summaryrefslogtreecommitdiffstats
path: root/xlators/cluster
Commit message (Collapse)AuthorAgeFilesLines
* fix leaks pointed out by cppcheck static analyzerCsaba Henk2009-11-301-1/+3
| | | | | | | | | | | | | | | | | | | See http://cppcheck.sf.net [./build/libglusterfs/src/y.tab.c:2104]: (error) Memory leak: cmd [./libglusterfs/src/md5.c:306]: (error) Resource leak: f [./xlators/debug/io-stats/src/io-stats.c:1396]: (error) Possible null pointer dereference: this - otherwise it is redundant to check if this is null at line 1402 [./xlators/cluster/stripe/src/stripe.c:3597]: (error) Memory leak: dup_str [./xlators/debug/trace/src/trace.c:2426]: (error) Possible null pointer dereference: this - otherwise it is redundant to check if this is null at line 2429 These appear in rarely hit error cases or test/demo code so nothing serious, in fact. Signed-off-by: Csaba Henk <csaba@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 420 (fix leaks pointed out by cppcheck static analyzer) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=420
* cluster/afr: Preserve generation number along with inode in lookup and ↵Vikas Gorur2009-11-303-3/+26
| | | | | | | | | | | | | creation fops. This fixes fuse_create_cbk conflict warnings and random errors while running dbench (typically open handle failure with ENOENT). Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 315 (generation number support) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=315
* cluster/afr: Refactored lookup_cbk and introduce precedence of errors.Vikