : 12829/13455
1305 Description: Changing primary group doesn't change the group lustre assigns to
1307 Details : When CRAY_XT3 is defined, the fsgid supplied by the client is
1308 overridden with the primary group provided by the group upcall,
1309 whereas the supplied fsgid can be trusted if it is in the list of
1310 supplementary groups returned by the group upcall.
1312 Severity : enhancement
1314 Description: Allow masking D_WARNING, D_ERROR messages from console
1315 Details : Console messages can now be disabled via lnet.printk.
1319 Description: User code with malformed file open parameter crashes client node
1320 Details : Before packing join_file req, all the related reference should be
1321 checked carefully in case some malformed flags cause fake
1322 join_file req on client.
1326 Description: LDLM_ENQUEUE races with LDLM_CP_CALLBACK
1327 Details : ldlm_completion_ast() assumes that a lock is granted when the req
1328 mode is equal to the granted mode. However, it should also check
1329 that LDLM_FL_CP_REQD is not set.
1333 Description: Heavy nfs access might result in deadlocks
1334 Details : After ELC code landed, it is now improper to enqueue any mds
1335 locks under och_sem, because enqueue might want to decide to
1336 cancel open locks for same inode we are holding och_sem for.
1340 Description: Client eviction while running blogbench
1341 Details : A lot of unlink operations with concurrent I/O can lead to a
1342 deadlock causing evictions. To address the problem, the number of
1343 oustanding OST_DESTROY requests is now throttled to
1344 max_rpcs_in_flight per OSC and LDLM_FL_DISCARD_DATA blocking
1345 callbacks are processed in priority.
1349 Description: enable ACLs on MDS by default
1350 Details : ACLs must be enabled on MDS by default.
1353 Frequency : PPC/PPC64 only
1355 Description: conflicts between asm-ppc64/types.h and lustre_types.h
1356 Details : fix duplicated definitions between asm-ppc64/types.h and
1357 lustre_types.h on PPC.
1360 Frequency : PPC/PPC64 only
1362 Description: asm-ppc/segment.h does not exist
1363 Details : fix compile issue on PPC.
1367 Description: better handle error messages in extents code
1370 Frequency : RHEL4 only
1372 Description: mkfs is very slow on IA64/RHEL4
1373 Details : A performance regression has been discovered in the MPT Fusion
1374 driver between versions 3.02.73rh and 3.02.99.00rh. As a
1375 consequence, we have downgraded the MPT Fusion driver in the RHEL4
1376 kernel from 3.02.99.00 to 3.02.73 until this problem is fixed.
1378 Severity : enhancement
1380 Description: SNMP support enhancement
1381 Details : Adding total number of sampled request for an MDS node in snmp
1384 Severity : enhancement
1386 Description: Optimize ldlm waiting list processing for PR extent locks
1387 Details : When processing waiting list for read extent lock and meeting read
1388 lock that is same or wider to it that is not contended, skip
1389 processing rest of the list and immediatelly return current
1390 status of conflictness, since we are guaranteed there are no
1391 conflicting locks in the rest of the list.
1395 Description: Time out and refuse to reconnect
1396 Details : When the failover node is the primary node, it is possible
1397 to have two identical connections in imp_conn_list. We must
1398 compare not conn's pointers but NIDs, otherwise we can defeat
1399 connection throttling.
1403 Description: port llog fixes from b1_6 into HEAD
1404 Details : Port llog reference couting and some llog cleanups from b1_6
1405 (bug 10800) into HEAD, for protect from panic and access to already
1406 free llog structures.
1410 Description: Detect stride IO mode in read-ahead
1411 Details : When a client does stride read, read-ahead should detect that and
1412 read-ahead pages according to the detected stride pattern.
1416 Description: data checksumming impacts single node performance
1417 Details : add support for several checksum algorithm. Currently, only CRC32
1418 and Adler-32 are supported. The checksum type can be changed on
1419 the fly via /proc/fs/lustre/osc/*/checksum_type.
1423 Description: use adler32 for page checksums
1424 Details : when available, use the Adler-32 algorithm instead of CRC32 for
1429 Description: build for x2 fails
1430 Details : fix compile issue on Cray systems.
1434 Description: Properly match for duplicate locks
1435 Details : Due to different lock order from skiplists code, we need to
1436 traverse entire list for now
1439 Frequency : only on PPC/SLES10
1441 Description: "BITS_PER_LONG is not 32 or 64" in linux/idr.h
1442 Details : On SLES10/PPC, fs.h includes idr.h which requires BITS_PER_LONG to
1443 be defined. Add a hack in mkfs_lustre.c to work around this compile
1448 Description: LASSERT on MDS when client holding flock lock dies
1449 Details : ldlm pool logic depends on number of granted locks equal to
1450 number of released locks which is not true for flock locks, so
1451 just exclude such locks from consideration.
1455 Description: MDS deadlock with many ll_sync_lov threads and I/O stalled
1456 Details : Use fsfilt_sync() for both the whole filesystem sync and
1457 individual file sync to eliminate dangerous inode locking
1458 with I_LOCK that can lead to a deadlock.
1462 Description: performance in 1.6.3
1463 Details : Force q->max_phys_segments to MAX_PHYS_SEGMENTS on SLES10 to be
1464 sure that 1MB requests are not fragmented by the block layer.
1466 Severity : enhancement
1468 Description: organize the server-side client stats on per-nid basis
1469 Details : Change the structure of stats under obdfilter and mds to
1477 The "uuid"s file would list the uuids of _active_ exports.
1478 And the clear entry is to clear all stats and stale nids.
1480 Severity : enhancement
1482 Description: eliminate client locks in face of contention
1483 Details : file contention detection and lockless i/o implementation
1484 for contended files.
1488 Description: Reinitialize optind to 0 so that interactive lfs works in all cases
1491 Frequency : very rare, if additional xattrs are used on kernels >= 2.6.12
1493 Description: MDS may lose file striping (and hence file data) in some cases
1494 Details : If there are additional extended attributes stored on the MDS,
1495 in particular ACLs, SELinux, or user attributes (if user_xattr
1496 is specified for the client mount options) then there is a risk
1497 of attribute loss. Additionally, the Lustre file striping
1498 needs to be larger than default (e.g. striped over all OSTs),
1499 and an additional attribute must be stored initially in the
1500 inode and then increase in size enough to be moved to the
1501 external attribute block (e.g. ACL growing in size) for file
1506 Description: skiplist implementation simplification
1507 Details : skiplists are used to group compatible locks on granted list
1508 that was implemented as tracking first and last lock of each
1509 lock group the patch changes that to using doubly linked lists
1513 Description: MDS LBUG: ASSERTION(!IS_ERR(dchild))
1514 Details : Change LASSERTs to client eviction (i.e. abort client's recovery)
1515 because LASSERT on both the data supplied by a client, and the
1516 data on disk is dangerous and incorrect.
1518 Severity : enhancement
1520 Description: Slow truncate/writes to huge files at high offsets.
1521 Details : Directly associate cached pages to lock that protect those pages,
1522 this allows us to quickly find what pages to write and remove
1523 once lock callback is received.
1527 Description: more ldlm soft lockups
1528 Details : In ldlm_resource_add_lock(), call to ldlm_resource_dump()
1529 starve other threads from the resource lock for a long time in
1530 case of long waiting queue, so change the debug level from
1531 D_OTHER to the less frequently used D_INFO.
1533 Severity : enhancement
1535 Description: add -gid, -group, -uid, -user options to lfs find
1539 Description: Hung threads in invalidate_inode_pages2_range
1540 Details : The direct IO path doesn't call check_rpcs to submit a new RPC
1541 once one is completed. As a result, some RPCs are stuck in the
1542 queue and are never sent.
1546 Description: filter threads hungs on waiting journal commit
1547 Details : Cleanup filter group llog code, then only filter group llog will
1548 be only created in the MDS/OST syncing process.
1552 Description: Procfs and llog threads access destoryed import sometimes.
1553 Details : Sync the import destoryed process with procfs and llog threads by
1554 the import refcount and semaphore.
1556 Severity : enhancement
1558 Description: openlock cache of b1_6 port to HEAD
1563 Description: kernel BUG at ldiskfs2_ext_new_extent_cb
1564 Details : If insertion of an extent fails, then discard the inode
1565 preallocation and free data blocks else it can lead to duplicate
1570 Description: don't always update ctime in ext3_xattr_set_handle()
1571 Details : Current xattr code updates inode ctime in ext3_xattr_set_handle.
1572 In some cases the ctime should not be updated, for example for
1573 2.0->1.8 compatibility it is necessary to delete an xattr and it
1574 should not update the ctime.
1578 Bugzilla : 15713/16362
1579 Description: Assertion in iopen_connect_dentry in 1.6.3
1580 Details : looking up an inode via iopen with the wrong generation number can
1581 populate the dcache with a disconneced dentry while the inode
1582 number is in the process of being reallocated. This causes an
1583 assertion failure in iopen since the inode's dentry list contains
1584 both a connected and disconnected dentry.
1588 Description: assertion failure in ldlm_handle2lock()
1589 Details : fix a race between class_handle_unhash() and class_handle2object()
1590 introduced in lustre 1.6.5 by bug 13622.
1595 Description: Kernel BUG: sd_iostats_bump: unexpected disk index
1596 Details : remove the limit of 256 scsi disks in the sd_iostat patch
1601 Description: oops in sd_iostats_seq_show()
1602 Details : unloading/reloading the scsi low level driver triggers a kernel
1603 bug when trying to access the sd iostat file.
1608 Description: Kernel panics during QLogic driver reload
1609 Details : REQ_BLOCK_PC requests are not handled properly in the sd iostat
1610 patch, causing memory corruption.
1615 Description: journal_dev option does not work in b1_6
1616 Details : pass mount option during pre-mount.
1618 Severity : enhancement
1620 Description: Add a FIEMAP(FIle Extent MAP) ioctl
1621 Details : FIEMAP ioctl will allow an application to efficiently fetch the
1622 extent information of a file. It can be used to map logical blocks
1623 in a file to physical blocks in the block device.
1627 Description: LDLM soft lockups - improvement
1628 Details : It is be possible to send the lock handle along with each read
1629 or write request because the client is already doing a lock match
1630 itself so there isn't any reason the OST should have to re-do that
1634 Frequency : only on Cray X2
1636 Description: X2 build failures
1637 Details : fix build failures on Cray X2.
1641 Description: xid & resent requests
1642 Details : Initialize RPC XID from clock at startup (randomly if clock is
1645 Severity : enhancement
1647 Description: Add lustre_start utility to start or stop multiple Lustre servers
1652 Description: Lustre GPF in {:ptlrpc:ptlrpc_server_free_request+373}
1653 Details : In case of memory pressure, list_del() can be called twice on
1654 req->rq_history_list, causing a kernel oops.
1658 Description: kptllnd_peer_check_sends()) ASSERTION(!in_interrupt()) failed
1659 Details : fix stack overflow in the distributed lock manager by defering
1660 export eviction after a failed AST to the elt thread instead of
1661 handling it in the dlm interpret routine.
1665 Description: Convert some comments to new format.
1666 Details : Update documenting comments to match doxygen conventions.
1670 Description: Grammar fixes.
1671 Details : A couple of trivial spelling fixes.
1675 Description: OSD_COUNTERS-mandatory
1676 Details : Make previously optional ->oti_{w,r}_locks sanity checks mandatory
1677 to simplify the code.
1681 Description: simplify cmm_device freeing logic.
1682 Details : Call cmm_device_free() in the failure path of cmm_device_alloc().
1686 Description: Add lockdep support to dt_object_operations locking interface.
1687 Details : Augment ->do_{read,write}_lock() prototypes with a `role'
1688 parameter indicating lock ordering. Update mdd code to use new
1693 Description: Introduce failloc constants for lockless IO tests.
1694 Details : Add two new failloc constants to test lockless IO. Only one of
1695 them in implemented---another is checked in yet to be landed
1700 Description: Add lockdep support for inode mutex.
1701 Details : Introduce and use new LOCK_INODE_MUTEX_PARENT() macro to be used
1702 in the situations where Lustre has to lock more than one inode
1707 Description: Add optional invariants checking support.
1708 Details : Add new LINVRNT() macro, optional on new --enable-invariants
1709 configure switch. This macro is to be used for consistency and
1710 sanity checks that are too expensive to be left in `production'
1715 Description: Zap lock->l_granted_mode with explicit LCK_MINMODE.
1716 Details : Use LCK_MINMODE rather than 0 to reset lock->l_granted_mode to
1721 Description: Add lockdep support for ldlm_lock and ldlm_resource.
1722 Details : Use spin_lock_nested() in (the only) situation where more than
1723 one ldlm_lock is locked simultaneously. Also, fix possible
1724 dead-lock in ldlm_lock_change_resource() by enforcing particular
1729 Description: Use struct ldlm_callback_suite in ldlm_lock_create().
1730 Details : Instead of specifying each ldlm_lock call-back through separate
1731 parameter, wrap them into struct ldlm_callback_suite.
1735 Description: Kill join_lru obd method and its callers.
1736 Details : CLIO uses lock weighting policy to keep locks over mmapped regions
1737 in memory---a requirement implemented through ->o_join_lru() obd
1738 method in HEAD. Remove this method and its users.
1742 Description: Add asynchronous ldlm ENQUEUE completion handler.
1743 Details : CLIO posts ENQUEUE requests asynchronously through ptlrpcd---a
1744 case that stock ldlm_completion_ast() cannot handle as it waits
1745 until lock is granted. Introduce new ldlm_completion_ast_async()
1746 for this. Also comment ldlm_completion_ast().
1750 Description: ldlm_error <-> errno conversion.
1751 Details : Add functions to map (rather arbitrary) between LDLM error codes
1752 and standard errno values. CLIO needs this to prevent LDLM specific
1753 constants from escaping ldlm and osc.
1757 Description: Kill unused ldlm_handle2lock_ns() function.
1758 Details : Kill unused ldlm_handle2lock_ns() function.
1762 Description: Add lu_ref support to ldlm_lock
1763 Details : lu_ref support for ldlm_lock and ldlm_resource. See lu_ref patch.
1764 lu_ref fields ->l_reference and ->lr_reference are added to ldlm_lock
1765 and ldlm_resource. LDLM interface has to be changed, because code that
1766 releases a reference on a lock, has to "know" what reference this is.
1767 In the most frequent case
1769 lock = ldlm_handle2lock(handle);
1771 LDLM_LOCK_PUT(lock);
1773 no changes are required. When any other reference (received _not_ from
1774 ldlm_handle2lock()) is released, LDLM_LOCK_RELEASE() has to be called
1775 instead of LDLM_LOCK_PUT().
1777 Arguably, changes are pervasive, and interface requires some discipline
1778 for proper use. On the other hand, it was very instrumental in finding
1779 a few leaked lock references.
1783 Description: Add ldlm_lock_addref_try().
1784 Details : Introduce ldlm_lock_addref_try() function (used by CLIO) that
1785 attempts to addref a lock that might be being canceled
1790 Description: Add ldlm_weigh_callback().
1791 Details : Add new ->l_weigh_ast() call-back to ldlm_lock. It is called
1792 by ldlm_cancel_shrink_policy() to estimate lock "value", instead of
1793 hard-coded `number of pages' logic.
1797 Description: Add lockdep annotations to llog code.
1798 Details : Use appropriately tagged _nested() locking calls in the places
1799 where llog takes more than one ->lgh_lock lock.
1803 Description: Add loi_kms_set().
1804 Details : Wrap kms updates into a helper function.
1808 Description: Constify instances of struct lsm_operations.
1809 Details : Constify instances of struct lsm_operations.
1813 Description: lu_conf support.
1814 Details : On a server, a file system object is uniquely identified
1815 by a fid, which is sufficient to locate and load all object
1816 state (inode). On a client, on the other hand, more data are
1817 necessary instantiate an object. Change lu_object_find() and
1818 friends to take additional `lu_conf' argument describing object.
1819 Typically this includes layout information.
1823 Description: lu_context fixes.
1824 Details : Introduce new lu_context functions that are needed on the client
1825 side, where some system threads (ptlrpcd) are shared by multiple
1826 modules, and so cannot be stopped during module shutdown.
1830 Description: Add start and stop methods to lu_device_type_operations.
1831 Details : Introduce two new methods in lu_device_type_operations, that are
1832 invoked when first instance of a given type is created and last one
1833 is destroyed respectively. This is need by CLIO.
1837 Description: Add lu_ref support to struct lu_device.
1838 Details : Add lu_ref support to lu_object and lu_device. lu_ref is used to
1839 track leaked references.
1843 Description: Introduce lu_kmem_descr.
1844 Details : lu_kmem_descr and its companion interface allow to create
1845 and destroy a number of kmem caches at once.
1849 Description: Fix lu_object finalization race.
1850 Details : Fix a race between lu_object_find() finding an object and its
1851 concurrent finalization. This race is (most likely) not possible
1852 on the server, but might happen on the client.
1856 Description: Introduce lu_ref interface.
1857 Details : lu_ref is a debugging module allowing to track references to
1858 a given object. It is quite cpu expensive, and has to be
1859 explicitly enabled with --enable-lu_ref. See usage description
1864 Description: Factor lu_site procfs stats into a separate function.
1865 Details : Separate lu_site stats printing code into a separate function
1866 that can be reused on a client.
1870 Description: Constify instances of struct {lu,dt,md}_device_operations.
1871 Details : Constify instances of struct {lu,dt,md}_device_operations.
1875 Description: Introduce struct md_site and move meta-data specific parts of
1876 struct lu_site here.
1877 Details : Move md-specific fields out of struct lu_site into special struct
1878 md_site, so that lu_site can be used on a client.
1882 Description: Kill mdd_lov_destroy().
1883 Details : Remove unused mdd code.
1887 Description: Add st_block checking to multistat.c.
1888 Details : Add st_block checking to multistat.c.
1892 Description: Add lu_ref support to struct obd_device.
1893 Details : Add lu_ref tracking to obd_device.
1897 Description: Kill obd_set_fail_loc().
1898 Details : Remove unused code.
1902 Description: Add special type for ptlrpc_request interpret functions.
1903 Details : Add lu_env parameter to ->rq_interpreter call-back. NULL is passed
1904 there. Actual usage will be in CLIO.
1908 Description: Replace RW_LOCK_UNLOCKED() macro with rwlock_init().
1909 Details : Replace RW_LOCK_UNLOCKED() with rwlock_init() as the former
1910 doesn't work with lockdep.
1914 Description: Add rwv.c test program.
1915 Details : New testing program exercising readv(2) and writev(2) (Qian).
1919 Description: Add sendfile.c test program.
1920 Details : New testing program exercising sendfile(2) (Jay).
1924 Description: Ratelimit a message that can be very frequent.
1925 Details : Ratelimit a memory allocation failure message that can
1930 Description: Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
1931 Details : Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
1935 Description: Liblustre build fixes.
1936 Details : Liblustre build fixes.
1940 Description: libcfs: add cfs_{need,cond}_resched() interface.
1941 Details : libcfs: add cfs_{need,cond}_resched() definition and
1942 implementations for Linux, NT, and liblustre.
1944 Severity : enhancement
1946 Description: More exported tunables for mballoc
1947 Details : Add support for tunable preallocation window and new tunables for
1948 large/small requests
1952 Description: Detect corruption of block bitmap and checking for preallocations
1953 Details : Checks validity of on-disk block bitmap. Also it does better
1954 checking of number of applied preallocations. When corruption is
1955 found, it turns filesystem readonly to prevent further corruptions.
1959 Description: (rw.c:1323:ll_read_ahead_pages()) ASSERTION(page_idx > ria->ria_stoff) failed
1960 Details : Once the unmatched stride IO mode is detected, shrink the stride-ahead
1961 window to 0. If it does hit cache miss, and read-pattern is still
1962 stride-io mode, does not reset the stride window, but also does not
1963 increase the stride window length in this case.
1967 Frequency : only for big-endian servers
1968 Description: Check if system is big-endian while mounting fs with extents feature
1969 Details : Mounting a filesystem with extents feature will fail on big-endian
1970 systems since ext3-based ldiskfs is not supported on big-endian
1971 systems. This can be over-riden with "bigendian_extents" mount option.
1973 Severity : enhancement
1975 Description: The root squash functionality
1976 Details : A security feature, which is to prevent users from being able
1977 to mount lustre on their desktop, run as root, and delete
1978 all of the files in the filesystem. The goal is accomplished by
1979 remapping user id (UID) and group id (GID) of the root user to
1980 a UID and GID specified by the system administartor via Lustre
1981 configuration management server (MGS). The functionality also
1982 allows to specify sets of clients for which the remapping does
1987 Description: Excessive recovery window
1988 Details : With AT enabled, the recovery window can be excessively long (6000+
1989 seconds). To address this problem, we no longer use
1990 OBD_RECOVERY_FACTOR when extending the recovery window (the connect
1991 timeout no longer depends on the service time, it is set to
1992 INITIAL_CONNECT_TIMEOUT now) and clients report the old service
1993 time via pb_service_time.
1997 Description: Watchdog triggered on MDS failover
1998 Details : enable OBD_CONNECT_MDT flag when connecting from the MDS so that
1999 the OSTs know that the MDS "UUID" can be reused for the same export
2000 from a different NID, so we do not need to wait for the export to be
2004 Frequency : rare, only if using MMP with Linux RAID
2006 Description: MMP doesn't work with Linux RAID
2007 Details : While using HA for Lustre servers with Linux RAID, it is possible
2008 that MMP will not detect multiple mounts. To make this work we
2009 need to unplug the device queue in RAID when the MMP block is being
2010 written. Also while reading the MMP block, we should read it from
2011 disk and not the cached one.
2013 Severity : enhancement
2015 Description: open file using fid
2016 Details : A file can be opened using just its fid, like
2017 <mntpt>/.lustre/fid/SEQ:OID:VER - this is needed for HSM and replication
2020 Frequency : Only in RHEL5 when mounting multiple ext3 filesystems
2023 Description: "kmem_cache_create: duplicate cache jbd_4k" error message
2024 Details : add proper locking for creation of jbd_4k slab cache
2028 Description: MMP check in ext3_remount() fails without displaying any error
2029 Details : When multiple mount protection fails during remount, proper error
2032 --------------------------------------------------------------------------------
2034 2007-08-10 Cluster File Systems, Inc. <info@clusterfs.com>
2036 * Support for kernels:
2037 2.6.5-7.283 (SLES 9),
2038 2.6.9-55.EL (RHEL 4),
2039 2.6.16.46-0.14 (SLES 10),
2040 2.6.18.8 vanilla (kernel.org)
2041 * Client support for unpatched kernels:
2042 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
2043 2.6.16 - 2.6.22 vanilla (kernel.org)
2044 * Due to recently discovered recovery problems, we do not recommend
2045 using patchless RHEL 4 clients with this or any earlier release.
2046 * Recommended e2fsprogs version: 1.39.cfs8
2047 * Note that reiserfs quotas are disabled on SLES 10 in this kernel.
2048 * Starting with this release, the ldiskfs backing filesystem required
2049 by Lustre is now in its own package, lustre-ldiskfs. This package
2050 should be installed. It is versioned separately from Lustre and
2051 may be released separately in future.
2055 Description: block reactivating mgc import until all deactivates complete
2056 Details : Fix race when failing back MDT/MGS to itself (testing)
2059 Frequency : at statup only
2061 Description: mds_lov_synchronize race leads to various problems
2062 Details : simultaneous MDT->OST connections at startup can cause the
2063 sync to abort, leaving the OSC in a bad state.
2065 Severity : enhancement
2067 Description: add optional extra BUILD_VERSION info
2068 Details : add a new environment variable (namely LUSTRE_VERS) which allows
2069 to override the lustre version.
2072 Frequency : 2.6.18 servers only
2074 Description: ll_kern_mount() doesn't release the module reference
2075 Details : The ldiskfs module reference count never drops down to 0
2076 because ll_kern_mount() doesn't release the module reference.
2081 Description: server LBUG when using old ost_num_threads parameter
2082 Details : Accept the old ost_num_threads parameter but warn that it
2083 is deprecated, and fix an off-by-one error that caused an LBUG.
2088 Description: Transient SCSI error results in persistent IO issue
2089 Details : iobuf->dr_error is not reinitialized to 0 between two
2093 Frequency : sometimes when underlying device returns I/O errors
2095 Description: OSTs not going read-only during write failures
2096 Details : OSTs are not remounted read-only when the journal commit threads
2097 get I/O errors because fsfilt_ext3 calls journal_start/stop()
2098 instead of the ext3 wrappers.
2102 Description: poor connect scaling with increasing client count
2103 Details : Don't run filter_grant_sanity_check for more than 100 exports
2104 to improve scaling for large numbers of clients.
2107 Frequency : SLES10 only
2109 Description: sanity-quota.sh quotacheck failed: rc = -22
2110 Details : Quotas cannot be enabled on SLES10.
2113 Frequency : liblustre clients only
2115 Description: getdirentries does not give error when run on compute nodes
2116 Details : getdirentries does not fail when the size specified as an argument
2117 is too small to contain at least one entry
2119 Severity : enhancement
2121 Description: Add LNET router traceability for debug purposes
2122 Details : If a checksum failure occurs with a router as part of the
2123 IO path, the NID of the last router that forwarded the bulk data
2124 is printed so it can be identified.
2129 Description: OST "spontaneously" evicts client; client has imp_pingable == 0
2130 Details : Due to a race condition, liblustre clients were occasionally
2131 evicted incorrectly.
2133 Severity : enhancement
2135 Description: lfs setstripe use optional parameters instead of postional
2138 Severity : enhancement
2140 Description: Nanosecond timestamp support for ldiskfs
2141 Details : The on-disk ldiskfs filesystem has added support for nanosecond
2142 resolution timestamps. There is not yet support for this at
2143 the Lustre filesystem level.
2146 Frequency : during server recovery
2148 Description: MDS failing to send precreate requests due to OSCC_FLAG_RECOVERING
2149 Details : request with rq_no_resend flag not awake l_wait_event if they get
2153 Frequency : nfs export on patchless client
2155 Description: connectathon hang when test nfs export over patchless client
2156 Details : Disconnected dentry cannot be found with lookup, so we do not need
2157 to unhash it or make it invalid
2160 Description: fix llapi_lov_get_uuids() to allow many OSTs to be returned
2161 Details: : Change llapi_lov_get_uuids() to read the UUIDs from /proc instead
2162 of using an ioctl. This allows lfsck for > 160 OSTs to succeed.
2167 Description: open req refcounting wrong on reconnect
2168 Details : If reconnect happened between getting open reply from server and
2169 call to mdc_set_replay_data in ll_file_open, we will schedule
2170 replay for unreferenced request that we are about to free.
2171 Subsequent close will crash in variety of ways.
2172 Check that request is still eligible for replay in
2173 mdc_set_replay_data().
2178 Description: disable writes to filesystem when reading health_check file
2179 Details : the default for reading the health_check proc file has changed
2180 to NOT do a journal transaction and write to disk, because this
2181 can cause reads of the /proc file to hang and block HA state
2182 checking on a healthy but otherwise heavily loaded system. It
2183 is possible to return to the previous behaviour during configure
2184 with --enable-health-write.
2186 Severity : enhancement
2188 Description: 64-bit inode version
2189 Details: : Add a on-disk 64-bit inode version for ext3 to track changes made
2190 to the inode. This will be required for version-based recovery.
2195 Description: MDS fails to start if a duplicate client export is detected
2196 Details : in some rare cases it was possible for a client to connect to
2197 an MDS multiple times. Upon recovery the MDS would detect this
2198 and fail during startup. Handle this more gracefully.
2200 Severity : enhancement
2202 Description: Add -o localflock option to simulate old noflock
2204 Details : This will achieve local-only flock/fcntl locks
2210 Description: log_commit_thread vs filter_destroy race leads to crash
2211 Details : Take import reference before releasing llog record semaphore
2216 Description: Wrong request locking in request set processing
2217 Details : ptlrpc_check_set wrongly uses req->rq_lock for proctect add to
2218 imp_delayed_list, in this place should be used imp_lock.
2221 Frequency : when reconnection
2223 Description: Grant Leak when osc reconnect to OST
2224 Details : When osc reconnect ost, OST(filter) should check whether it
2225 should grant more space to client by comparing fed_grant and
2226 cl_avail_grant, and return the granted space to client instead
2227 of "new granted" space, because client will call osc_init_grant
2228 to update the client grant space info.
2231 Frequency : when client reconnect to OST
2233 Description: Grant Leak when osc do resend and replay bulk write
2234 Details : When osc reconnect to OST, OST(filter)should clear grant info of
2235 bulk write request, because the grant info while be sync between
2236 OSC and OST when reconnect, and we should ignore the grant info
2237 these of resend/replay write req.
2242 Description: Grant space more than avaiable left space sometimes.
2243 Details : When then OST is about to be full, if two bulk writing from
2244 different clients came to OST. Accord the avaliable space of the
2245 OST, the first req should be permitted, and the second one
2246 should be denied by ENOSPC. But if the seconde arrived before
2247 the first one is commited. The OST might wrongly permit second
2248 writing, which will cause grant space > avaiable space.
2251 Frequency : when client is evicted
2253 Description: Grant might be wrongly erased when osc is evicted by OST
2254 Details : when the import is evicted by server, it will fork another
2255 thread ptlrpc_invalidate_import_thread to invalidate the
2256 import, where the grant will be set to 0. While the original
2257 thread will update the grant it got when connecting. So if
2258 the former happened latter, the grant will be wrongly errased
2259 because of this race.
2264 Description: Checking Stale with correct fid
2265 Details : ll_revalidate_it should uses de_inode instead of op_data.fid2
2266 to check whether it is stale, because sometimes, we want the
2267 enqueue happened anyway, and op_data.fid2 will not be initialized.
2269 Severity : enhancement
2271 Description: update patchless client
2272 Details : Add support for patchless client with 2.6.20, 2.6.21 and RHEL 5
2275 Frequency : only with 2.4 kernel
2277 Description: random memory corruption
2278 Details : size of struct ll_inode_info is to big for union inode.u and this
2279 can be cause of random memory corruption.
2284 Description: Memory leak in recovery
2285 Details : Lov_mds_md was not free in an error handler in mds_create_object.
2286 It should also check obd_fail before fsfilt_start, otherwise if
2287 fsfilt_start return -EROFS,(failover mds during mds recovery).
2288 then the req will return with repmsg->transno = 0 and rc = EROFS.
2289 and we met hit the assert LASSERT(req->rq_reqmsg->transno ==
2290 req->rq_repmsg->transno) in ptlrpc_replay_interpret. Fcc should
2291 be freed no matter whether fsfilt_commit success or not.
2294 Frequency : only with huge count clients
2296 Description: Prevents from taking the superblock lock in llap_from_page for
2298 Details : using LL_ORIGIN_REMOVEPAGE origin flag instead of LL_ORIGIN_UNKNOW
2299 for llap_from_page call in ll_removepage prevents from taking the
2300 superblock lock for a soon died page.
2305 Description: Not check open intent error before release open handle
2306 Details : in some rare cases, the open intent error is not checked before
2307 release open handle, which may cause
2308 ASSERTION(open_req->rq_transno != 0), because it tries to release
2309 the failed open handle.
2314 Description: Set cat log bitmap only after create log success.
2315 Details : in some rare cases, the cat log bitmap is set too early. and it
2316 should be set only after create log success.
2320 Description: Accessing a block bevice can re-enable I/O when Lustre is
2321 tearing down a device.
2322 Details : dev_clear_rdonly(bdev) must be called in kill_bdev() instead of
2327 Description: service threads may hog cpus when there are a lot of requests
2328 Details : Insert cond_resched to give other threads a chance to use some CPU
2333 Description: the cat log was not initialized in recovery
2334 Details : When mds(mgs) do recovery, the tgt_count might be zero, so the
2335 unlink log on mds will not be initialized until mds post
2336 recovery. And also in mds post recovery, the unlink log will
2337 initialization will be done asynchronausly, so there will be race
2338 between add unlink log and unlink log initialization.
2342 Description: brw_stats were being printed incorrectly
2343 Details : brw_stats were being printed as log2 but all of them were not
2344 recorded as log2. Also remove some code duplication arising from
2345 filter_tally_{read,write}.
2349 Frequency : rare, only in recovery.
2350 Description: ASSERTION(req->rq_type != LI_POISON) failed
2351 Details : imp_lock should be held while iterating over imp_sending_list for
2352 prevent destroy request after get timeout in ptlrpc_queue_wait.
2356 Description: replay-single.sh test 52 fails
2357 Details : A lock's skiplist need to be cleanup when it being unlinked
2358 from its resource list.
2362 Description: Short directio read returns full requested size rather than
2364 Details : Direct I/O operations should return actual amount of bytes
2365 transferred rather than requested size.
2367 Severity : enhancement
2369 Description: metadata RPC reduction (e.g. for rm performance)
2370 Details : decrease the amount of synchronous RPC between clients and servers
2371 by canceling conflicing lock before the operation on the client
2372 and packing thier handles into the main operation RPC to server.
2374 Severity : enhancement
2376 Description: Async OSC create to avoid the blocking unnecessarily.
2377 Details : If a OST has no remain object, system will block on the creating
2378 when need to create a new object on this OST. Now, ways use
2379 pre-created objects when available, instead of blocking on an
2380 empty osc while others are not empty. If we must block, we block
2381 for the shortest possible period of time.
2385 Description: improve handling recoverable errors
2386 Details : if request processig with error which can be recoverable on server
2387 request should be resend, otherwise page released from cache and
2390 Severity : enhancement
2392 Description: refine locking for avoid write wrong info into lov_objid file
2393 Details : fix possible races with add new target and write/update data in
2396 --------------------------------------------------------------------------------
2398 2007-05-03 Cluster File Systems, Inc. <info@clusterfs.com>
2403 Frequency : on some architectures
2405 Description: 1.6 client sometimes fails to mount from a 1.4 MDT
2406 Details : Uninitialized flags sometimes cause configuration commands to
2410 Frequency : patchless clients only
2412 Description: missing __iget() symbol export
2413 Details : The __iget() symbol export is missing. To avoid the need for
2414 this on patchless clients the deathrow inode reaper is turned
2415 off, and we depend on the VM to clean up old inodes. This
2416 dependency was during via the fix for bug 12181.
2418 --------------------------------------------------------------------------------
2420 2007-04-19 Cluster File Systems, Inc. <info@clusterfs.com>
2422 * CONFIGURATION CHANGE. This version of Lustre WILL NOT
2423 INTEROPERATE with older versions automatically. In many cases a
2424 special upgrade step is needed. Please read the
2425 user documentation before upgrading any part of a live system.
2426 * WIRE PROTOCOL CHANGE from previous 1.6 beta versions. This
2427 version will not interoperate with 1.6 betas before beta5 (1.5.95).
2428 * WARNING: Lustre configuration and startup changes are required with
2429 this release. See https://mail.clusterfs.com/wikis/lustre/MountConf
2433 Severity : enhancement
2435 Description: MountConf
2436 Details : Lustre configuration is now managed via mkfs and mount
2437 commands instead of lmc and lconf. New obd types (MGS, MGC)
2438 are added for dynamic configuration management. See
2439 https://mail.clusterfs.com/wikis/lustre/MountConf for
2442 Severity : enhancement
2444 Description: dynamic OST addition
2445 Details : OSTs can now be added to a live filesystem
2447 Severity : enhancement
2449 Description: startup order invariance
2450 Details : MDTs and OSTs can be started in any order. Clients only
2451 require the MDT to complete startup.
2453 Severity : enhancement
2455 Description: parallel, asynchronous orphan cleanup
2456 Details : orphan cleanup is now performed in separate threads for each
2457 OST, allowing parallel non-blocking operation.
2459 Severity : enhancement
2461 Description: optimized stripe assignment
2462 Details : stripe assignments are now made based on ost space available,
2463 ost previous usage, and OSS previous usage, in order to try
2464 to optimize storage space and networking resources.
2466 Severity : enhancement
2468 Description: Permanently set tunables
2469 Details : All writable /proc/fs/lustre tunables can now be permanently
2470 set on a per-server basis, at mkfs time or on a live system.
2472 Severity : enhancement
2474 Description: Lustre message v2
2475 Details : Add lustre message format v2.
2477 Severity : enhancement
2479 Description: client OST exclusion list
2480 Details : Clients can be started with a list of OSTs that should be
2481 declared "inactive" for known non-responsive OSTs.
2485 Description: SPEC SFS validation failure on NFS v2 over lustre.
2486 Details : Changes the blocksize for regular files to be 2x RPC size,
2487 and not depend on stripe size.
2489 Severity : enhancement
2491 Description: Multiple MD RPCs in flight.
2492 Details : Further unserialise some read-only MDS RPCs - learn about intents.
2493 To avoid overly-overloading MDS, introduce a limit on number of
2494 MDS RPCs in flight for a single client and add /proc controls
2495 to adjust this limit.
2497 Severity : enhancement
2499 Description: client read/write statistics
2500 Details : Add client read/write call usage stats for performance
2501 analysis of user processes.
2502 /proc/fs/lustre/llite/*/offset_stats shows non-sequential
2503 file access. extents_stats shows chunk size distribution.
2504 extents_stats_per_process show chunk size distribution per
2507 Severity : enhancement
2509 Description: mds statistics
2510 Details : Add detailed mds operations statistics in
2511 /proc/fs/lustre/mds/*/stats.
2515 Description: Failure of copying files with lustre special EAs.
2516 Details : Client side always return success for setxattr call for lustre
2517 special xattr (currently only "trusted.lov").
2522 Description: Refcount LNET uuids
2523 Details : The global LNET uuid list grew linearly with every startup;
2524 refcount repeated list entries instead of always adding to
2527 Severity : enhancement
2529 Description: Dynamic service threads
2530 Details : Within a small range, start extra service threads
2531 automatically when the request queue builds up.
2534 Frequency : mixed-endian client/server environments
2536 Description: mixed-endian crashes
2537 Details : The new msg_v2 system had some failures in mixed-endian
2540 Severity : enhancement
2542 Description: Easy OST removal
2543 Details : OSTs can be permanently deactivated with e.g. 'lctl
2544 conf_param lustre-OST0001.osc.active=0'
2546 Severity : enhancement
2548 Description: MGS proc entries
2549 Details : Added basic proc entries for the MGS showing what filesystems
2552 Severity : enhancement
2554 Description: provide MGS failover
2555 Details : Added config lock reacquisition after MGS server failover.
2557 Severity : enhancement
2559 Description: add Linux 2.4 support
2560 Details : Added support for RHEL 2.4.21 kernel for 1.6 servers and clients
2564 Description: a large application tries to do I/O to the same resource and dies
2565 in the middle of it.
2566 Details : Check the req->rq_arrival time after the call to
2567 ost_brw_lock_get(), but before we do anything about
2568 processing it & sending the BULK transfer request. This
2569 should help move old stale pending locks off the queue as
2570 quickly as obd_timeout.
2573 Frequency : when an incorrect nid is specified during startup
2575 Description: ptlrpc connect to non-existant node causes kernel crash
2576 Details : LNET can't be re-entered from an event callback, which
2577 happened when we expire a message after the export has been
2578 cleaned up. Instead, hand the zombie cleanup off to another
2581 Severity : enhancement
2583 Description: plain/inodebits lock performance improvement
2584 Details : Grouping plain/inodebits in granted list by their request modes
2585 and bits policy, thus improving the performance of search through
2589 Frequency : only if OST filesystem is corrupted
2591 Description: client incorrectly hits assertion in ptlrpc_replay_req()
2592 Details : for a short time RPCs with bulk IO are in the replay list,
2593 but replay of bulk IOs is unimplemented. If the OST filesystem
2594 is corrupted due to disk cache incoherency and then replay is
2595 started it is possible to trip an assertion. Avoid putting
2596 committed RPCs into the replay list at all to avoid this issue.
2599 Frequency : liblustre (e.g. catamount) on a large cluster with >= 8 OSTs/OSS
2601 Description: System hang on startup
2602 Details : This bug allowed the liblustre (e.g. catamount) client to
2603 return to the app before handling all startup RPCs. This
2604 could leave the node unresponsive to lustre network traffic
2605 and manifested as a server ptllnd timeout.
2607 Severity : enhancement
2609 Description: Add "/proc/sys/lustre/debug_peer_on_timeout"
2610 Details : liblustre envirable: LIBLUSTRE_DEBUG_PEER_ON_TIMEOUT
2611 boolean to control whether to print peer debug info when a
2612 client's RPC times out.
2615 Frequency : only for kernels with patches from Lustre below 1.4.3
2617 Description: Remove old rdonly API
2618 Details : Remove old rdonly API which unused from at least lustre 1.4.3
2621 Frequency : only for devices with external journals
2623 Description: Set external device read-only also
2624 Details : During a commanded failover stop, we set the disk device
2625 read-only while the server shuts down. We now also set any
2626 external journal device read-only at the same time.
2629 Frequency : when upgrading from 1.4 while trying to change parameters
2631 Description: The wrong (new) MDC name was used when setting parameters for
2632 upgraded MDT's. Also allows changing of OSC (and MDC)
2633 parameters if --writeconf is specified at tunefs upgrade time.
2636 Frequency : when setting specific ost indicies
2638 Description: QOS code breaks on skipped indicies
2639 Details : Add checks for missing OST indicies in the QOS code, so OSTs
2640 created with --index need not be sequential.
2642 Severity : enhancement
2644 Description: Add uninit_groups feature to ldiskfs2 to speed up e2fsck
2645 Details : The uninit_groups feature works in conjunction with the kernel
2646 filesystem code (ldiskfs2 only) and e2fsprogs-1.39-cfs6 to speed
2647 up the pass1 processing of e2fsck. This is a read-only feature
2648 in ldiskfs2 only, so older kernels and current ldiskfs cannot
2649 mount filesystems that have had this feature enabled.
2651 Severity : enhancement
2653 Description: Improve multi-block allocation algorithm to avoid fragmentation
2654 Details : The mballoc3 code (ldiskfs2 only) adds new mechanisms to improve
2655 allocation locality and avoid filesystem fragmentation.
2657 ------------------------------------------------------------------------------
2659 2007-04-01 Cluster File Systems, Inc. <info@clusterfs.com>
2661 * Support for kernels:
2662 2.4.21-47.0.1.EL (RHEL 3)
2663 2.6.5-7.283 (SLES 9)
2664 2.6.9-42.0.10.EL (RHEL 4)
2665 2.6.12.6 vanilla (kernel.org)
2666 2.6.16.27-0.9 (SLES 10)
2667 * Recommended e2fsprogs version: 1.39.cfs5
2669 * Note that reiserfs quotas are disabled on SLES 10 in this kernel
2673 Frequency : occasional, depends on client load and configuration
2674 Bugzilla : 12181, 12203
2675 Description: data loss for recently-modified files
2677 Details : In some cases it is possible that recently written or created
2678 files may not be written to disk in a timely manner (this should
2679 normally be within 30s unless client IO load is very high).
2680 The problem appears as zero-length files or files that are a
2681 multiple of 1MB in size after a client crash or client eviction
2682 that are missing data at the end of the file.
2684 This problem is more likely to be hit on clients where files are
2685 repeatedly created and unlinked in the same directory, clients
2686 have a large amount of RAM, have many CPUs, the filesystem has
2687 many OSTs, the clients are rebooted frequently, and/or the files
2688 are not accessed by other nodes after being written.
2690 The presence of the problem can be detected by looking at
2691 /proc/sys/fs/inode-state. If the first number (nr_inodes) is
2692 smaller than the second (nr_unused) then dirty files will not
2693 be flushed automatically to disk. "sync; sleep 10" should be
2694 run several times on the node before unmounting it to update
2695 Lustre (this is also safe to run on nodes without this problem).
2697 There is also a related kernel bug in the RHEL4 4 2.6.9 kernel
2698 that can cause this same problem, so customers using that kernel
2699 also need to update the kernel in addition to Lustre. In order
2700 to properly fix this bug, the RHEL3 2.4.21 kernel is also updated.
2702 It is normal that files written just before a client crash (less
2703 than 30s) may not yet have been flushed to disk, even for local
2707 Frequency : frequent on thin XT3 nodes
2709 Description: UUID collision on thin XT3 Linux nodes
2710 Details : UUIDs on Compute Node Linux XT3 nodes were not generated
2711 randomly, since we relied on an insufficiently-seeded PRNG.
2716 Description: OSS hangs after "All ost request buffers busy"
2717 Details : A deadlock between quota and journal operations caused OSS
2718 hangs after printing "All ost request buffers busy."
2721 Frequency : always on liblustre builds
2723 Description: Cleanup compiler warnings on liblustre
2726 Frequency : always on liblustre builds on XT3
2728 Description: LC_CONFIG_CDEBUG don't run while build liblustre on XT3.
2732 Description: Addition of EXT3_FEATURE_RO_COMPAT_DIR_NLINKS flag for
2733 > 32000 subdirectories
2734 Details : Add EXT3_FEATURE_RO_COMPAT_DIR_NLINK flag to
2735 EXT3_FEATURE_RO_COMPAT_SUPP. This flag will be set whenever
2736 subdirectory count crosses 32000. This will aid e2fsck to
2737 correctly handle more than 32000 subdirectories.
2740 Frequency : liblustre (e.g. catamount) on a large cluster with >= 8 OSTs/OSS
2742 Description: System hang on startup
2743 Details : This bug allowed the liblustre (e.g. catamount) client to
2744 return to the app before handling all startup RPCs. This
2745 could leave the node unresponsive to lustre network traffic
2746 and manifested as a server ptllnd timeout.
2748 Severity : enhancement
2750 Description: Add "/proc/sys/lustre/debug_peer_on_timeout"
2751 (liblustre envirable: LIBLUSTRE_DEBUG_PEER_ON_TIMEOUT)
2752 boolean to control whether to print peer debug info when a
2753 client's RPC times out.
2758 Description: make O_SYNC working on 2.6 kernels
2759 Details : 2.6 kernels use different method for mark pages for write,
2760 so need add a code to lustre for O_SYNC work.
2765 Description: Failure to close file and release space on NFS
2766 Details : Put inode details into lock acquired in ll_intent_file_open.
2767 Use mdc_intent_lock in ll_intent_open to properly
2768 detect all kind of errors unhandled by mdc_enqueue.
2773 Description: proc file read during shutdown sometimes raced obd removal,
2775 Details : Add lock to prevent obd access after proc file removal.
2778 Frequency : Only for files larger than 4GB on 32-bit clients.
2780 Description: improperly doing page alignment of locks
2781 Details : Modify lustre core code to use CFS_PAGE_* defines instead of
2782 PAGE_*. Make CFS_PAGE_MASK a 64-bit mask.
2787 Description: RPCs being resent when they shouldn't be
2788 Details : Some RPCs that should not be resent are being resent. This
2789 can cause inconsistencies in the RPC state machine. Do not
2790 resend such requests.
2793 Frequency : rare, only with NFS export
2795 Description: Crash on NFS re-export node
2796 Details : under very unusual load conditions an assertion is hit in
2797 ll_intent_file_open()
2800 Frequency : only if OST filesystem is corrupted
2802 Description: client incorrectly hits assertion in ptlrpc_replay_req()
2803 Details : for a short time RPCs with bulk IO are in the replay list,
2804 but replay of bulk IOs is unimplemented. If the OST filesystem
2805 is corrupted due to disk cache incoherency and then replay is
2806 started it is possible to trip an assertion. Avoid putting
2807 committed RPCs into the replay list at all to avoid this issue.
2812 Description: large O_DIRECT requests fail under memory pressure/fragmentation
2813 Details : Large single O_DIRECT read and write calls can fail to allocate
2814 a sufficiently large buffer to process the request. In case of
2815 allocation failure the allocation is retried with a smaller
2816 buffer and broken into smaller requests.
2818 Severity : enhancement
2820 Description: Add -o localflock option to simulate old noflock behaviour.
2821 Details : This will achieve local-only flock/fcntl locks coherentness.
2826 Description: versioning check is incomplete
2827 Details : Checking the version difference of client vs. server, report
2828 error if the gap is too big.
2833 Description: add support PG_writeback bit
2834 Details : add support for PG_writeback bit for Lustre, for more carefull
2835 work with page cache in 2.6 kernel. This also fix some deadlocks
2836 and remove hack for work O_SYNC with 2.6 kernel.
2838 Severity : enhancement
2840 Description: Add uninit_groups feature to ldiskfs2 to speed up e2fsck
2841 Details : The uninit_groups feature works in conjunction with the kernel
2842 filesystem code (ldiskfs2 only) and e2fsprogs-1.39-cfs6 to speed
2843 up the pass1 processing of e2fsck. This is a read-only feature
2844 in ldiskfs2 only, so older kernels and current ldiskfs cannot
2845 mount filesystems that have had this feature enabled.
2847 Severity : enhancement
2849 Description: Improve multi-block allocation algorithm to avoid fragmentation
2850 Details : The mballoc3 code (ldiskfs2 only) adds new mechanisms to improve
2851 allocation locality and avoid filesystem fragmentation.
2853 ------------------------------------------------------------------------------
2855 2007-02-09 Cluster File Systems, Inc. <info@clusterfs.com>
2857 * Support for kernels:
2858 2.6.9-42.0.3.EL (RHEL 4)
2859 2.6.5-7.276 (SLES 9)
2860 2.4.21-47.0.1.EL (RHEL 3)
2861 2.6.12.6 vanilla (kernel.org)
2862 2.6.16.21-0.8 (SLES10)
2863 * Recommended e2fsprogs version: 1.39.cfs2-0
2865 * The backwards-compatible /proc/sys/portals symlink has been removed
2866 in this release. Before upgrading, please ensure that you change
2867 any configuration scripts or /etc/sysctl.conf files that access
2868 /proc/sys/portals/* or sysctl portals.* to use the corresponding
2869 entry in /proc/sys/lnet or sysctl lnet.*. This change can be made
2870 in advance of the upgrade on any system running Lustre 1.4.6 or
2871 newer, since /proc/sys/lnet was added in that version.
2872 * Note that reiserfs quotas are disabled on SLES 10 in this kernel
2876 Frequency : only when quota is used
2878 Description: avoid scanning export list for quota master
2879 Details : Change the algorithms to avoid scanning export list in order
2880 to improve the efficiency.
2883 Frequency : MDS failover only, very rarely
2885 Description: "went back in time" messages on mds failover
2886 Details : The greatest transno may be lost when the current operation
2887 finishes with an error (transno==0) and the client's last_rcvd
2888 record is over-written. Save the greatest transno in the
2889 mds_last_transno for this case.
2892 Frequency : always for specific kernels and striping counts
2894 Description: client may get "Matching packet too big" without ACL support
2895 Details : Clients compiled without CONFIG_FS_POSIX_ACL get an error message
2896 when trying to access files in certain configurations. The
2897 clients should in fact be denied when mounting because they do
2898 not understand ACLs.
2901 Frequency : Cray XT3 with more than 4000 clients and multiple jobs
2903 Description: many clients connecting with IO in progress causes connect timeouts
2904 Details : Avoid synchronous journal commits to avoid delays caused by many
2905 clients connecting/disconnecting when bulk IO is in progress.
2906 Queue liblustre connect requests on OST_REQUEST_PORTAL instead of
2907 OST_IO_PORTAL to avoid delays behind potentially many pending
2911 Frequency : occasionally with multiple writers to a single file
2913 Description: shared writes to file may result in wrong size reported by stat()
2914 Details : Allow growing of kms when extent lock is cancelled
2917 Frequency : always with random mmap IO to multi-striped file
2919 Description: mmap write might be lost if we are writing to a 'hole' in stripe
2920 Details : Only if the hole is at the end of OST object so that kms is too
2921 small. Fix is to increase kms accordingly in ll_nopage.
2924 Frequency : rare, only if OST filesystem is inconsistent with MDS filesystem
2926 Description: writes to a missing object would leak memory on the OST
2927 Details : If there is an inconsistency between the MDS and OST filesystems,
2928 such that the MDS references an object that doesn't exist, writes
2929 to that object will leak memory due to incorrect cleanup in the
2930 error handling path, eventually running out of memory on the OST.
2935 Description: Creating too long symlink causes lustre errors
2936 Details : Check symlink and name lengths before sending requests to MDS.
2939 Frequency : only if flock is enabled (not on by default)
2941 Description: posix locks not released on fd closure on 2.6.9+
2942 Details : We failed to add posix locks to list of inode locks on 2.6.9+
2943 kernels, this caused such locks not to be released on fd close and
2944 then assertions on fs unmount about still used locks.
2947 Frequency : MDS failover only, very rarely
2949 Description: clients may get ASSERTION(granted_lock != NULL)
2950 Details : When request was taking a long time, and a client was resending
2951 a getattr by name lock request. The were multiple lock requests
2952 with the same client lock handle and
2953 mds_getattr_name->fixup_handle_for_resent_request found one of the
2954 lock handles but later failed with ASSERTION(granted_lock != NULL).
2959 Description: handle->h_buffer_credits > 0, assertion failure
2960 Details : h_buffer_credits is zero after truncate, causing assertion
2961 failure. This patch extends the transaction or creates a new
2965 Frequency : NFS re-export or patchless client
2966 Bugzilla : 11179, 10796
2967 Description: Crash on NFS re-export node (__d_move)
2968 Details : We do not want to hash the dentry if we don't have a lock.
2969 But if this dentry is later used in d_move, we'd hit uninitialised
2970 list head d_hash, so we just do this to init d_hash field but
2971 leave dentry unhashed.
2974 Frequency : NFS re-export or patchless client
2976 Description: NFS exports has problem with symbolic link
2977 Details : lustre client didn't properly install dentry when re-exported
2978 to NFS or running patchless client.
2981 Frequency : NFS re-export or patchless client
2983 Description: Various nfs/patchless fixes.
2984 Details : fixes reuse disconected alias for lookup process - this fixes
2985 warning "find_exported_dentry: npd != pd",
2986 fix permission error with open files at nfs.
2987 fix apply umask when do revalidate.
2990 Frequency : occasional
2992 Description: Crash on NFS re-export node
2993 Details : calling clear_page() on the wrong pointer triggered oops in
2994 generic_mapping_read().
2997 Frequency : rarely, using O_DIRECT IO
2999 Description: unaligned directio crashes client with LASSERT
3000 Details : check for unaligned buffers before trying any requests.
3003 Frequency : rarely, using CFS RAID5 patches in non-standard kernel series
3005 Description: stale data returned from RAID cache
3006 Details : If only a small amount of IO is done to the RAID device before
3007 reading it again it is possible to get stale data from the RAID
3008 cache instead of reading it from disk.
3011 Frequency : always for sles10 kernel
3013 Description: sles10 support
3014 Details : ll_follow_link: compile fixes and using of nd_set_link
3015 under newer kernels.
3018 Frequency : depends on arch, kernel and compiler version, always on sles10
3021 Description: recursive or deep enough symlinks cause stack overflow
3022 Details : getting rid of large stack-allocated variable in
3026 Frequency : depends on hardware
3028 Description: lustre write performance loss in the SLES10 kernel
3029 Details : the performance loss is caused by using of write barriers in the
3030 ext3 code. The SLES10 kernel turns barrier support on by
3031 default. The fix is to undo that change for ldiskfs.
3033 ------------------------------------------------------------------------------
3035 2006-12-09 Cluster File Systems, Inc. <info@clusterfs.com>
3037 * Support for kernels:
3038 2.6.9-42.0.3EL (RHEL 4)
3039 2.6.5-7.276 (SLES 9)
3040 2.4.21-47.0.1.EL (RHEL 3)
3041 2.6.12.6 vanilla (kernel.org)
3045 Frequency : quota enabled and large files being deleted
3047 Description: releasing more than 4GB of quota at once hangs OST
3048 Details : If a user deletes more than 4GB of files on a single OST it
3049 will cause the OST to spin in an infinite loop. Release
3050 quota in < 4GB chunks, or use a 64-bit value for 1.4.7.1+.
3055 Description: statfs data retrieved from /proc may be stale or zero
3056 Details : When reading per-device statfs data from /proc, in the
3057 {kbytes,files}_{total,free,avail} files, it may appear
3058 as zero or be out of date.
3061 Frequency : systems with MD RAID1 external journal devices
3063 Description: lconf's call to blkid is confused by RAID1 journal devices
3064 Details : Use the "blkid -l" flag to locate the MD RAID device instead
3065 of returning all block devices that match the journal UUID.
3068 Frequency : always, for aggregate stripe size over 4GB
3070 Description: "lfs setstripe" fails assertion when setting 4GB+ stripe width
3071 Details : Using "lfs setstripe" to set stripe size * stripe count over 4GB
3072 will fail the kernel with "ASSERTION(lsm->lsm_xfersize != 0)"
3075 Frequency : always if "lfs find" used on a local file/directory
3077 Description: "lfs find" segfaults if used on a local file/directory
3078 Details : The case where a directory component was not specified wasn't
3079 handled correctly. Handle this properly.
3082 Frequency : always on ppc64
3084 Description: the write to an ext3 filesystem mounted with mballoc got stuck
3085 Details : ext3_mb_generate_buddy() uses find_next_bit() which does not
3086 perform endianness conversion.
3089 Frequency : rarely (truncate to non-zero file size after write under load)
3090 Bugzilla : 10730, 10687
3091 Description: Files padded with zeros to next 4K multiple
3092 Details : With filesystems mounted using the "extents" option (2.6 kernels)
3093 it is possible that files that are truncated to a non-zero size
3094 immediately after being written are filled with zero bytes beyond
3095 the truncated size. No file data is lost.
3097 Severity : enhancement
3099 Description: Allow recovery/failover for liblustre clients.
3100 Details : liblustre clients were unaware of failover configurations until
3103 Severity : enhancement
3105 Description: user file locks should fail when not mounting with flock option
3106 Details : Set up an error-returning stub in ll_file_operations.lock field
3107 to prevent incorrect behaviour when client is mounted without
3108 flock option. Also, set up properly f_op->flock field for
3112 Frequency : always on ia64
3114 Description: "lfs df" loops on printing out MDS statfs information
3115 Details : The obd_ioctl_data was not initialized and in some systems
3116 this caused a failure during the ioctl that did not return
3117 an error. Initialize the struct and return an error on failure.
3120 Frequency : SLES 9 only
3122 Description: Error of copying files with lustre special EAs as root
3123 Details : Client side always return success for setxattr call for lustre
3124 special xattr (currently only "trusted.lov").
3127 Frequency : rarely on clusters with both ia64+i386 clients
3129 Description: ia64+i686 clients doing shared IO on the same file may LBUG
3130 Details : In rare cases when both ia64+i686 (or other mixed-PAGE_SIZE)
3131 clients are doing concurrent writes to the same file it is
3132 possible that the ia64 clients may LASSERT because the OST
3133 extent locks are not PAGE_SIZE aligned. Ensure that grown
3134 locks are always aligned on the request boundary.
3137 Frequency : specific use, occasional
3139 Description: Overwriting in use executable truncates on-disk binary image
3140 Details : If one node attempts to overwrite an executable in use by
3141 another node, we now correctly return ETXTBSY instead of
3142 truncating the file.
3144 Severity : enhancement
3146 Description: Async OSC create to avoid the blocking unnecessarily.
3147 Details : If a OST has no remain object, system will block on the creating
3148 when need to create a new object on this OST. Now, ways use
3149 pre-created objects when available, instead of blocking on an
3150 empty osc while others are not empty. If we must block, we block
3151 for the shortest possible period of time.
3156 Description: chmod on Lustre root is propagated to other clients
3157 Details : Re-validate root's dentry in ll_lookup_it to avoid having it
3158 invalid by the follow_mount time.
3161 Frequency : liblustre clients only
3163 Description: Race in 'instant cancel' lock handling could lead to such locks
3164 never to be granted in case of SMP MDS
3165 Details : Do not destroy not yet granted but cbpending locks in
3169 Frequency : replay/resend of open
3171 Description: non null lock assetion failure in mds_intent_policy
3172 Details : Trying to replay/resend lockless open requests resulted in
3173 mds_open() returning 0 with no lock. Now it sets a flag if
3174 a lock is going to be returned.
3176 Severity : enhancement
3178 Description: Checksum enhancements
3179 Details : New checksum enhancements allow for resending RPCs that failed
3182 Severity : enhancement
3184 Description: Tunables on number of dirty pages in cacche
3185 Details : Allow to set limit on number of dirty pages cached.
3190 Description: client crash on unmount - lock still has references
3191 Details : In some error handling cases it was possible to leak a lock
3192 reference on a client while accessing a file. This was not
3193 harmful to the client during operation, but would cause the
3194 client to crash when the filesystem is unmounted.
3197 Frequency : specific case, rare
3199 Description: ETXTBSY on mds though file not in use by client
3200 Details : ETXTBSY is no longer incorrectly returned when attempting to
3201 chmod or chown a directory that the user previously tried to
3202 execute or a currently-executing binary.
3205 Frequency : extremely rare except on liblustre-based clients
3207 Description: Lustre space not freed when files are deleted
3208 Details : Clean up open-unlinked files after client eviction. Previously
3209 the unlink was skipped and the files remained as orphans.
3214 Description: OST failure "would be an LBUG" in waiting_locks_callback()
3215 Details : In some cases it was possible to send a blocking callback to
3216 a client doing a glimpse, even though that client didn't get
3217 a lock granted. When the glimpse lock is cancelled on the OST
3218 the freed lock is left on the waiting list and corrupted the list.
3221 Frequency : all core dumps
3223 Description: Broke core dumps to lustre
3224 Details : Negative dentry may be unhashed if parent does not have UPDATE
3225 lock, but some callers, e.g. do_coredump, expect dentry to be
3226 hashed after successful create, hash it in ll_create_it.
3228 ------------------------------------------------------------------------------
3230 2006-09-13 Cluster File Systems, Inc. <info@clusterfs.com>
3232 * Support for kernels:
3233 2.6.9-42.0.2.EL (RHEL 4)
3234 2.6.5-7.276 (SLES 9)
3235 2.4.21-40.EL (RHEL 3)
3236 2.6.12.6 vanilla (kernel.org)
3240 Frequency : always on RHEL 3
3242 Description: Number of open files grows over time
3243 Details : The number of open files grows over time, whether or not
3244 Lustre is started. This was due to a filp leak introduced
3245 by one of our kernel patches.
3247 ------------------------------------------------------------------------------
3249 08-20-2006 Cluster File Systems, Inc. <info@clusterfs.com>
3251 * Support for kernels:
3252 2.6.9-42.EL (RHEL 4)
3253 2.6.5-7.276 (SLES 9)
3254 2.4.21-40.EL (RHEL 3)
3255 2.6.12.6 vanilla (kernel.org)
3260 Bugzilla : 5719, 9635, 9792, 9684
3261 Description: OST (or MDS) trips assertions in (re)connection under heavy load
3262 Details : If a server is under heavy load and cannot reply to new
3263 connection requests before the client resends the (re)connect,
3264 the connection handling code can behave badly if two service
3265 threads are concurrently handing separate (re)connections from
3266 the same client. Add better locking to the connection handling
3267 code, and ensure that only a single connection will be processed
3268 for a given client UUID, even if the lock is dropped.
3270 Severity : enhancement
3272 Description: add TCP zero-copy support to kernel
3273 Details : Add support to the kernel TCP stack to allow zero-copy bulk
3274 sends if the hardware supports scatter-gather and checksumming.
3275 This allows socklnd to do client-write and server-read more
3276 efficiently and reduce CPU utilization from skbuf copying.
3279 Frequency : only if NFS exporting from client
3281 Description: NULL pointer deref in ll_iocontrol() if chattr mknod file
3282 Details : If setting attributes on a file created under NFS that had
3283 never been opened it would be possible to oops the client
3284 if the file had no objects.
3287 Frequency : always for liblustre
3289 Description: liblustre client does MDS+OSTs setattr RPC for each write
3290 Details : When doing a write from a liblustre client, the client
3291 incorrectly issued an RPC to the MDS and each OST the file was
3292 striped over in order to update the timestamps. When writing
3293 with small chunks and many clients this could overwhelm the MDS
3294 with RPCs. In all cases it would slow down the write because
3295 these RPCs are unnecessary.
3297 Severity : enhancement
3299 Description: allow number of MDS service threads to be changed at module load
3300 Details : It is now possible to change the number of MDS service threads
3301 running. Adding "options mds mds_num_threads={N}" to the MDS's
3302 /etc/modprobe.conf will set the number of threads for the next
3303 time Lustre is restarted (assuming the "mds" module is also
3304 reloaded at that time). The default number of threads will
3305 stay the same, 32 for most systems.
3310 Description: OST crash if filesystem is unformatted or corrupt
3311 Details : If an OST is started on a device that has never been formatted
3312 or if the filesystem is corrupt and cannot even mount then the
3313 error handling cleanup routines would dereference a NULL pointer.
3318 Description: NULL pointer deref in llap_from_page.
3319 Details : get_cache_page_nowait can return a page with NULL (or otherwise
3320 incorrect) mapping if the page was truncated/reclaimed while it was
3321 searched for. Check for this condition and skip such pages when
3322 doing readahead. Introduce extra check to llap_from_page() to
3323 verify page->mapping->host is non-NULL (so page is not anonymous).
3326 Frequency : Sometimes when using sys_sendfile
3328 Description: "page not covered by a lock" warnings from ll_readpage
3329 Details : sendfile called ll_readpage without right page locks present.
3330 Now we introduced ll_file_sendfile that does necessary locking
3331 around call to generic_file_sendfile() much like we do in
3335 Frequency : with certain MDS communication failures at client mount time
3337 Description: NULL pointer deref after failed client mount
3338 Details : a client connection request may delayed by the network layer
3339 and not be sent until after the PTLRPC layer has timed out the
3340 request. If the client fails the mount immediately it will try
3341 to clean up before the network times out the request. Add a
3342 reference from the request import to the obd device and delay
3343 the cleanup until the network drops the request.
3346 Frequency : occasionally during client (re)connect
3348 Description: assertion failure during client (re)connect
3349 Details : processing a client connection request may be delayed by the
3350 client or server longer than the client connect timeout. This
3351 causes the client to resend the connection request. If the
3352 original connection request is replied in this interval, the
3353 client may trip an assertion failure in ptlrpc_connect_interpret()
3354 which thought it would be the only running connect process.
3357 Frequency : only with obd_echo servers and clients that are rebooted
3359 Description: kernel BUG accessing uninitialized data structure
3360 Details : When running an obd_echo server it did not start the ping_evictor
3361 thread, and when a client was evicted an uninitialized data
3362 structure was accessed. Start the ping_evictor in the RPC
3363 service startup instead of the OBD startup.
3365 Severity : enhancement
3366 Bugzilla : 10193 (patchless)
3367 Description: Remove dependency on various unexported kernel interfaces.
3368 Details : No longer need reparent_to_init, exit_mm, exit_files,
3369 sock_getsockopt, filemap_populate, FMODE_EXEC, put_filp.
3372 Frequency : rare (only users of deprecated and unsupported LDAP config)
3374 Description: write_conf for zeroconf mount queried LDAP incorrectly for client
3375 Details : LDAP apparently contains 'lustreName' attributes instead of
3376 'name'. A simple remapping of the name is sufficient.
3379 Frequency : rare (only with non-default dump_on_timeout debug enabled)
3381 Description: waiting_locks_callback trips kernel BUG if client is evicted
3382 Details : Running with the dump_on_timeout debug flag turned on makes
3383 it possible that the waiting_locks_callback() can try to dump
3384 the Lustre kernel debug logs from an interrupt handler. Defer
3385 this log dumping to the expired_lock_main() thread.
3387 Severity : enhancement
3389 Description: Support NFS exporting on 2.6 kernels.
3390 Details : Implement non-rawops metadata methods for NFS server to use without
3391 changing NFS server code.
3394 Frequency : very rare (synthetic metadata workload only)
3396 Description: two racing renames might cause an MDS thread to deadlock
3397 Details : Running the "racer" program may cause one MDS thread to rename
3398 a file from being the source of a rename to being the target of
3399 a rename at exactly the same time that another thread is doing
3400 so, and the second thread has already enqueued these locks after
3401 doing a lookup of the target and is trying to relock them in
3402 order. Ensure that we don't try to re-lock the same resource.
3405 Frequency : only very large systems with liblustre clients
3407 Description: slow eviction of liblustre clients with the "evict_by_nid" RPC
3408 Details : Use asynchronous set_info RPCs to send the "evict_by_nid" to
3409 all OSTs in parallel. This allows the eviction of stale liblustre
3410 clients to proceed much faster than if they were done in series,
3411 and also offers similar improvements for other set_info RPCs.
3416 Description: excessive CPU usage during initial read phase on client
3417 Details : During the initial read phase on a client, it would agressively
3418 retry readahead on the file, consuming too much CPU and impacting
3419 performance (since 1.4.5.8). Improve the readahead algorithm
3420 to avoid this, and also improve some other common cases (read
3421 of small files in particular, where "small" is files smaller than
3422 /proc/fs/lustre/llite/*/max_read_ahead_whole_mb, 2MB by default).
3427 Description: MDS crash when receiving packet with unknown intent.
3428 Details : Do not LBUG in unknown intent case, just return -EFAULT
3430 Severity : enhancement
3431 Bugzilla : 9293, 9385
3432 Description: MDS RPCs are serialised on client. This is unnecessary for some.
3433 Details : Do not serialize getattr (non-intent version) and statfs.
3436 Frequency : occasional, when OST network is overloaded/intermittent
3438 Description: client evicted by OST after bulk IO timeout
3439 Details : If a client sends a bulk IO request (read or write) the OST
3440 may evict the client if it is unresposive to its data GET/PUT
3441 request. This is incorrect if the network is overloaded (takes
3442 too long to transfer the RPC data) or dropped the OST GET/PUT
3443 request. There is no need to evict the client at all, since
3444 the pinger and/or lock callbacks will handle this, and the
3445 client can restart the bulk request.
3448 Frequency : Always when mmapping file with no objects
3450 Description: client crashes when mmapping file with no objects
3451 Details : Check that we actually have objects in a file before doing any
3452 operations on objects in ll_vm_open, ll_vm_close and
3458 Description: Request leak when working with deleted CWD
3459 Details : Introduce advanced request refcount tracking for requests
3460 referenced from lustre intent.
3462 Severity : Enhancement
3464 Description: Cache open file handles on client.
3465 Details : MDS now will return special lock along with openhandle, if
3466 requested and client is allowed to hold openhandle, even if unused,
3467 until such a lock is revoked. Helps NFS a lot, since NFS is opening
3468 closing files for every read/write openration.
3470 Severity : Enhancement
3472 Description: Cache open negative dentries on client when possible.
3473 Details : Guard negative dentries with UPDATE lock on parent dir, drop
3474 negative dentries on lock revocation.
3479 Description: Remounting a client read-only wasn't possible with a zconf mount
3480 Details : It wasn't possible to remount a client read-only with llmount.
3482 Severity : enhancement
3483 Description: Include MPICH 1.2.6 Lustre ADIO interface patch
3484 Details : In lustre/contrib/ or /usr/share/lustre in RPM a patch for
3485 MPICH is included to add Lustre-specific ADIO interfaces.
3486 This is based closely on the UFS ADIO layer and only differs
3487 in file creation, in order to allow the OST striping to be set.
3488 This is user-contributed code and not supported by CFS.
3493 Description: extended inode attributes (immutable, append-only) work improperly
3494 when 2.4 and 2.6 kernels are used on client/server or vice versa
3495 Details : Introduce kernel-independent values for these flags.
3497 Severity : enhancement
3500 Description: Allow fractional MB tunings for lustre in /proc/ filesystem.
3501 Details : Many of the /proc/ tunables can only be tuned at a megabyte
3502 granularity. Now, Fractional MB granularity is be supported,
3503 this is very useful for low memory system.
3505 Severity : enhancement
3507 Description: Getattr by fid
3508 Details : Getting a file attributes by its fid, obtaining UPDATE|LOOKUP
3509 locks, avoids extra getattr rpc requests to MDS, allows '/' to
3510 have locks and avoids getattr rpc requests for it on every stat.
3513 Frequency : Always, for filesystems larger than 2TB
3515 Description: ldiskfs crash at mount for filesystem larger than 2TB with mballoc
3516 Details : Kenrel kmalloc limits allocations to 128kB and this prevents
3517 filesystems larger than 2TB to be mounted with mballoc enabled.
3520 Frequency : Always, for 32-bit kernel without CONFIG_LBD and filesystem > 2TB
3522 Description: ldiskfs crash at mount for filesystem larger than 2TB with mballoc
3523 Details : If a 32-bit kernel is compiled without CONFIG_LBD enabled and a
3524 filesystems larger than 2TB is mounted then the kernel will
3525 silently corrupt the start of the filesystem. CONFIG_LBD is
3526 enabled for all CFS-supported kernels, but the possibility of
3527 this happening with a modified kernel config exists.
3529 Severity : enhancement
3531 Description: add client O_DIRECT support for 2.6 kernels
3532 Details : It is now possible to do O_DIRECT reads and writes to files
3533 in the Lustre client mountpoint on 2.6 kernel clients.
3535 Severity : enhancement
3537 Description: parallel glimpse, setattr, statfs, punch, destroy requests
3538 Details : Sends glimpse, setattr, statfs, punch, destroy requests to OSTs in
3539 parallel, not waiting for response from every OST before sending
3540 a rpc to the next OST.
3545 Description: setattr vs write race when updating file timestamps
3546 Details : Client processes that update a file timestamp into the past
3547 right after writing to the file (e.g. tar) it is possible that
3548 the updated file modification time can be reset to the current
3549 time due to a race between processing the setattr and write RPC.
3551 Severity : enhancement
3553 Description: Bring 'lfs find' closer in line with regular Linux find.
3554 Details : lfs find util supports -atime, -mtime, -ctime, -maxdepth, -print,
3555 -print0 options and obtains all the needed info through the lustre
3558 Severity : enhancement
3560 Description: support up to 1024 configured devices on one node
3561 Details : change obd_dev array from statically allocated to dynamically
3562 allocated structs as they are first used to reduce memory usage
3567 Description: Flush dirty partially truncated pages during truncate
3568 Details : Immediatelly flush partially truncated pages in filter_setattr,
3569 this way we completely avoid having any pages in page cache on OST
3570 and can retire ugly workarounds during writes to flush such pages.
3575 Description: i_sem vs transaction deadlock in mds_obd_destroy during unlink.
3576 Details : protect inode from truncation within vfs_unlink() context
3577 just take a reference before calling vfs_unlink() and release it
3578 when parent's i_sem is free.
3583 Description: last_id value checked outside lock on OST caused LASSERT failure
3584 Details : If there were multiple MDS->OST object precreate requests in
3585 flight, it was possible that the OST's last object id was checked
3586 outside a lock and incorrectly tripped an assertion. Move checks
3587 inside locks, and discard old precreate requests.
3590 Frequency : always, if extents are used on OSTs
3592 Description: index ei_leaf_hi (48-bit extension) is not zeroed in extent index
3593 Details : OSTs using the extents format would not zero the high 16 bits of
3594 the index physical block number. This is not a problem for any
3595 OST filesystems smaller than 16TB, and no kernels support ext3
3596 filesystems larger than 16TB yet. This is fixed in 1.4.7 (all
3597 new/modified files) and can be fixed for existing filesystems
3598 with e2fsprogs-1.39-cfs1.
3603 Description: import connection selection may be incorrect if timer wraps
3604 Details : Using a 32-bit jiffies timer with HZ=1000 may cause backup
3605 import connections to be ignored if the 32-bit jiffies counter
3606 wraps. Use a 64-bit jiffies counter.
3609 Frequency : very large clusters immediately after boot
3611 Description: LNET request buffers exhausted under heavy short-term load
3612 Details : If a large number of client requests are generated on a service
3613 that has previously never seen so many requests it is possible
3614 that the request buffer growth cannot keep up with the spike in
3615 demand. Instead of dropping incoming requests, they are held in
3616 the LND until the RPC service can accept more requests.
3619 Frequency : Sometimes during replay
3621 Description: Assertion failure in ll_local_open after replay.
3622 Details : If replay happened on an open request reply before we were able
3623 to set replay handler, reply will become not swabbed tripping the
3624 assertion in ll_local_open. Now we set the handler right after
3625 recognising of open request
3628 Frequency : very rare
3630 Description: kernel reports "badness in vsnprintf"
3631 Details : Reading from the "recovery_status" /proc file in small chunks
3632 may cause a negative length in lprocfs_obd_rd_recovery_status()
3633 call to vsnprintf() (which is otherwise harmless). Exit early
3634 if there is no more space in the output buffer.
3636 Severity : enhancement
3638 Description: clear OBD RPC statistics by writing to them
3639 Details : It is now possible to clear the OBD RPC statistics by writing
3640 to the "stats" file.
3645 Description: Client mtime is not the same on different clients after utimes
3646 Details : In some cases, the client was using the utimes() syscall on
3647 a file cached on another node. The clients now validate the
3648 ctime from the MDS + OSTs to determine which one is right.
3653 Description: Inability to activate failout mode
3654 Details : lconf script incorrectly assumed that in pythong string's numeric
3655 value is used in comparisons.
3658 Frequency : always with multiple stripes per file
3660 Description: Inefficient object allocation for mutli-stripe files
3661 Details : When selecting which OSTs to stripe files over, for files with
3662 a stripe count that divides evenly into the number of OSTs,
3663 the MDS is always picking the same starting OST for each file.
3664 Return the OST selection heuristic to the original design.
3669 Description: mount failures may take full timeout to return an error
3670 Details : Under some heavy load conditions it is possible that a
3671 failed mount can wait for the full obd_timeout interval,
3672 possibly several minutes, before reporting an error.
3673 Instead return an error as soon as the status is known.
3675 Frequency : quota enabled and large files being deleted
3677 Description: releasing more than 4GB of quota at once hangs OST
3678 Details : If a user deletes more than 4GB of files on a single OST it
3679 will cause the OST to spin in an infinite loop. Release
3680 quota in < 4GB chunks, or use a 64-bit value for 1.4.7.1+.
3685 Description: statfs data retrieved from /proc may be stale or zero
3686 Details : When reading per-device statfs data from /proc, in the
3687 {kbytes,files}_{total,free,avail} files, it may appear
3688 as zero or be out of date.
3691 Frequency : systems with MD RAID1 external journal devices
3693 Description: lconf's call to blkid is confused by RAID1 journal devices
3694 Details : Use the "blkid -l" flag to locate the MD RAID device instead
3695 of returning all block devices that match the journal UUID.
3698 Frequency : always, for aggregate stripe size over 4GB
3700 Description: assertion fails when trying to use 4GB stripe size
3701 Details : Use "setstripe" to set stripe size over 4GB will fail the kernel,
3702 complaining "ASSERTION(lsm->lsm_xfersize != 0)"
3705 Frequency : always on ppc64
3707 Description: the first write on an ext3 filesystem with mballoc got stuck
3708 Details : ext3_mb_generate_buddy() uses find_next_bit() which does not
3709 perform endianness conversion.
3711 ------------------------------------------------------------------------------
3713 02-14-2006 Cluster File Systems, Inc. <info@clusterfs.com>
3715 * WIRE PROTOCOL CHANGE. This version of Lustre networking WILL NOT
3716 INTEROPERATE with older versions automatically. Please read the
3717 user documentation before upgrading any part of a live system.
3718 * WARNING: Lustre networking configuration changes are required with
3719 this release. See https://bugzilla.clusterfs.com/show_bug.cgi?id=10052
3722 * Support for kernels:
3723 2.6.9-22.0.2.EL (RHEL 4)
3724 2.6.5-7.244 (SLES 9)
3725 2.6.12.6 vanilla (kernel.org)
3728 Severity : enhancement
3729 Bugzilla : 7981/8208
3730 Description: Introduced Lustre Networking (LNET)
3731 Details : LNET is new networking infrastructure for Lustre, it includes
3732 a reorganized network configuration mode (see the user
3733 documentation for full details) as well as support for routing
3734 between different network fabrics. Lustre Networking Devices
3735 (LNDs) for the supported network fabrics have also been
3736 created for this new infrastructure.
3738 Severity : enhancement
3739 Description: Introduced Access control lists
3740 Details : clients can set ACLs on files and directories in order to have
3741 more fine-grained permissions than the standard Unix UGO+RWX.
3742 The MDS must be started with the "-o acl" mount option.
3744 Severity : enhancement
3745 Description: Introduced filesystem quotas
3746 Details : Administrators may now establish per-user quotas on the
3749 Severity : enhancement
3751 Description: Configuration change for the XT3
3752 The PTLLND is now used to run Lustre over Portals on the XT3
3753 The configure option(s) --with-cray-portals are no longer used.
3754 Rather --with-portals=<path-to-portals-includes> is used to
3755 enable building on the XT3. In addition to enable XT3 specific
3756 features the option --enable-cray-xt3 must be used.
3761 Description: Running on many-way SMP OSTs can trigger oops in llcd_send()
3762 Details : A race between allocating a new llcd and re-getting the llcd_lock
3763 allowed another thread to grab newly-allocated llcd.
3765 Severity : enhancement
3767 Description: 2.6 OST async journal commit and locking fix to improve performance
3768 Details : The filter_direct_io()+filter_commitrw_write() journal commits for
3769 2.6 kernels are now async as they already were in 2.4 kernels so
3770 that they can commit concurrently with the network bulk transfer.
3771 For block-allocated files the filter allocation semaphore is held
3772 to avoid filesystem fragmentation during allocation. BKL lock
3773 removed for 2.6 xattr operations where it is no longer needed.
3778 Description: lconf incorrectly determined whether two IP networks could talk
3779 Details : In some more complicated routing and multiple-network
3780 configurations, lconf will avoid trying to make a network
3781 connection to a disjoint part of the IP space. It was doing the
3782 math incorrectly for one set of cases.
3787 Description: Fix for potential infinite loop processing records in an llog.
3788 Details : If an llog record is corrupted/zeroed, it is possible to loop
3789 forever in llog_process(). Validate the llog record length
3790 and skip the remainder of the block on error.
3793 Frequency : occasional (liblustre only)
3795 Description: liblustre could not open files whose last component is a symlink
3796 Details : sysio_path_walk() would incorrectly pass the open intent to
3797 intermediate path components.
3800 Frequency : rare (liblustre only with non-standard tuning)
3801 Bugzilla : 7201 (7350)
3802 Description: Tuning the MDC DLM LRU size to zero triggers client LASSERT
3803 Details : llu_lookup_finish_locks() tries to set lock data on a lock
3804 after it has been released, only do this for referenced locks
3806 Severity : enhancement
3808 Description: specifying an (invalid) directory default stripe_size of -1
3809 would reset the directory default striping
3810 Details : stripe_size -1 was used internally to signal directory stripe
3811 removal, now use "all default" to signal dir stripe removal
3812 as a directory striping of "all default" is not useful
3815 Frequency : common for large clusters running liblustre clients
3817 Description: doing an ls when liblustre clients are running is slow
3818 Details : sending a glimpse AST to a liblustre client waits for every AST
3819 to time out, as liblustre clients will not respond. Since they
3820 cannot cache data we refresh the OST lock LVB from disk instead.
3822 Severity : enhancement
3824 Description: doing an ls at the same time as file IO can be slow
3825 Details : enqueue and other "small" requests can be blocked behind many
3826 large IO requests. Create a new OST IO portal for non-IO
3827 requests so they can be processed faster.
3830 Frequency : rare (only HPUX clients mounting unsupported re-exported NFS vol)
3832 Description: an HPUX NFS client would get -EACCESS when ftruncate()ing a newly
3833 created file with mode 000
3834 Details : the Linux NFS server relies on an MDS_OPEN_OWNEROVERRIDE hack to
3835 allow an ftruncate() as a non-root user to a file with mode 000.
3836 Lustre now respects this flag to disable mode checks when
3837 truncating a file owned by the user
3840 Frequency : liblustre-only, when liblustre client dies unexpectedly or becomes
3843 Description: Revoking locks from clients that went dead or catatonic might take
3845 Details : New lock flags FL_CANCEL_ON_BLOCK used by liblustre makes
3846 cancellation of such locks instant on servers without waiting for
3847 any reply from clients. Clients drops these locks when cancel
3848 notification from server is received without replying.
3851 Frequency : liblustre-only, when liblustre client dies or becomes busy
3853 Description: Doing ls on Linux clients can take a long time with active
3855 Details : Liblustre client cannot handle ASTs in timely manner, so avoid
3856 granting such locks to it in the first place if possible. Locks
3857 are taken by proxy on the OST during the read or write and
3858 dropped immediately afterward. Add connect flags handling, do
3859 not grant locks to liblustre clients for glimpse ASTs.
3861 Severity : enhancement
3863 Description: Improve read-ahead algorithm to avoid excessive IO for random reads
3864 Details : Existing read-ahead algorithm is tuned for the case of streamlined
3865 sequential reads and behaves badly with applications doing random
3866 reads. Improve it by reading ahead at least read region, and
3867 avoiding excessive large RPC for small reads.
3869 Severity : enhancement
3871 Description: Creating more than 1000 files for a single job may cause a load
3872 imbalance on the OSTs if there are also a large number of OSTs.
3873 Details : qos_prep_create() uses an OST index reseed value that is an
3874 even multiple of the number of available OSTs so that if the
3875 reseed happens in the middle of the object allocation it will
3876 still utilize the OSTs as uniformly as possible.
3881 Description: OST or MDS may oops in ping_evictor_main()
3882 Details : ping_evictor_main() drops obd_dev_lock if deleting a stale export
3883 but doesn't restart at beginning of obd_exports_timed list
3886 Severity : enhancement
3888 Description: improve by-nid export eviction on the MDS and OST
3889 Details : allow multiple exports with the same NID to be evicted at one
3890 time without re-searching the exports list.
3893 Frequency : rare, only with supplementary groups enabled on SMP 2.6 kernels
3895 Description: MDS may oops in groups_free()
3896 Details : in rare race conditions a newly allocated group_info struct is
3897 freed again, and this can be NULL. The 2.4 compatibility code
3898 for groups_free() checked for a NULL pointer, but 2.6 did not.
3901 Frequency : common for liblustre clients doing little filesystem IO
3902 Bugzilla : 9352, 7313
3903 Description: server may evict liblustre clients accessing contended locks
3904 Details : if a client is granted a lock or receives a completion AST
3905 with a blocking AST already set it would not reply to the AST
3906 for LDLM_FL_CANCEL_ON_BLOCK locks. It now replies to such ASTs.
3909 Frequency : lfs setstripe, only systems with more than 160 OSTs
3911 Description: unable to set striping with a starting offset beyond OST 160
3912 Details : llapi_create_file() incorrectly limited the starting stripe
3913 index to the maximum single-file stripe count.
3916 Frequency : LDAP users only
3918 Description: lconf did not handle in-kernel recovery with LDAP properly
3919 Details : lconf/LustreDB get_refs() is searching the wrong namespace
3921 Severity : enhancement
3923 Description: bind OST threads to NUMA nodes to improve performance
3924 Details : all OST threads are uniformly bound to CPUs on a single NUMA
3925 node and do their allocations there to localize memory access
3927 Severity : enhancement
3929 Description: llmount can determine client NID directly from Myrinet (GM)
3930 Details : the client NID code from gmnalnid was moved directly into
3931 llmount, removing the need to use this or specifying the
3932 client NID explicitly when mounting GM clients with zeroconf
3935 Frequency : if client is started with down MDS
3937 Description: if client is started with down MDS mount hangs in ptlrpc_queue_wait
3938 Details : Having an LWI_INTR() wait event (interruptible, but no timeout)
3939 will wait indefinitely in ptlrpc_queue_wait->l_wait_event() after
3940 ptlrpc_import_delayed_req() because we didn't check if the
3941 request was interrupted, and we also didn't break out of the
3942 event loop if there was no timeout
3947 Description: data loss during non-page-aligned writes to a single file from
3948 both multiple nodes and multiple threads on one node at same time
3949 Details : updates to KMS and lsm weren't protected by common lock. Resulting
3950 inconsistency led to false short-reads, that were cached and later
3951 used by ->prepare_write() to fill in partially written page,
3952 leading to data loss.
3955 Frequency : always, if lconf --abort_recovery used
3957 Description: lconf --abort_recovery fails with 'Operation not supported'
3958 Details : lconf was attempting to abort recovery on the MDT device and not
3961 Severity : enhancement
3963 Description: remove cleanup logs
3964 Details : replace lconf-generated cleanup logs with lustre internal
3965 cleanup routines. Eliminates the need for client-cleanup and
3968 Severity : enhancement
3970 Description: add support for EAs (user and system) on lustre filesystems
3971 Details : it is now possible to store extended attributes in the Lustre
3972 client filesystem, and with the user_xattr mount option it
3973 is possible to allow users to store EAs on their files also
3975 Severity : enhancement
3977 Description: Add possibility (config option) to show minimal available OST free
3979 Details : When compiled with --enable-mindf configure option, statfs(2)
3980 (and so, df) will return least minimal free space available from
3981 all OSTs as amount of free space on FS, instead of summary of
3982 free spaces of all OSTs.
3984 Severity : enhancement
3986 Description: do not expand extent locks acquired on OST-side
3987 Details : Modify ldlm_extent_policy() to not expand local locks, acquired
3988 by server: they are not cached anyway.
3991 Frequency : when mmap is used/binaries executed from Lustre
3993 Description: Unmmap pages before throwing them away from read cache.
3994 Details : llap_shrink cache now attempts to unmap pages before discarding
3995 them (if unmapping failed - do not discard). SLES9 kernel has
3996 extra checks that trigger if this unmapping is not done first.
4001 Description: lconf didn't resolve symlinks before checking to see whether a
4002 given mountpoint was already in use
4005 Frequency : when migrating failover services
4006 Bugzilla : 6395, 9514
4007 Description: When migrating a subset of services from a node (e.g. failback
4008 from a failover service node) the remaining services would
4009 time out and evict clients.
4010 Details : lconf --force (implied by --failover) sets the global obd_timeout
4011 to 5 seconds in order to quickly disconnect, but this caused
4012 other RPCs to time out too quickly. Do not change the global
4013 obd_timeout for force cleanup, only set it for DISCONNECT RPCs.
4015 Severity : enhancement
4016 Frequency : if MDS is started with down OST
4017 Bugzilla : 9439,5706
4018 Description: Allow startup/shutdown of an MDS without depending on the
4019 availability of the OSTs.
4020 Details : Asynchronously call mds_lov_synchronize during MDS startup.
4021 Add appropriate locking and lov-osc refcounts for safe
4022 cleaning. Add osc abort_inflight calls in case the
4023 synchronize never started.
4026 Frequency : occasional (Cray XT3 only)
4028 Description: root not authorized to access files in CRAY_PORTALS environment
4029 Details : The client process capabilities were not honoured on the MDS in
4030 a CRAY_PORTALS/CRAY_XT3 environment. If the file had previously
4031 been accessed by an authorized user then root was able to access
4032 the file on the local client also. The root user capabilities
4033 are now allowed on the MDS, as this environment has secure UID.
4036 Frequency : occasional
4038 Description: ldiskfs "too long searching" message happens too often
4039 Details : A debugging message (otherwise harmless) prints too often on
4040 the OST console. This has been reduced to only happen when
4041 there are fragmentation problems on the filesystem.
4046 Description: Division by zero in statfs when all OSCs are inactive
4047 Details : lov_get_stripecnt() returns zero due to incorrect order of checks,
4048 lov_statfs divides by value returned by lov_get_stripecnt().
4052 Bugzilla : 9489, 3273
4053 Description: First write from each client to each OST was only 4kB in size,
4054 to initialize client writeback cache, which caused sub-optimal
4055 RPCs and poor layout on disk for the first writen file.
4056 Details : Clients now request an initial cache grant at (re)connect time
4057 and so that they can start streaming writes to the cache right
4058 away and always do full-sized RPCs if there is enough data.
4059 If the OST is rebooted the client also re-establishes its grant
4060 so that client cached writes will be honoured under the grant.
4065 Description: Slow ls (and stat(2) syscall) on files residing on IO-loaded OSTs
4066 Details : Now I/O RPCs go to different portal number and (presumably) fast
4067 lock requests (and glimses) and other RPCs get their own service
4068 threads pool that should be able to service those RPCs
4071 Severity : enhancement
4073 Description: Ability to exchange lustre version between client and servers and
4074 issue warnings at client side if client is too old. Also for
4075 liblustre clients there is ability to refuse connection of too old
4077 Details : New 'version' field is added to connect data structure that is
4078 filled with version info. That info is later checked by server and
4082 Frequency : rare, liblustre only.
4083 Bugzilla : 9296, 9581
4084 Description: Two simultaneous writes from liblustre at offset within same page
4085 might proceed at the same time overwriting eachother with stale
4087 Details : I/O lock withing llu_file_prwv was released too early, before data
4088 actually was hitting the wire. Extended lock-holding time until
4089 server acknowledges receiving data.
4092 Frequency : extremely rare. Never observed in practice.
4094 Description: avoid generating lustre_handle cookie of 0.
4095 Details : class_handle_hash() generates handle cookies by incrementing
4096 global counter, and can hit 0 occasionaly (this is unlikely, but
4097 not impossible, because initial value of cookie counter is
4098 selected randonly). Value of 0 is used as a sentinel meaning
4099 "unassigned handle" --- avoid it. Also coalesce two critical
4100 sections in this function into one.
4102 Severity : enhancement
4104 Description: allow liblustre clients to delegate truncate locking to OST
4105 Details : To avoid overhead of locking, liblustre client instructs OST to
4106 take extent lock in ost_punch() on client's behalf. New connection
4107 flag is added to handle backward compatibility.
4109 Severity : enhancement
4110 Bugzilla : 4928, 7341, 9758
4111 Description: allow number of OST service threads to be specified
4112 Details : a module parameter allows the number of OST service threads
4113 to be specified via "options ost ost_num_threads={N}" in the
4114 OSS's /etc/modules.conf or /etc/modprobe.conf.
4118 Bugzilla : 6146, 9635, 9895
4119 Description: servers crash with bad pointer in target_handle_connect()
4120 Details : In rare cases when a client is reconnecting it was possible that
4121 the connection request was the last reference for that export.
4122 We would temporarily drop the export reference and get a new
4123 one, but this may have been the last reference and the export
4124 was just destroyed. Get new reference before dropping old one.
4126 Severity : enhancement
4127 Frequency : if client is started with failover MDS
4129 Description: Allow multiple MDS hostnames in the mount command
4130 Details : Try to read the configuration from all specified MDS
4131 hostnames during a client mount in case the "primary"
4134 Severity : enhancement
4136 Description: Stop sending data to evicted clients as soon as possible.
4137 Details : Check if the client we are about to send or are sending data to
4138 was evicted already. (Check is done every second of waiting,
4139 for which l_wait_event interface was extended to allow checking
4140 of exit condition at specified intervals).
4143 Frequency : rare, normally only when NFS exporting is done from client
4145 Description: 'bad disk LOV MAGIC: 0x00000000' error when chown'ing files
4147 Details : Make mds_get_md() recognise empty md case and set lmm size to 0.
4150 Frequency : always, if srand() is called before liblustre initialization
4152 Description: Liblustre uses system PRNG disturbing its usage by user application
4153 Details : Introduce internal to lustre fast and high-quality PRNG for
4154 lustre usage and make liblustre and some other places in generic
4155 lustre code to use it.
4157 Severity : enhancement
4158 Bugzilla : 9477, 9557, 9870
4159 Description: Verify that the MDS configuration logs are updated when xml is
4160 Details : Check if the .xml configuration logs are newer than the config
4161 logs stored on the MDS and report an error if this is the case.
4162 Request --write-conf, or allow starting with --old_conf.
4164 Severity : enhancement
4166 Description: Handle symlinks in the path when checking if Lustre is mounted.
4167 Details : Resolve intermediate symlinks when checking if a client has
4168 mounted a filesystem to avoid duplicate client mounts.
4173 Description: lconf can hit an error exception but still return success.
4174 Details : The lconf command catches the Command error exception at the top
4175 level script context and will exit with the associated exit
4176 status, but doesn't ensure that this exit status is non-zero.
4181 Description: failure of ptlrpc thread startup can cause oops
4182 Details : Starting a ptlrpc service thread can fail if there are a large
4183 number of threads or the server memory is very fragmented.
4184 Handle this without oopsing.
4187 Frequency : always, only if liblustre and non-default acceptor port was used
4189 Description: liblustre cannot connect to servers with non-default acceptor port
4190 Details : tcpnal_set_default_params() was not called and was therefore
4191 ignoring the environment varaible TCPNAL_PORT, as well as other
4192 TCPNAL_ environment variables
4197 Description: two objects could be created on the same OST for a single file
4198 Details : If an OST is down, in some cases it was possible to create two
4199 objects on a single OST for a single file. No problems other
4200 than potential performance impact and spurious error messages.
4204 Bugzilla : 5681, 9562
4205 Description: Client may oops in ll_unhash_aliases
4206 Details : Client dcache may become inconsistent in race condition.
4207 In some cases "getcwd" can fail if the current directory is
4213 Description: Inode refcounting problems in NFS export code
4214 Details : link_raw functions used to call d_instantiate without obtaining
4215 extra inode reference first.
4219 Bugzilla : 9942, 9903
4220 Description: Referencing freed requests leading to crash, memleaks with NFS.
4221 Details : We used to require that call to ll_revalidate_it was always
4222 followed by ll_lookup_it. Also with revalidate_special() it is
4223 possible to call ll_revalidate_it() twice for the same dentry
4224 even if first occurence returned success. This fix changes semantic
4225 between DISP_ENQ_COMPLETE disposition flag to mean there is extra
4226 reference on a request referred from the intent.
4227 ll_intent_release() then releases such a request.
4230 Frequency : rare, normally benchmark loads only
4232 Description: unlinked inodes were kept in memory on the client
4233 Details : If a client is repeatedly creating and unlinking files it
4234 can accumulate a lot of stale inodes in the inode slab cache.
4235 If there is no other client load running this can cause the
4236 client node to run out of memory. Instead flush old inodes
4237 from client cache that have the same inode number as a new inode.
4240 Frequency : SLES9 2.6.5 kernel and long filenames only
4241 Bugzilla : 9969, 10379
4242 Description: utime reports stale NFS file handle
4243 Details : SLES9 uses out-of-dentry names in some cases, which confused
4244 the lustre dentry revalidation. Change it to always use the
4248 Frequency : rare, unless heavy write-truncate concurrency is continuous
4249 Bugzilla : 4180, 6984, 7171, 9963, 9331
4250 Description: OST becomes very slow and/or deadlocked during object unlink
4251 Details : filter_destroy() was holding onto the parent directory lock
4252 while truncating+unlinking objects. For very large objects this
4253 may block other threads for a long time and slow overall OST
4254 responsiveness. It may also be possible to get a lock ordering
4255 deadlock in this case, or run out of journal credits because of
4256 the combined truncate+unlink. Solution is to do object truncate
4257 first in one transaction without parent lock, and then do the
4258 final unlink in a new transaction with the parent lock. This
4259 reduces the lock hold time dramatically.
4262 Frequency : rare, 2.4 kernels only
4264 Description: MDS or OST cleanup may trip kernel BUG when dropping kernel lock
4265 Details : mds_cleanup() and filter_cleanup() need to drop the kernel lock
4266 before unmounting their filesystem in order to avoid deadlock.
4267 The kernel_locked() function in 2.4 kernels only checks whether
4268 the kernel lock is held, not whether it is this process that is
4269 holding it as 2.6 kernels do.
4274 Description: MDS or OST may oops/LBUG if a client is connecting multiple times
4275 Details : The client ptlrpc code may be trying to reconnect to a down
4276 server before a previous connection attempt has timed out.
4277 Increase the reconnect interval to be longer than the connection
4278 timeout interval to avoid sending duplicate connections to
4282 Frequency : echo_client brw_test command
4284 Description: fix echo_client to work with OST preallocated code
4285 Details : OST preallocation code (5137) didn't take echo_client IO path
4286 into account: echo_client calls filter methods outside of any
4287 OST thread and, hence, there is no per-thread preallocated
4288 pages and buffers to use. Solution: hijack pga pages for IO. As
4289 a byproduct, this avoids unnecessary data copying.
4293 Bugzilla : 3555, 5962, 6025, 6155, 6296, 9574
4294 Description: Client can oops in mdc_commit_close() after open replay
4295 Details : It was possible for the MDS to return an open request with no
4296 transaction number in mds_finish_transno() if the client was
4297 evicted, but without actually returning an error. Clients
4298 would later try to replay that open and may trip an assertion
4299 Simplify the client close codepath, and always return an error
4300 from the MDS in case the open is not successful.
4303 Frequency : rare, 2.6 OSTs only
4305 Description: OST may deadlock under high load on fragmented files
4306 Details : If there was a heavy load and highly-fragmented OST filesystems
4307 it was possible to have all the OST threads deadlock waiting on
4308 allocation of biovecs, because the biovecs were not released
4309 until the entire RPC IO was completed. Instead, release biovecs
4310 as soon as they are complete to ensure forward IO progress.
4312 Severity : enhancement
4314 Description: Support for specifying external journal device at mount
4315 Details : If an OST or MDS device is formatted with an external journal
4316 device, this device major/minor is stored in the ext3 superblock
4317 and may not be valid for failover. Allow detecting and
4318 specifying the external journal at mount time.
4323 Description: Mounting an MDS with pending unlinked files may cause oops
4324 Details : target_finish_recovery() calls mds_postrecov() which returned
4325 the number of orphans unlinked. mds_lov_connect->mds_postsetup()
4326 considers this an error and immediately begins cleaning up the
4327 lov, just after starting the mds_lov process
4329 Severity : enhancement
4331 Description: Implement 'lfs df' to report actual free space on per-OST basis
4332 Details : Add sub-command 'df' on 'lfs' to report the disk space usage of
4333 MDS/OSDs. Usage: lfs df [-i][-h]. Command Options: '-i' to report
4334 usage of objects; '-h' to report in human readable format.
4336 ------------------------------------------------------------------------------
4338 08-26-2005 Cluster File Systems, Inc. <info@clusterfs.com>
4345 Description: Mounting an ldiskfs file system with mballoc may crash OST node.
4346 Details : ldiskfs mballoc code may reference an uninitialized buddy struct
4347 at startup during orphan unlinking. Instead, skip buddy update
4348 before setup, as it will be regenerated after recovery is complete.
4353 Description: If an OST is inactive, its locks might reference stale inodes.
4354 Details : lov_change_cbdata() must iterate over all namespaces, even if
4355 they are inactive to clear inode references from the lock.
4357 Severity : enhancement
4358 Frequency : occasional, if non-standard max_dirty_mb used
4360 Description: Client will block write RPCs if not enough grant
4361 Details : If a client has max_dirty_mb smaller than max_rpcs_in_flight,
4362 then the client will block writes while waiting for another RPC
4363 to complete instead of consuming its dirty limit. With change
4364 we get improved performance when max_dirty_mb is small.
4366 Severity : enhancement
4367 Bugzilla : 3389, 6253
4368 Description: Add support for supplementary groups on the MDS.
4369 Details : The MDS has an upcall /proc/fs/lustre/mds/{mds}/group_upcall
4370 (set to /usr/sbin/l_getgroups if enabled) which will do MDS-side
4371 lookups for user supplementary groups into a cache.
4375 Description: O_CREAT|O_EXCL open flags in liblustre always return -EEXIST
4376 Details : Make libsysio to not enforce O_EXCL by clearing the flag,
4377 for liblustre O_EXCL is enforced by MDS.
4381 Description: readdir never returns NULL in liblustre.
4382 Details : Corrected llu_iop_getdirentries logic, to return offset of next
4383 dentry in struct dirent.
4387 Frequency : liblustre only, depends on application IO pattern
4388 Description: liblustre clients evicted if not contacting servers
4389 Details : Don't put liblustre clients into the ping_evictor list, so
4390 they will not be evicted by the pinger ever.
4392 Severity : enhancement
4394 Description: Add ability to evict clients by NID from MDS.
4395 Details : By echoing "nid:$NID" string into
4396 /proc/fs/lustre/mds/.../evict_client client with nid that equals to
4397 $NID would be instantly evicted from this MDS and from all active
4398 OSTs connected to it.
4402 Description: Do not query file size twice, somewhat slowing stat(2) calls.
4403 Details : lookup_it_finish() used to query file size from OSTs that was not
4408 Description: service threads change working directory to that of init
4409 Details : Starting lustre service threads may pin the working directory
4410 of the parent thread, making that filesystem busy. Threads
4411 now change to the working directory of init to avoid this.
4415 Frequency : during shutdown only
4416 Description: shutdown with a failed MDS or OST can cause unmount to hang
4417 Details : Don't resend DISCONNECT messages in ptlrpc_disconnect_import()
4422 Frequency : 2.6 only
4423 Description: chmod/chown may include an extra supplementary group
4424 Details : ll{,u}_mdc_pack_op_data() does not properly initialize the
4425 supplementary group and if none is specified this is used.
4428 Bugzilla : 5479 (6816)
4430 Description: Racing open + rm can assert client in mdc_set_open_replay_data()
4431 Details : If lookup is in progress on a file that is unlinked we might try
4432 to revalidate the inode and fail in revalidate after lookup is
4433 complete and ll_file_open() enqueues the open again but
4434 it_open_error() was not checking DISP_OPEN_OPEN errors correctly.
4437 Frequency : always, if lconf --abort_recovery used
4439 Description: lconf --abort_recovery fails with 'Operation not supported'
4440 Details : lconf was attempting to abort recovery on the MDT device and not
4443 ------------------------------------------------------------------------------
4445 2005-08-08 Cluster File Systems, Inc. <info@clusterfs.com>
4450 Frequency : rare (only unsupported configurations with a node running as an
4452 Bugzilla : 6514, 5137
4453 Description: Mounting a Lustre file system on a node running as an OST could
4455 Details : OSTs now preallocates memory needed to write out data at
4456 startup, instead of when needed, to avoid having to
4457 allocate memory in possibly low memory situations.
4458 Specifically, if the file system is mounted on on OST,
4459 memory pressure could force it to try to write out data,
4460 which it needed to allocate memory to do. Due to the low
4461 memory, it would be unable to do so and the node would
4462 become unresponsive.
4464 Severity : enhancement
4466 Description: Addition of lconf --service command line option
4467 Details : lconf now accepts a '--service <arg>' option, which is
4468 shorthand for 'lconf --group <arg> --select <arg>=<hostname>'
4470 Severity : enhancement
4472 Description: Failover mode is now the default for OSTs.
4473 Details : By default, OSTs will now run in failover mode. To return to
4474 the old behaviour, add '--failout' to the lmc line for OSTs.
4476 Severity : enhancement
4478 Description: Health checks are now provided for MDS and OSTs
4479 Details : Additional detailed health check information on MSD and OSTs
4480 is now provided through the procfs health_check value.
4483 Frequency : occasional, depends on IO load
4485 Description: Disk fragmentation on the OSTs could eventually cause slowdowns
4486 after numerous create/delete cycles
4487 Details : The ext3 inode allocation policy would not allocate new inodes
4488 very well on the OSTs because there are no new directories
4489 being created. Instead we look for groups with free space if
4490 the parent directories are nearly full.
4495 Description: Network or server problems during mount may cause partially
4496 mounted clients instead of returning an error.
4497 Details : The config llog parsing code may overwrite the error return
4498 code during mount error handling, returning success instead
4504 Description: MDS can fail to allocate large reply buffers
4505 Details : After long uptimes the MDS can fail to allocate large reply
4506 buffers (e.g. zconf client mount config records) due to memory
4507 fragmentation or consumption by the buffer cache. Preallocate
4508 some large reply buffers so that these replies can be sent even
4509 under memory pressure.
4513 Frequency : rare (liblustre)
4514 Description: fsx running with liblustre complained that using truncate() to
4515 extend the file doesn't work. This patch corrects that issue.
4516 Details : This is the liblustre equivalent of the fix for bug 6196. Fixes
4517 ATTR_SIZE and lsm use in llu_setattr_raw.
4521 Frequency : rare, only 2.6 kernels
4522 Description: Unusual file access patterns on the MDS may result in inode
4523 data being lost in very rare circumstances.
4524 Details : Bad interaction between the ea-in-inode patch and the "no-read"
4525 code in the 2.6 kernel caused the inode and/or EA data not to
4526 be read from disk, causing single-file corruption.
4530 Frequency : rare, only 2.6 filesystems using extents
4531 Description: Heavy concurrent write and delete load may cause data corruption.
4532 Details : It was possible under high-load situations to have an extent
4533 metadata block in the block device cache from a just-unlinked
4534 file overwrite a newly-allocated data block. We now unmap any
4535 metadata buffers that alias just-allocated data blocks.
4539 Frequency : filesystems with default stripe_count larger than 77
4540 Description: lconf+mke2fs fail when formatting filesystem with > 77 stripes
4541 Details : lconf specifies an inode size of 4096 bytes when the default
4542 stripe_count is larger than 77. This conflicts with the default
4543 inode density of 1 per 4096 bytes. Allocate smaller inodes in
4544 this case to avoid pinning too much memory for large EAs.
4546 ------------------------------------------------------------------------------
4548 2005-07-07 Cluster File Systems, Inc. <info@clusterfs.com>
4553 Frequency : rare (extremely heavy IO load with hundreds of clients)
4555 Description: Client is evicted, gets IO error writing to file
4556 Details : lock ordering changes for bug 5492 reintroduced bug 3267 and
4557 caused clients to be evicted for AST timeouts. The fixes in
4558 bug 5192 mean we no longer need to have such short AST timeouts
4559 so ldlm_timeout has been increased.
4562 Frequency : occasional during --force or --failover shutdown under load
4563 Bugzilla : 5949, 4834
4564 Description: Server oops/LBUG if stopped with --force or --failover under load
4565 Details : a collection of import/export refcount and cleanup ordering
4566 issues fixed for safer force cleanup
4569 Frequency : only filesystems larger than 120 OSTs
4570 Bugzilla : 5990, 6223
4571 Description: lfs getstripe would oops on a very large filesystem
4572 Details : lov_getconfig used kfree on vmalloc'd memory
4575 Frequency : only filesystems exporting via NFS to Solaris 10 clients
4576 Bugzilla : 6242, 6243
4577 Description: reading from files that had been truncated to a non-zero size
4578 but never opened returned no data
4579 Details : ll_file_read() reads zeros from no-object files to EOF
4584 Description: A bug in MDS/OSS recovery could cause the OSS to fail an assertion
4585 Details : There's little harm in aborting MDS/OSS recovery and letting it
4586 try again, so I removed the LASSERT and return an error instead.
4588 Severity : enhancement
4590 Description: New debugging infrastructure for tracking down data corruption
4591 Details : The I/O checksum code was replaced to: (a) control it at runtime,
4592 (b) cover more of the client-side code path, and (c) try to narrow
4593 down where problems occurred
4597 Bugzilla : 3819, 4364, 4397, 6313
4598 Description: Racing close and eviction MDS could cause assertion in mds_close
4599 Details : It was possible to get multiple mfd references during close and
4600 client eviction, leading to one thread referencing a freed mfd.
4602 Severity: : enhancement
4603 Bugzilla : 3262, 6359
4604 Description: Attempts to reconnect to servers are now more aggressive.
4605 Details : This builds on the enhanced upcall-less recovery that was added
4606 in 1.4.2. When trying to reconnect to servers, clients will
4607 now try each server in the failover group every 10 seconds. By
4608 default, clients would previously try one server every 25 seconds.
4613 Description: After recovery, certain operations trigger a failed
4614 assertion on a client.
4615 Details : Failing over an mds, using lconf -d --failover, while a
4616 client was doing a readdir() call would cause the client to
4617 LBUG after recovery completed and the readdir() was resent.
4619 Severity : enhancement
4621 Description: Default groups are now added by lconf
4622 Details : You can now run lconf --group <servicename> without having to
4623 manually add groups with lmc.
4626 Frequency : occasional
4628 Description: Nodes with an elan id of 0 trigger a failed assertion
4631 Frequency : always when accessing e.g. tty/console device nodes
4633 Description: tty and some other devices nodes cannot be used on lustre
4634 Details : file's private_data field is used by device data and lustre
4635 values in there got lost. New field was added to struct file to
4636 store fs-specific private data.
4639 Frequency : when exporting Lustre via NFS
4641 Description: NFSD failed occasionally when looking up a path component
4642 Details : NFSD is looking up ".." which was broken in ext3 directories
4643 that had grown large enough to become hashed.
4646 Frequency : Clusters with multiple interfaces not on the same subnet
4648 Description: Nodes will repeatedly try to reconnect to an interface which it
4649 cannot reach and report an error to the log.
4650 Details : Extra peer list entries will be created by lconf with some peers
4651 unreachable. lconf now validates the peer before adding it.
4654 Frequency : Only if a default stripe is set on the filesystem root.
4656 Description: Setting a default stripe on the filesystem root prevented the
4657 filesystem from being remounted.
4658 Details : The client was sending extra request flags in the root getattr
4659 request and did not allocate a reply buffer for the dir EA.
4662 Frequency : occasional, higher if lots of files are accessed by one client
4663 Bugzilla : 6159, 6097
4664 Description: Client trips assertion regarding lsm mismatch/magic
4665 Details : While revalidating inodes the VFS looks up inodes with ifind()
4666 and in rare cases can find an inode that is being freed.
4667 The ll_test_inode() code will free the lsm during ifind()
4668 when it finds an existing inode and then the VFS later attaches
4669 this free lsm to a new inode.
4673 Bugzilla : 6422, 7030
4674 Description: MDS deadlock between mkdir and client eviction
4675 Details : Creating a new file via mkdir or mknod (starting a transaction
4676 and getting the ns lock) can deadlock with client eviction
4677 (gets ns lock and trying to finish a synchronous transaction).
4680 Frequency : occasional
4681 Description: While starting a server, the fsfilt_ext3 module could not be
4683 Details : CFS's improved ext3 filesystem is named ldiskfs for 2.6
4684 kernels. Previously, lconf would still use the ext3 name
4685 when trying to load modules. Now, it will correctly use
4686 ext3 on 2.4 and ldiskfs on 2.6.
4688 Severity : enhancement
4689 Description: The default stripe count has been changed to 1
4690 Details : The interpretation of the default stripe count (0, to lfs
4691 or lmc) has been changed to mean striping across a single
4692 OST, rather than all available. For general usage we have
4693 found a stripe count of 1 or 2 works best.
4695 Severity : enhancement
4696 Description: Add support for compiling against Cray portals.
4697 Details : Conditional compiling for some areas that are different
4701 Frequency : occasional
4702 Bugzilla : 6409, 6834
4703 Description: Creating files with an explicit stripe count may lead to
4704 a failed assertion on the MDS
4705 Details : If some OSTs are full or unavailable, creating files may
4706 trigger a failed assertion on the MDS. Now, Lustre will
4707 try to use other servers or return an error to the
4711 Frequency : occasional
4713 Description: Multiple concurrent overlapping read+write on multiple SMP nodes
4714 caused lock timeout during readahead (since 1.4.2).
4715 Details : Processes doing readahead might match a lock that hasn't been
4716 granted yet if there are overlapping and conflicting lock
4717 requests. The readahead process waits on ungranted lock
4718 (original lock is CBPENDING), while OST waits for that process
4719 to cancel CBPENDING read lock and eventually evicts client.
4721 Severity : enhancement
4723 Description: Initial enabling of flock support for clients
4724 Details : Implements fcntl advisory locking and file status functions.
4725 This feature is provided as an optional mount flag (default
4726 off), and is NOT CURRENTLY SUPPORTED. Not all types of record
4727 locking are implemented yet, and those that are are not guaranteed
4728 to be completely correct in production environments.
4729 mount -t lustre -o [flock|noflock] ...
4732 Frequency : occasional
4734 Description: OSTs running 2.4 kernels but with extents enabled might trip an
4735 assertion in the ext3 JBD (journaling) layer.
4736 Details : The b_committed_data struct is protected by the big kernel lock
4737 in 2.4 kernels, serializing journal_commit_transaction() and
4738 ext3_get_block_handle->ext3_new_block->find_next_usable_block()
4739 access to this struct. In 2.6 kernels there is finer grained
4740 locking to improve SMP performance of the JBD layer.
4744 Description: Changes the "SCSI I/O Stats" kernel patch to default to "enabled"
4749 Description: merge and cleanup kernel patches.
4752 -----------------------------------------------------------------------------
4754 2005-05-05 Cluster File Systems, Inc. <info@clusterfs.com>
4756 NOTE: Lustre 1.4.2 uses an incompatible network protocol than previous
4757 versions of Lustre. Please update all servers and clients to
4758 version 1.4.2 or later at the same time. You must also run
4759 "lconf --write-conf {config}.xml" on the MDS while it is stopped
4760 to update the configuration logs.
4762 - fix for HPUX NFS client breakage when NFS exporting Lustre (5781)
4763 - mdc_enqueue does not need max_mds_easize request buffer on send (5707)
4764 - swab llog records of type '0' so we get proper header size/idx (5861)
4765 - send llog cancel req to DLM cancel portal instead of cb portal (5515)
4766 - fix rename of one directory over another leaking an inode (5953)
4767 - avoid SetPageDirty on 2.6 (5981)
4768 - don't re-add just-being-destroyed locks to the waiting list (5653)
4769 - when creating new directories, inherit the parent's custom
4770 striping settings if present parent (3048)
4771 - flush buffers from cache before direct IO in 2.6 obdfilter (4982)
4772 - don't hold i_size_sem in ll_nopage() and ll_ap_refresh_count (6077)
4773 - don't hold client locks on temporary worklist from l_lru (5666)
4774 - handle IO errors in 2.6 obdfilter bio completion routine (6046)
4775 - automatically evict dead clients (5921)
4776 - Update file size properly in create+truncate+fstat case (6196)
4777 - Do not unhash mountpoint dentries, do not allow removal of
4779 - Avoid lock ordering deadlock issue with write/truncate (6203,5654)
4780 - reserve enough journal credits in fsfilt_start_log for setattr (4554)
4781 - ldlm_enqueue freed-export error path would always LBUG (6149,6184)
4782 - don't reference lr_lvb_data until after we hold lr_lvb_sem (6170)
4783 - don't overwrite last_rcvd if there is a *_client_add() error (6086)
4784 - Correctly handle reads of files with no objects (6243)
4785 - lctl recover will also mark a device active if deactivate used (5933)
4787 - by default create 1 inode per 4kB space on MDS, per 16kB on OSTs
4788 - allow --write-conf on an MDS with different nettype than client (5619)
4789 - don't write config llogs to MDS for mounts not from that MDS (5617)
4790 - lconf should create multiple TCP connections from a client (5201)
4791 - init scripts are now turned off by default; run chkconfig --on
4792 lustre and chkconfig --on lustrefs to use them
4793 - upcalls are no longer needed for clients to recover to failover
4795 - add --abort-recovery option to lconf to abort recovery on device
4797 - add support for an arbitrary number of OSTs (3026)
4798 - Quota support protocol changes.
4799 - forward compatibility changes to wire structs (6007)
4800 - rmmod NALs that might be loaded because of /etc/modules.conf (6133)
4801 - support for mountfsoptions and clientoptions to the Lustre LDAP (5873)
4802 - improved "lustre status" script
4803 - initialize blocksize for non-regular files (6062)
4804 - added --disable-server and --disable-client configure options (5782)
4805 - introduce a lookup cache for lconf to avoid repeated DB scans (6204)
4806 - Vanilla 2.4.29 support
4807 - increase maximum number of obd devices to 520 (6242)
4808 - remove the tcp-zero-copy patch from the suse-2.4 series (5902)
4809 - Quadrics Elan drivers are now included for the RHEL 3 2.4.21 and
4810 SLES 9 2.6.5 kernels
4811 - limit stripes per file to 160 (the maximum EA size) (6093)
4813 2005-03-22 Cluster File Systems, Inc. <info@clusterfs.com>
4816 - don't LASSERT in ll_release on NULL lld with NFS export (4655, 5760)
4817 - hold NS lock when calling handle_ast_error->del_waiting_lock (5746)
4818 - fix setattr mtime regression from lovcleanup merge (4829, 5669)
4819 - workaround for 2.6 crash in ll_unhash_aliases (5687, 5210)
4820 - small ext3 extents cleanups and fixes (5733)
4821 - improved mballoc code, several small races and bugs fixed (5733, 5638)
4822 - kernel version 43 - fix remove_suid bugs in both 2.4 and 2.6 (5695)
4823 - avoid needless client->OST connect, fix handle mismatch (5317)
4824 - fix DLM error path that led to out-of-sync client, long delays (5779)
4825 - support common vfs-enforced mount options (nodev,nosuid,noexec) (5637)
4826 - fix several locking issues related to i_size (5492,5624,5654,5672)
4827 - don't move pending lock onto export if it is already evicted (5683)
4828 - fix kernel oops when creating .foo in unlinked directory (5548)
4829 - fix deadlock in obdfilter statistics vs. object create (5811)
4830 - use time_{before,after} to avoid timer jiffies wrap (5882)
4831 - shutdown --force/--failover stability (3607,3651,4797,5203,4834)
4832 - Do not leak request if server was not able to process it (5154)
4833 - If mds_open unable to find parent dir, make that negative lookup(5154)
4834 - don't create new directories with extent-mapping (5909, 5936)
4836 - fix lustre/lustrefs init scripts for SuSE (patch from Scali, 5702)
4837 - don't hold the pinger_sem in ptlrpc_pinger_sending_on_import
4838 - change obd_increase_kms to obd_adjust_kms (up or down) (5654)
4839 - lconf, lmc search both /usr/lib and /usr/lib64 for Python libs (5800)
4840 - support for RHEL4 kernel on i686 (5773)
4841 - provide error messages when incompatible logs are encountered (5898)
4843 2005-02-18 Cluster File Systems, Inc. <info@clusterfs.com>
4844 * version 1.4.0.10 (1.4.1 release candidate 1)
4846 - don't keep a lock reference when lock is not granted (4238)
4847 - unsafe list practices (rarely) led to infinite eviction loop (4908)
4848 - add per-fs limit of Lustre pages in page cache, avoid OOM (4699)
4849 - drop import inflight refcount on signal_completed_replay error (5255)
4850 - unlock page after async write error during send (3677)
4851 - handle missing objects in filter_preprw_read properly (5265)
4852 - no transno return for symlink open, don't save no-trasno open (3440)
4853 - don't try to complete elan receive that already failed (4012)
4854 - free RPC server reply state on error (5406)
4855 - clean up thread from ptlrpc_start_thread() on error (5160)
4856 - readahead could read extra page into cache that wasn't ejected (5388)
4857 - prevent races in class_attach/setup/cleanup/detach (5260)
4858 - don't dereference de->d_inode after l_dput of de (5458)
4859 - use "int" for stripe value returned from lock_to_stripe (5544)
4860 - mballoc allocation and error-checking fixes in 2.6 (5504)
4861 - block device patches to fix I/O request sizes in 2.6 (5482)
4862 - look up hostnames for IB nals (5602)
4863 - 2.6 changed lock ordering of 2 semaphores, caused deadlock (5654)
4864 - don't start multiple acceptors for the same port (5277)
4865 - fix incorrect LASSERT in mds_getattr_name (5635)
4866 - export a proc file for general "ping" checking (5628)
4867 - fix "lfs check" to not block when the MDS is down (5628)
4869 - service request history (4965)
4870 - put {ll,lov,osc}_async_page structs in a single slab (4699)
4871 - create an "evict_client" /proc entry on OSTs, like the MDS has
4872 - fix mount usage message, return errors per mount(8) (5168)
4873 - change grep [] to grep "[]" in tests so they work in more UMLs
4874 - fix ppc64/x86_64 spec to use %{_libdir} instead of /usr/lib (5389)
4875 - remove ancient LOV_MAGIC_V0 EA support (5047)
4876 - add "disk I/Os in flight" and "I/O req time" stats in obdfilter
4877 - align r/w RPCs to PTLRPC_MAX_BRW_SIZE boundary for performance (3451)
4878 - allow readahead allocations to fail when low on memory (5383)
4879 - mmap locking landed again, after considerable improvement (2828)
4880 - add get_hostaddr() to lustreDB.py for LDAP support (5459)
4882 2004-11-23 Cluster File Systems, Inc. <info@clusterfs.com>
4885 - send OST transaction number in read/write reply to free req (4966)
4886 - don't ASSERT in ptl_send_rpc() if we run out of memory (5119)
4887 - lock /proc/sys/portals/routes internal state, avoiding oops (4827)
4888 - the watchdog thread now runs as interruptible (5246)
4889 - flock/lockf fixes (but it's still disabled, pending 5135)
4890 - don't use EXT3 constants in llite code (5094)
4891 - memory shortage at startup could cause assertion (5176)
4893 - reorganization of lov code
4894 - single portals codebase
4896 - add extents/mballoc support (5025)
4897 - direct I/O reads in the obdfilter (4048)
4898 - kernel patches from LNXI for 2.6 (bluesmoke, perfctr, mtd, kexec)
4900 tbd Cluster File Systems, Inc. <info@clusterfs.com>
4903 - send OST transaction number in read/write reply to free req (4966)
4904 - don't ASSERT in ptl_send_rpc() if we run out of memory (5119)
4905 - lock /proc/sys/portals/routes internal state, avoiding oops (4827)
4906 - the watchdog thread now runs as interruptible (5246)
4907 - handle missing objects in filter_preprw_read properly (5265)
4908 - unsafe list practices (rarely) led to infinite eviction loop (4908)
4909 - drop import inflight refcount on signal_completed_replay error (5255)
4910 - unlock page after async write error during send (3677)
4911 - return original error code on reconstructed replies (3761)
4912 - no transno return for symlink open, don't save no-trasno open (3440)
4914 - add pid to ldlm debugging output (4922)
4915 - bump the watchdog timeouts -- we can't handle 30sec yet
4916 - extra debugging for orphan dentry/inode bug (5259)
4918 2004-11-16 Cluster File Systems, Inc. <info@clusterfs.com>
4921 - fix TCP_NODELAY bug, which caused extreme perf regression (5134)
4922 - allocate qswnal tx descriptors singly to avoid fragmentation (4504)
4923 - don't LBUG on obdo_alloc() failure, use OBD_SLAB_ALLOC() (4800)
4924 - fix NULL dereference in /proc/sys/portals/routes (4827)
4925 - allow failed mdc_close() operations to be interrupted (4561)
4926 - stop precreate on OST before MDS would time out on it (4778)
4927 - don't send partial-page writes before EOF from client (4410)
4928 - discard client grant for sub-page writes on large-page clients (4520)
4929 - don't free dentries not owned by NFS code, check generation (4806)
4930 - fix lsm leak if mds_create_objects() fails (4801)
4931 - limit debug_daemon file size, always print CERROR messages (4789)
4932 - use transno after validating reply (3892)
4933 - process timed out requests if import state changes (3754)
4934 - update mtime on OST during writes, return in glimpse (4829)
4935 - add mkfsoptions to LDAP (4679)
4936 - use ->max_readahead method instead of zapping global ra (5039)
4937 - don't interrupt __l_wait_event() during strace
4939 - add software watchdogs to catch hung threads quickly (4941)
4940 - make lustrefs init script start after nfs is mounted
4941 - fix CWARN/ERROR duplication (4930)
4942 - return async write errors to application if possible (2248)
4943 - add /proc/sys/portal/memused (bytes allocated by PORTALS_ALLOC)
4944 - print NAL number in %x format (4645)
4945 - update barely-supported suse-2.4.21-171 series (4842)
4946 - support for sles 9 %post scripts
4947 - support for building 2.6 kernel-source packages
4948 - support for sles km_* packages
4950 2004-10-07 Cluster File Systems, Inc. <info@clusterfs.com>
4953 - ignore -ENOENT errors in osc_destroy (3639)
4954 - notify osc create thread that OSC is being cleaned up (4600)
4955 - add nettype argument for llmount in #5d in conf-sanity.sh (3936)
4956 - reconstruct ost_handle() like mds_handle() (4657)
4957 - create a new thread to do import eviction to avoid deadlock (3969)
4958 - let lconf resolve symlinked-to devices (4629)
4959 - don't unlink "objects" from directory with default EA (4554)
4960 - hold socknal file ref over connect in case target is down (4394)
4961 - allow more than 32000 subdirectories in a single directory (3244)
4962 - fix blocks count for O_DIRECT writes (3751)
4963 - OST returns ENOSPC from object create when no space left (4539)
4964 - don't send truncate RPC if file size isn't changing (4410)
4965 - limit OSC precreate to 1/2 of value OST considers bogus (4778)
4966 - bind to privileged port in socknal and tcpnal (3689)
4968 - rate limit CERROR/CWARN console message to avoid overload (4519)
4969 - GETFILEINFO dir ioctl returns LOV EA + MDS stat in 1 call (3327)
4970 - basic mmap support (3918)
4971 - kernel patch series update from b1_4 (4711)
4973 2004-09-16 Cluster File Systems, Inc. <info@clusterfs.com>
4976 - avoid crash during MDS cleanup with OST shut down (2775)
4977 - fix loi_list_lock/oig_lock inversion on interrupted IO (4136)
4978 - don't use bad inodes on the MDS (3744)
4979 - dynamic object preallocation to improve recovery speed (4236)
4980 - don't hold spinlock over lock dumping or change debug flags (4401)
4981 - don't zero obd_dev when it is force cleaned (3651)
4982 - print grants to console if they go negative (4431)
4983 - "lctl deactivate" will stop automatic recovery attempts (3406)
4984 - look for existing locks in ldlm_handle_enqueue() (3764)
4985 - don't resolve lock handle twice in recovery avoiding race (4401)
4986 - revalidate should check working dir is a directory (4134)
4988 - don't always mark "slow" obdfilter messages as errors (4418)
4990 2004-08-24 Cluster File Systems, Inc. <info@clusterfs.com>
4993 - don't close LustreDB during write_conf until it is done (3860)
4994 - fix typo in lconf for_each_profile (3821)
4995 - allow dumping logs from multiple threads at one time (3820)
4996 - don't allow multiple threads in OSC recovery (3812)
4997 - fix debug_size parameters (3864)
4998 - fix mds_postrecov to initialize import for llog ctxt (3121)
4999 - replace config semaphore with spinlock (3306)
5000 - be sure to send a reply for a CANCEL rpc with bad export (3863)
5001 - don't allow enqueue to complete on a destroyed export (3822)
5002 - down write_lock before checking llog header bitmap (3825)
5003 - recover from lock replay timeout (3764)
5004 - up llog sem before sending rpc (3652)
5005 - reduce ns lock hold times when setting kms (3267)
5006 - change a dlm LBUG to LASSERTF, to maybe learn something (4228)
5007 - fix NULL deref and obd_dev leak on setup error (3312)
5008 - replace some LBUG about llog ops with error handling (3841)
5009 - don't match INVALID dentries from d_lookup and spin (3784)
5010 - hold dcache_lock while marking dentries INVALID and hashing (4255)
5011 - fix invalid assertion in ptlrpc_set_wait (3880)
5013 - add libwrap support for the TCP acceptor (3996)
5014 - add /proc/sys/portals/routes for non-root route listing (3994)
5015 - allow setting MDS UUID in .xml (2580)
5016 - print the stack of a process that LBUGs (4228)
5018 2004-07-14 Cluster File Systems, Inc. <info@clusterfs.com>
5021 - don't cleanup request in ll_file_open() on failed MDS open (3430)
5022 - make sure to unset replay flag from failed open requests (3440)
5023 - if default stripe count is 0, use OST count for inode size (3636)
5024 - update parent mtime/ctime on client for create/unlink (2611)
5025 - drop dentry ref in ext3_add_link from open_connect_dentry (3266)
5026 - free recovery state on server during a forced cleanup (3571)
5027 - unregister_reply for resent reqs (3063)
5028 - loop back devices mounting and status check on 2.6 (3563)
5029 - fix resource-creation race that can provoke i_size == 0 (3513)
5030 - don't try to use bad inodes returned from MDS/OST fs lookup (3688)
5031 - more debugging for page-accounting assertion (3746)
5032 - return -ENOENT instead of asserting if ost getattr+unlink race (3558)
5033 - avoid deadlock after precreation failure (3758)
5034 - fix race and lock order deadlock in orphan handling (3450, 3750)
5035 - add validity checks when grabbing inodes from l_ast_data (3599)
5037 - add /proc/.../recovery_status to obdfilter (3428)
5038 - lightweight CDEBUG infrastructure, debug daemon (3668)
5039 - change default OSC RPC parameters to be better on small clusters
5040 - turn off OST read cache for files smaller than 32MB
5041 - install man pages and include them in rpms (3100)
5042 - add new init script for (un)mounting lustre filesystems (2593)
5043 - run chkconfig in %post for init scripts (3701)
5044 - drop scimac NAL (unmaintained)
5046 2004-06-17 Cluster File Systems, Inc. <info@clusterfs.com>
5049 - clean kiobufs before and after use (3485)
5050 - strip trailing '/'s before comparing paths with /proc/mounts (3486)
5051 - remove assertions to work around "in-flight rpcs" recovery bug (3063)
5052 - change init script to fail more clearly if not run as root (1528)
5053 - allow clients to reconnect during replay (1742)
5054 - fix ns_lock/i_sem lock ordering deadlock for kms update (3477)
5055 - don't do DNS lookups on NIDs too small for IP addresses (3442)
5056 - re-awaken ptlrpcd if new requests arrive during check_set (3554)
5057 - fix cond_resched (3554)
5058 - only evict unfinished clients after recovery (3515)
5059 - allow bulk resend, prevent data loss (3570)
5060 - dynamic ptlrpc request buffer allocation (2102)
5061 - don't allow unlinking open directory if it isn't empty (2904)
5062 - set MDS/OST threads to umask 0 to not clobber client modes (3359)
5063 - remove extraneous obd dereference causing LASSERT failure (3334)
5064 - don't use get_cycles() when creating temp. files on the mds (3156)
5065 - hold i_sem when setting i_size in ll_extent_lock() (3564)
5066 - handle EEXIST for set-stripe, set proper directory name (3336)
5068 - servers can dump a log evicting a client - lustre.dump_on_timeout=1
5069 - fix ksocknal_fmb_callback() error messages (2918)
5071 2004-05-27 Cluster File Systems, Inc. <info@clusterfs.com>
5074 - don't copy lvb into (possibly NULL) reply on error (2983)
5075 - don't deref dentry after dput, don't free lvb on error (2922)
5076 - use the kms to determine writeback rpc length (2947)
5077 - increment oti_logcookies when osc is inactive (2948)
5078 - update client's i_blocks count via lvb messages (2543)
5079 - handle intent open/close of special files properly (1557)
5080 - mount MDS with errors=remount-ro, like obdfilter (2009)
5081 - initialize lock handle to avoid ASSERT on error cleanup (3057)
5082 - don't use cancelling-locks' kms values (2947)
5083 - use highest lock extent for kms, not last one (2925)
5084 - don't dereference ERR_PTR() dentry in error handling path (3107)
5085 - fix thread race in portals_debug_dumplog() (3122)
5086 - create lprocfs device entries at setup instead of at attach (1519)
5087 - common AST error handler, don't evict client on completion race (3145)
5088 - zero nameidata in detach_mnt in 2.6 (3118)
5089 - verify d_inode after revalidate_special is valid in 2.6 (3116)
5090 - use lustre_put_super() to handle zconf unmounts in 2.6 (3064)
5091 - initialize RPC timeout timer earlier for 2.6 (3219)
5092 - don't dereference NULL reply buffer if mdc_close was never sent (2410)
5093 - print nal/nid for unknown nid (3258)
5094 - additional checks for oscc recovery before doing precreate (3284)
5095 - fix ll_extent_lock() error return code for 64-bit systems (3043)
5096 - don't crash in mdc_close for bad permissions on open (3285)
5097 - zero i_rdev for non-device files (3147)
5098 - clear page->private before handing to FS, better assertion (3119)
5099 - tune the read pipeline (3236)
5100 - fix incorrect decref of invalidated dentry (2350)
5101 - provide read-ahead stats and refine rpc in flight stats (3328)
5102 - don't hold journal transaction open across create RPC (3313)
5103 - update atime on MDS at close time (3265)
5104 - close LDAP connection when recovering to avoid server load (3315)
5105 - update iopen-2.6 patch with fixes from 2399,2517,2904 (3301)
5106 - don't leak open file on MDS after open resend (3325)
5107 - serialize filter_precreate and filter_destroy_precreated (3329)
5108 - loop device shouldn't call sync_dev() for nul device (3092)
5109 - clear page cache after eviction (2766)
5110 - resynchronize MDS->OST in background (2824)
5111 - refuse to mount the same filesystem twice on same mountpoint (3394)
5112 - allow llmount to create routes for mounting behind routers (3320)
5113 - push lock cancellation to blocking thread for glimpse ASTs (3409)
5114 - don't call osc_set_data_with_check() for TEST_LOCK matches (3159)
5115 - fix rare problem with rename on htree directories (3417)
5117 - allow default OST striping configuration per directory (1414)
5118 - fix compilation for qswnal for 2.6 kernels (3125)
5119 - increase maximum number of MDS request buffers for large systems
5120 - change liblustreapi to be useful for external progs like lfsck (3098)
5121 - increase local configuration timeout for slow disks (3353)
5122 - allow configuring ldlm AST timeout - lustre.ldlm_timeout=<seconds>
5124 2004-03-22 Cluster File Systems, Inc. <info@clusterfs.com>
5127 - fixes for glimpse AST timeouts / incorrectly 0-sized files (2818)
5128 - don't overwrite extent policy data in reply if lock was blocked (2901)
5129 - drop filter export grants atomically with removal from device (2663)
5130 - del obd_self_export from work_list in class_disconnect_exports (2908)
5131 - don't LBUG if MDS recovery times out during orphan cleanup (2530)
5132 - swab reply message in mdc_close, other PPC fixes (2464)
5133 - fix destroying of named logs (2325)
5134 - overwrite old logs when running lconf --write_conf (2264)
5135 - bump LLOG_CHUNKSIZE to 8k to allow for larger clusters (2306)
5136 - fix race in target_handle_connect (2898)
5137 - mds_reint_create() should take same inode create lock (2926)
5138 - correct journal credits calculated for CANCEL_UNLINK_LOG (2931)
5139 - don't close files for self_export to avoid uninitialized obd (2936)
5140 - allow MDS with the same name as client node (2939)
5141 - hold dentry reference for closed log files for unlink (2325)
5142 - reserve space for all logs during transactions (2059)
5143 - don't evict page beyond end of stripe extent (2925)
5144 - don't oops on a deleted current working directory (2399)
5145 - handle hard links to targets without a parent properly (2517)
5146 - don't dereference NULL lock when racing during eviction (2867)
5147 - don't grow lock extents when lots of conflicting locks (2919)
5149 2004-03-04 Cluster File Systems, Inc. <info@clusterfs.com>
5152 - account for cache space usage on clients to avoid data loss (974)
5153 - lfsck support in lustre kernel code (2349)
5154 - reduce journal credits needed for BRW writes (2370)
5155 - orphan handling to avoid losing space on client/server crashes
5156 - ptlrpcd can be blocked, stopping ALL progress (2477)
5157 - use lock value blocks to assist in proper KMS, faster stat (1021)
5158 - takes i_sem instead of DLM locks internally on obdfilter (2720)
5159 - recovery for initial connections (2355)
5160 - fixes for mds_cleanup_orphans (1934)
5161 - abort_recovery crashes MDS in b_eq (mds_unlink_orphan) (2584)
5162 - block all file creations until orphan recovery completes (1901)
5163 - client remove rq_connection from request struct (2423)
5164 - conf-sanity test_5, proper cleanup in umount log not availale (2640)
5165 - recovery timer race (2670)
5166 - mdc_close recovey bug (2532)
5167 - ptlrpc cleanup bug (2710)
5168 - mds timeout on local locks (2588)
5169 - namespace lock held during RPCs (2431)
5170 - handle interrupted sync write properly (2503)
5171 - don't try to handle a message that hasn't been replied to (2699)
5172 - client assert failure during cleanup after abort recovery (2701)
5173 - leak mdc device after failed mount (2712)
5174 - ptlrpc_check_set allows timedout requests to complete (2714)
5175 - wait for inflight reqs when ptlrpcd finishes (2710)
5176 - make sure unregistered services are removed from the srv_list
5177 - reset bulk XID's when resending them (caught by 1138 test)
5178 - unregister_bulk after timeout
5179 - fix lconf error (2694)
5180 - handle write after unfinished setstripe, stripe-only getstripe (2388)
5181 - readahead locks pages, leaves pending causing memory pressure (2673)
5182 - increase OST request buffers to 4096 on large machines (2729)
5183 - fix up permission of existing directories in simple_mkdir (2661)
5184 - init deleted item, add assertions ptlrpc_abort_inflight() (2725)
5185 - don't assign transno to errored transactions (2742)
5186 - don't delete objects on OST if given a bogus objid from MDS (2751)
5187 - handle large client PAGE_SIZE readdir on small PAGE_SIZE MDS (2777)
5188 - if rq_no_resend, then timeout request after recovery (2432)
5189 - fix MDS llog_logid record size, 64-bit array alignment (2733)
5190 - don't call usermode_helper from ptlrpcd, DEFAULT upcall (2773)
5191 - put magic in mount.lustre data, check for bad/NULL mount data (2529)
5192 - MDS recovery shouldn't delete objects that it has given out (2730)
5193 - if enqueue arrives after completion, don't clobber LVB (2819)
5194 - don't unlock pages twice when trigger_group_io returns error (2814)
5195 - don't deref NULL rq_repmsg if ldlm_handle_enqueue failed (2822)
5196 - don't write pages to disk if there was an error (1450)
5197 - don't ping imports that have recovery disabled (2676)
5198 - take buffered bytes into account when balancing socknal conn (2817)
5199 - hold a DLM lock over readdir always, use truncate_inode_pages (2706)
5200 - reconnect unlink llog connection after MDS reconnects to OST (2816)
5201 - remove little-endian swabbing of llog records (1987)
5202 - set/limit i_blksize to LL_MAX_BLKSIZE on client (2884)
5203 - retry reposting request buffers if they fail (1191)
5204 - grow extent at grant time to avoid granting a revoked lock (2809)
5205 - lock revoke doesn't evict page if covered by a second lock (2765)
5206 - disable VM readahead to avoid reading outside lock extents (2805)
5208 - return LL_SUPER_MAGIC from statfs for the filesystem type (1972)
5209 - updated kernel patches for hp-2.4.20 kernel (2681)
5211 2004-02-07 Cluster File Systems, Inc. <info@clusterfs.com>
5214 - fix truncated write corruption (2366)
5215 - fix for failed assertion in iopen_connect_dentry (1792,2517)
5217 - don't flag the ptlrpcd thread with PF_MEMALLOC (2636)
5218 - ensure len(uuid) < 37 in lmc (1171)
5219 - fix ia64 OOPS in llog_test (2255)
5220 - zero end of page at obdfilter for partial page writes (2648)
5221 - don't leave stale dentries around after renames (bug 2428)
5222 - fix timeouts when evicting a client with a single lock held (2642)
5223 - set deadline for the initial HELLO message to drain (2634)
5224 - print out dotted-quad IP addresses in the socknal (2302)
5226 - additional debugging for MDS client eviction problem (2443)
5227 - fix mkfsoptions support for osts (2603, 2604)
5229 2004-01-27 Cluster File Systems, Inc. <info@clusterfs.com>
5232 - add series for the vanilla 2.6.0 kernel
5233 - add series for the vanilla 2.4.24 kernel
5234 - add series for a cray x86/64 UL kernel drop
5235 - fix xattr patches for the vanilla 2.4.19 series
5237 - generate true UUIDs in lmc (1171)
5238 - have portals stack dumping break in UML (2466)
5239 - avoid bad dchild deref; avoid inum lock w/o creation (2362)
5240 - allocate with _NOFS in ldlm to avoid deadlock (1933)
5241 - wake callback waiting threads on client eviction (2460)
5242 - Add --ptldebug and --subsystem to lmc (1719)
5243 - update assertion to allow safe interrupt allocation
5244 - set rq_no_resend for cancel requests (2432)
5245 - recalculate ptlrpcd timeout after resend (2494)
5246 - call vfs_rmdir when removing pending directories (2368)
5247 - fix renaming a file to itself (2429)
5248 - lmc creates a default one-stripe lov (2454)
5249 - expand procfs space to handle large clusters (2326)
5250 - increase UML stack to avoid overflow
5251 - update lconf's list of debug and subsystem masks
5252 - fix lfs find --obd (2510)
5253 - /proc tunable for disabling filter read caching (2591)
5254 - stop rpm packages from altering slapd.conf (2301)
5255 - disable nagle in the socknal under 0conf (2578)
5256 - choose mds inode size based on stripe count (2572)
5257 - fix kernel-source rpm problems (2516)
5259 - add --disable-doc to avoid pdf generation (2421)
5260 - update documentation, tests, type-os, comments
5261 - avoid format warnings on ia64
5262 - remove the TOE NAL
5263 - tiny code cleanups by removing unused fields
5265 2004-01-07 Cluster File Systems, Inc. <info@clusterfs.com>
5268 - fix obvious semaphore misuse in as-yet-unused setattr path (2348)
5269 - remove the most blatant lies from BUILDING file (2371)
5270 - change default debug level to reasonable production setting
5271 - reduce client side cache size to reduce cache flush time
5272 - reduce max RPCs in flight to avoid unnecessary file fragmentation
5273 - make TCP zerocopy and pinger support enabled by default (2476)
5274 - sync writes completed after process exits caused crashes (2319)
5275 - maintain correct mount count on the MDS (2356)
5276 - backout 1557, because 2316 wasn't really fixed
5277 - better file I/O statistics gathering in /proc
5278 - don't take unnecessary, deadlock-inducing bug in readpage (2383)
5279 - another kernel patch to fix zero-copy TCP function export
5280 - don't take duplicate lock when processing re-sent getattr (2420)
5281 - lctl uses obd_self_export instead of creating new conn (2353)
5282 - MDS/OST recovery case which requires object creation asserted (2425)
5283 - move lfs from /usr/sbin to /usr/bin in packages
5284 - fix race between mds_client_add and mds_client_free (2417)
5285 - use kmalloc instead of slabs in portals (2430)
5286 - don't create duplicate records when a failover MDS is present (2442)
5287 - remove unnecessary mount age check (2332)
5288 - don't remove directory inodes from locks prematurely (2451)
5289 - don't break if MDS service name is the same as hostname (2103)
5290 - fix races in client write RPC generation when cache full (2482)
5292 2003-12-13 Cluster File Systems, Inc. <info@clusterfs.com>
5295 - remove now-unused request->rq_obd (278)
5296 - if an allocation fails, print out how much memory we've used (1933)
5297 - use PORTAL_SLAB_ALLOC for structures, to get GFP_MEMALLOC (1933)
5298 - add the "configurable stack size" patch to most series files (1256)
5299 - ability to write large log records, for 100+ OST configs (2306)
5300 - fix NULL deref when filter_prep fails (2314)
5301 - fix operator precedence error in filter_sync
5302 - dynamic allocation of socknal TX descriptors (2315)
5303 - fix a missed case in the GFP_MEMALLOC patch, can cause deadlock (2310)
5304 - fix gcc 2.96 compilation problem in xattr kernel patch (2294)
5305 - ensure that CWARN messages in Portals always get to the syslog
5306 - __init/__exit are not for prototype decls (ldlm_init/exit)
5307 - x86-64 compile warning fixes
5308 - fix gateway LMC keyword conflict (2318)
5309 - fix MDS lock inversions in getattr/reint paths (1844)
5310 - fix a rare lock re-ordering bug, which caused deadlock (2322)
5311 - fix i_sem/journal inversion in fsfilt_ext3_write_record (2306)
5312 - DLM race condition prevented some lock evictions (2328)
5313 - ENOMEM detection and retry on socknal sends (2230)
5314 - use GFP_NOFS throughout Lustre, to combat ENOMEM (2230)
5315 - move osc_rpcd into ptlrpc, for use in MDC and others (2329)
5316 - protect MDS inode fsdata with stronger locking; fixes assertion (2313)
5317 - better error messages when a client is rejected during recovery (1505)
5318 - avoid cancelling locks which were never granted, after failure (2330)
5319 - fix i_sem/journal inversion in mds_client_add (2333)
5320 - fix truncate/getattr lock cycle deadlock (2334)
5321 - use rpcd to send close; allows resend after timeout, avoid leak (1897)
5322 - fix two rare exit paths which could leak an l_lock() ref (2321)
5323 - fencepost error in MDS/OST orphan recovery (2226)
5324 - make log record alignment 8 bytes (1988)
5325 - lstripe now fails when requested offset > ost_count (2237)
5326 - ensure that all kernel series have a complete list.h (1607)
5327 - fix crashes in special-file operations (2316)
5328 - lctl create/brw OID mismatch, caused by obsolete filter loop (2339)
5330 - allow configurable automake binary, for testing new versions
5331 - small update to the lfs documentation
5333 2003-12-03 Cluster File Systems, Inc. <info@clusterfs.com>
5335 * fix negative export reference count in fsfilt_sync (2312)
5337 2003-12-01 Cluster File Systems, Inc. <info@clusterfs.com>
5338 * release candidate 0.9.1
5340 - orphans are moved into the PENDING directory for possible recovery
5341 - replayed opens now open by fid for orphan/rename safety (1042)
5342 - last close of an orphan inode generates a transno (683)
5343 - chdir() and mount() now pin the directory entry (1020)
5344 - avoid CERROR in normal ll_setattr_raw() error case (1500)
5345 - discard very old requests without processing them (1502)
5346 - remove some common, well-understood CERRORs (1505)
5347 - require O_DIRECT I/O to be page-sized to workaround IA64 crash (1609)
5348 - clear "grant" flags in OST replies until OST grant code lands (1644)
5349 - fix read performance by not clobbering i_blksize on client (1598)
5350 - fix __ldlm_handle2lock oops by not dereferencing lock after PUT (1625)
5351 - make LRU size a /proc tunable, clears locks when reduced (707)
5352 - fix some lprocfs rot that prevented ptlbd from loading (1732)
5353 - server locks take references on exports now (1558)
5354 - build fixes for 2.4.20-rh trees (1663)
5355 - return an error from lov_create if all OSCs are inactive (1751)
5356 - fix import levels when a reconnect happens without a timeout (1597)
5357 - exit early from mds_open if we get a lookup error (1749)
5358 - partial page read at EOF wouldn't wait for disk before sending (1642)
5359 - avoid NULL deref in obdfilter when reading page past EOF (1592)
5360 - avoid LASSERT in ll_intent_lock if server failed very early (1090)
5361 - fix LBUG in ll_it_open_error with rc = -2 (1861)
5362 - write/truncate lock inversion (1639)
5363 - Don't auto-load obdclass, portals modules during cleanup (1495)
5364 - fix timestamps from jumping to "now" (1763)
5365 - extra journal assertions (1648)
5366 - add an extra multiunlink test (1771)
5367 - fix read_record/write_record API (1776)
5368 - fix leak of offset_extent, possible incorrect i_size later (1772)
5369 - fix lasserts in mis-matched transnos during open-unlink testing (1541)
5370 - Debugging for the kqswnal_get_idle_tx problems (1820)
5371 - Allow recovery to be attempted multiple times (1536)
5372 - Write out MDS last_rcvd file after it is first created (1600)
5373 - Fix tx_descriptor leak in failed transmit situations (1827)
5374 - ext3 journaling fixes for assertion failure after IO error (1871)
5375 - class_export_put() on freed export after completion AST error (1896)
5376 - Fix revalidate looping in VFS (1322)
5377 - Don't access a freed export during MDS_REINT timeout (1521)
5378 - Add open-unlink recovery support on the MDS (1673,1764)
5379 - Return an error if no MDS data was read from last_rcvd (1946)
5380 - Fix for lookup "." or ".." crash on error (1932,1931,1935)
5381 - Don't setup a disk device that doesn't match exported UUID (317)
5382 - Reduce bulk RPC timeout to avoid cascading client/OST failures (1845)
5383 - avoid committing NULL handle in force close
5384 - local.sh is now a one-stripe LOV configuration
5385 - POSIX utime.4 -EPERM on FIFO not owned by user (56)
5386 - fix ext3 htree duplicate directory entry corruption (1516)
5387 - POSIX creat.13, fstat.1, open.18, stat.3 new file atime/mtime (2020)
5388 - update to new LOV EA format (2097)
5389 - interoperability for different PAGE_SIZE/wordsize (686,1821,1343,2042)
5391 2003-06-15 Phil Schwan <phil@clusterfs.com>
5394 - imports and exports cleanup too early, need refcounts (349, 879, 1045)
5395 - per-import/export recovery handling (958, 931, 959)
5396 - multiple last-rcvd slots, for serving multiple FSes (949)
5397 - connections are again shared between multiple imp/exports (963, 964)
5398 - "umount -f" would hang if any requests needed to be sent (393, 978)
5399 - avoid pinning large req buffer by copying for queued messages (989)
5400 - add "uuid" to "lctl device" command to help upcalls (991)
5401 - "open" RPCs with transnos would confuse recovery counters (1037)
5402 - do proper endian conversion of all wire messages (288, 340, 891)
5403 - remove OST bulk get LBUGs, fix ost_brw_write cleanup (1126)
5404 - call waiting locks callback from LDLM recovery thread (1127, 1151)
5405 - fix ptlrpc_connection leak in target_handle_connect (1174)
5406 - fix import refcounting bug in OST and MDS cleanup (1134)
5407 - if an invalid-at-open-time OSC returned before close(), LBUG (1150)
5408 - fix very unlikely obd_types race condition (501)
5409 - remove osc_open hack for echo_client (1187)
5410 - we leaked exports/dlmimps for forcibly disconnected clients (1143)
5411 - a failure in read_inode2 leads to deadlock (1139)
5412 - cancel ack-locks as soon as transaction is committed (1072)
5413 - fix major leaks and crashes in the bulk I/O path (937, 1057)
5414 - make sure to commitrw after any preprw to avoid deadlock (1162)
5415 - failing to execute a file in a lustre FS would lock inode (1203)
5416 - small DEBUG_REQ fix to avoid dereferencing a NULL (1227)
5417 - don't ASSERT while cleaning up an incompletely-setup obd (1248)
5418 - obd_uuid2tgt would walk off the end of the list (1255)
5419 - on IA64 the osc would give portals incorrect bulk size (1258)
5420 - fix debug daemon ioctl interface; allows daemon on ia64 (1274)
5421 - fix lock inversion caused by new llite matching code (1282)
5422 - limit the number of dirty pages on a client to 10MB (1286)
5423 - timed out locks were not being corrected cancelled (1289)
5424 - fix O_DIRECT above 4GB on IA-32 (1292)
5425 * major user-visible changes
5426 - fail out/fail over policy now controlled by the upcall (993)
5428 - add OBD_PING to check server availability and failure (954)
5429 - lustre messages are now sent in sending host order (288, 340, 891)
5430 - add eadatalen to MDS getattr reply (340)
5431 - OST read replies may contain second buffer, with per-page status (593)
5433 2003-03-11 Phil Schwan <phil@clusterfs.com>
5436 - LDLM_DEBUG macro fix, for gcc 3.2 (850)
5437 - failed open()s could cause deadlock; fixed (867, 869)
5438 - stop cancelling OST locks when files are closed (481)
5439 - overlapping XID spaces caused network corruption (851, 853)
5440 - fix unsafe fsfilt counter arithmetic; change to atomic_t
5441 - setattr_raw added, to do single-RPC, server-side setattrs
5442 - lmc/lconf syntax change for OST UUIDs
5443 - fix crashy race condition between ptlrpc_free_req and osc_close
5444 - don't use request in mdc_enqueue if we hit a timeout (889)
5445 - don't set the inode i_size for regular files from the MDS (896)
5446 - handle out of order completion AST (842)
5447 - don't LBUG if a lock request times out after receiving AST (913)
5448 - avoid d_rehash race in ll_find_alias by rehashing inside dcache_lock
5449 - if a bad lock AST arrives, send an error instead of dropping entirely
5450 - return 0 from revalidate2 if ll_intent_lock returns -EINTR (912)
5451 - fix leak in bulk IO when only partially completed (899, 900, 926)
5452 - fix O_DIRECT for ia64 (55)
5453 - (almost) eliminate Lustre-kernel-thread effects on load average (722)
5454 - C-z after timeout could hang a process forever; fixed (977)
5456 - client-side I/O cache (678, 924, 929, 941, 970)
5458 - READPAGE and SETATTRs which don't take server-side locks get
5461 2003-02-11 Phil Schwan <phil@clusterfs.com>
5464 - Fix ldlm_lock_match on the MDS to avoid matching remote locks (592)
5465 - Fix fsfilt_extN_readpage() to read a full page of directory
5466 entries, or fake the remainder if PAGE_SIZE != blocksize (500)
5467 - Avoid extra mdc_getattr() in ll_intent_lock when possible (534, 604)
5468 - Fix imbalanced LOV object allocation and out-of-bound access (469)
5469 - Most intent operations were removed, in favour of a new RPC mode
5470 that does a single RPC to the server and bypasses most of the VFS
5471 - All LDLM resource ID arrays were removed in favour of ldlm_res_id
5472 - Aggressively cancel local locks on DLM servers
5473 - mds_reint_unlink sends EA to the client if it's the last nlink.
5474 client uses that EA to unlink OST objects.
5475 - mds_reint_{rename,unlink,link} were rewritten to take ordered locks
5476 - recursive symlinks were fixed (439)
5477 - fixed NULL deref in DEBUG_REQ
5478 - filter_update_lastobjid no longer calls sync, which annoyed extN
5479 - fixed multi-client small-writes to a single file problem (445)
5480 - fixed mtime updates during file writes (607)
5481 - fixed vector writes on obdfilter causing problems when ENOSPC (670)
5482 - fixed bug in obd_brw_read/write() (under guise of testing 367)
5483 - fixed Linux OST size reporting problem (444, 656)
5484 - OST now updates object mtime with writes or setattr (607, 619)
5485 - client verifies file size before zeroing page past EOF (445)
5486 - OST now writes last allocated objid to disk with allocation (108)
5487 - LOV on echo now works (409)
5489 - mds_reint_unlink sends a new buffer, with the EA included. this
5490 buffer is only valid if body->valid & OBD_MD_FLEASIZE, which is only
5491 set if a regular file was being unlinked, and it was the last link
5492 - use PtlGet from the target for bulk writes (315)
5493 - OST now updates object mtime with writes or setattr (607, 619)
5494 - LDLM now has a grant-time callback to revalidate locked items, if
5496 - Many MDS operations were reorganized to combat race conditions
5498 - Merge b_intel branch (updated lprocfs code) - now at /proc/fs/lustre
5499 - configure check to avoid gcc version 2.96 20000731-2.96-98) (606)
5501 2003-01-06 Andreas Dilger <adilger@clusterfs.com>
5504 - Fully reactivate OST imports after reconnection (512, others)
5505 - Make sure client sees our -ENOTCONN from mds_handle (513 - partial)
5506 - More graceful error handling for truncating on dead OST (515)
5507 - Don't error out unless we're actually accessing dead stripes (474)
5508 - Fix garbage sizes when stripes are missing (410)
5509 - LRU counters were broken, causing constant lock purge (433, 432)
5510 - garbage on read from stripes with failed OSTs (441)
5511 - mark OSCs as active before reconnecting during recovery (438)
5512 - lov_enqueue and lov_cancel need to handle inactive OSTs (403)
5513 - lfind did not preserve OST order in output (443)
5514 - symlinks cause hung clients, incorrect data (439)
5515 - stop dereferencing request after dropping refcount (457)
5516 - don't LASSERT(spin_is_locked) on non-SMP (455)
5517 - fixes for many rename() bugs
5518 - fstat didn't correctly synchronize attributes (399)
5519 - server must handle lock cancellation during blocking AST prep (487)
5520 - bulk descriptors were free()d too soon (511)
5521 - fix paths in lconf, which would load incorrect modules (451, 507)
5522 - fix confusing lconf 'host not found' error message (386)
5523 - fix lock order deadlock on OST (O/R i_sem before journal ops, 478)
5524 - fix race condition in mdc_blocking_ast() for inode access (526)
5525 - fix lov_unpackmd() unpacking wrong number of stripes (537)
5526 - fix lov_set_osc_active() marking wrong OSC inactive (440)
5527 - fix bad lstripe lov_unpackmd() assertion (fix layering too) (527)
5528 - fix multiple writes of stripe MD to MDS (358, maybe 519)
5529 - fix lstripe in several ways (kernel side) (527)
5530 - fix request leak in ldlm_cli_enqueue (262)
5531 - incorrect OSC was marked inactive after OST failure
5532 - call mds_fs_cleanup before unmounting filesystem (524)
5533 - fix races between taking ns_lock and ldlm_lock_change_resource
5534 - fix races updating LOV export open file list
5535 - fix lov_enqueue error path, avoid decref-ing bad lock handle
5536 - fix recovery NULL deref in ldlm_cli_cancel_unused
5537 - fix some DLM races by using new hash table for lock handles (419)
5538 - permit the client to specify desired inodes, at replay
5539 - duplicate requests when we queue them for replay reintegration
5540 - fix last_rcvd offset calculation
5541 - sync after each recovered transaction, so we always make progress
5542 - never, not always, ERESTART requests without transnos
5543 - store the lov_desc in the MDS, so we don't depend on getlovinfo to
5545 - skip replay if the MDS says that the client is already connected
5546 - don't check for a recovery-enabled export to match lctl's UUID
5547 - don't INC_USE_COUNT for phantom exports
5548 - don't crash when cleaning up phantom exports (567)
5549 - don't double-finish or set replay data for errored mdc_open requests
5550 - abort requests when they time out, so we don't get old replies
5551 - send/receive replies for AST messages again
5552 - if the client says that it doesn't have the lock, cancel it on the
5554 - if we timeout during I/O, don't try to cancel an in-use lock; instead
5555 mark it as destroyed, it will all work out when decref is called
5556 - fix module use counts (22, 581)
5558 - ASTs now expect a reply (server cancels lock on error reply)
5560 2002-12-02 Andreas Dilger <adilger@clusterfs.com>
5563 - fix many simultaneous client startup (392)
5564 - fix dentry->d_it clobbering
5565 - credentials weren't being shipped for readdir/getattr operations
5566 - remove invalid assertions triggered during some concurrent MD
5568 - proper Lustre versions added (336, 389)
5569 - fix memory leak for create error case (398)
5570 - fix LOV locking bug that would get cli/srv out of sync
5571 - fix echo client over LOV (409)
5572 - fix dbench 2, extN refcount problem (170, 258, 356, 418)
5573 - fix double-O_EXCL intent crash (424)
5574 - avoid sending multiple lock CANCELs (352)
5576 - MDS can do multi-client recovery (modulo bugs in new code)
5578 - many updates, edits, cleanups
5580 2002-11-18 Phil Schwan <phil@clusterfs.com>
5583 - fix null d_it dereference (346)
5584 - fix full OST/dbench hang (333)
5585 - fix permission problem with file removal (286)
5586 - fix removal of OSCs from LOV when they fail
5587 - fix NULL deref during bulk timeout (214)
5588 - fix problems related to multiple filesystems on one MDS (241)
5589 - fixed serious subtle metadata locking bugs
5590 - free locks on clients when inodes are removed due to memory
5592 - fix inode pointer in lock data (285)
5593 - partial support for multiple MDS on a single host (241)
5594 - data locks weren't cancelled at clear_inode time (290, 311)
5595 - intent locks could lead to unbounded lock growth (205)
5596 - added a maximum lock count, an LRU list, and a flusher
5597 - fix multiple rename (365)
5598 - properly abstracted the echo client
5599 - OSC locked 1 byte too many; fixed
5600 - rewrote brw callback code:
5601 - fixed recovery bugs related to LOVs (306)
5602 - fixed too-many-pages-in-one-write crash (191)
5603 - fixed (again) crash in sync_io_timeout (214)
5604 - probably fixed callback-related race (385)
5606 - Add capability to MDS protocol
5607 - LDLM cancellations and callbacks on different portals
5609 2002-10-28 Andreas Dilger <adilger@clusterfs.com>
5612 - limit client IOV size to PTL_MD_MAX_IOV (611336, 191)
5613 - defer open object destruction to close time (601981, 138)
5614 - open/close OST file handle in obdo (OBD_MD_FLHANDLE) (601981, 138)
5615 - move LDLM_ENQUEUE/CONVERT back to MDS portal (625069)
5616 - abstract ll_lookup2, fix ll_revalidate2 to use abstraction (256)
5617 - don't call obd_setattr in ll_file_release for destroyed objects
5618 * protocol change to lustre_msg: move |version| and add |flags|
5619 * protocol change to osc_punch: "start" in "o_size", "end" in "o_blocks"
5620 * lock replay: for LDLM_FL_REPLAY trust client to do right thing
5621 * added replay of create, unlink, link and rename operations during
5622 MDS failover; recovery should be much more robust now
5623 * remove failed OSCs from LOVs (only lov_create uses this so far)
5624 * the lustre-HOWTO was brought (more) up to date (582544)
5626 2002-10-23 Phil Schwan <phil@clusterfs.com>
5629 - in-use dentries weren't being reused properly (617851)
5630 - prevent multiple LDLM setup (599178)
5631 - fix LOV size calculations for truncate (617853)
5632 - fix client handling of MDS intent errors (POSIX)
5633 - fix permission bug in lovstripe.c test (624321)
5634 - fix MDS thread deadlock - move LDLM handler to DLM portal (625069)
5635 - truncate past end of file could corrupt data
5636 - proper cleanup after timeouts, crashes, etc (592524, 550815)
5637 - a race in recovery could return ETIMEDOUT to apps (623947)
5638 - building outside the source directory was fixed
5639 * the lustre-HOWTO was brought (more) up to date (582544)
5640 * major progress was made on recovery functionality
5642 2002-10-10 Phil Schwan <phil@clusterfs.com>
5645 - recovery deadlock fix
5646 - rm -rf causes LBUG fix (617817)
5647 - file open by multiple tasks fix (618962)
5648 - directory permissions bugs (602707 and 620007)
5649 - journal_stop fixed with locking (611313)
5650 - O_APPEND failures resolved (618273, perhaps 614459)
5651 - lconf PATH fix (619770)
5652 - IA64 build fix (621450)
5653 - RPC buffer sizes scale with amount of memory
5655 2002-10-01 Phil Schwan <phil@clusterfs.com>
5658 - locks would be cancelled without throwing away data pages,
5659 resulting in inconsistent data (605627)
5660 - inode attributes were not always being refreshed (605627, 612449)
5661 - lconf now continues to cleanup after lctl reports an error
5662 - MDS now enforces user permissions (602707)
5663 - lprocfs cleanup fixed, but not yet enabled (614157)
5664 - fixed infinite server hang, should a client not respond to an AST
5665 - avoid going into recovery if user calls readlink() with a buffer
5666 that's too small (613941)
5667 - AST RPCs no longer require replies (614867) -- this may be changed
5668 - don't crash server if client sends an IOV that's too big (611336)
5669 - fixed lock conversion deadlock (611892)
5670 - fixed the following of symlinks (614622)
5671 * recovery: the server can remove locks from a client that dies, other
5672 clients can make progress
5673 * more extN patch fixes
5674 * compile-time configurable ptlrpc buffer allocations
5676 - collaborative read cache document
5677 - Lustre Lite Performance CDR document-in-progress
5679 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
5682 - fix typo in patch-2.4.18
5684 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
5687 - clear ptlrpc request each time in handle_incoming_request()
5688 - unlink of files now destroys the object on the OST
5690 2002-09-19 Peter Braam <braam@clusterfs.com>
5692 * add hard link support
5693 * change obdfile creation method
5694 * kernel patch changed
5696 2002-09-19 Peter Braam <braam@clusterfs.com>
5699 - stack overflow bug in extN fixed
5701 2002-09-18 Andreas Dilger <adilger@clusterfs.com>
5703 * documentation updates
5704 - add man pages for config tools
5705 - update tests/README to describe testing with new config tools
5706 - finish metadata API descriptions
5707 * bug fixes and cleanups
5708 - statfs workaround for 16TB limit
5709 - LOV stripe allocation improved, can stripe on subset of OSTs
5710 - LOV file size/IO offset was wrong for files > 4GB in size
5711 - object EA data was being dropped, caused files to be unreadable
5712 - memory overflow with non-LOV OST caused memory corruption
5713 - fixed regression tests to work with new config tools, obdfilter
5714 - fixed bug when directory size became larger than 1 block
5715 - fixed bug (for single client case) when PWD was deleted
5716 - invalidate local directory pages when doing intent-based ops
5717 - avoid LDLM oops when lock callback contained bad data
5719 2002-09-09 Andreas Dilger <adilger@clusterfs.com>
5721 * documentation updates
5722 * bug fixes and cleanups
5723 - configuration tools
5726 - 64-bit compile warnings
5727 - 64-bit internal statfs data
5729 * test_brw on persistent OST devices
5731 * lprocfs (disabled)
5733 2002-09-04 Andreas Dilger <adilger@clusterfs.com>
5735 * documentation updates
5736 * bug fixes and cleanups
5737 * configuration tools
5739 2002-08-30 Peter J. Braam <braam@clusterfs.com>
5742 * many small fixes to 0_5_4
5743 * io/network handling
5744 * thinkos in MDS operations
5746 2002-08-24 Peter J. Braam <braam@clusterfs.com>
5749 * crucial basic fixes to 0.5.3
5750 * IOR, Iozone work over Elan
5753 2002-08-07 Phil Schwan <phil@clusterfs.com>
5754 * version 0_5_3, our first alpha
5755 * we use the new Portals iovs
5756 * documentation updates
5757 * bug fixes and cleanups
5758 * small changes in the DLM wire protocol
5760 2002-07-25 Peter J. Braam <braam@clusterfs.com>
5761 * version 0_5_1 with some initial stability,
5762 * locking on MD and file I/O.
5763 * documentation updates
5764 * several bug fixes since 0.5.0
5765 * small changes in wire protocol
5767 2002-07-18 Phil Schwan <phil@clusterfs.com>
5769 * delivered as Lustre Light Alpha
5770 * fixed a crash after handling invalid MDS requests
5771 * fixed directory pages for architectures with non-4k pages sizes
5773 2002-07-11 Andreas Dilger <adilger@clusterfs.com>
5774 * release version v0_4_4
5775 * Moves TCP acceptor to be on port 2432 (unused Coda port) instead
5777 * Fixes a number of interruption problems with OST operations.
5778 * Update documentation for portals header changes
5779 * Move all wire protocol structs/defines to lustre_idl.h
5780 * Fixes symlink length bug.
5781 * Add tcpdump to repository.
5783 2002-07-05 Andreas Dilger <adilger@clusterfs.com>
5784 * release version v0_4_3
5785 * Fixes statfs for inodes on extN.
5786 * Fixes bug in runtests which would delete /etc/hosts.
5787 * Use 64-bit object IDs wherever possible (not into VFS though)
5788 Remove ost_get_info, which is unused by lustre, and out of date.
5790 2002-07-03 Peter Braam <braam@clusterfs.com>
5791 * release version v0_4_2 Fixes a lookup error (type not passed)
5792 * move forward to head of Portals
5793 * move forward to latest Lustre kernel
5795 2002-06-25 Peter Braam <braam@clusterfs.com>
5796 * release version v0_4_1. Hopefully stable on single node use.
git://git.whamcloud.com / fs / lustre-release.git / blob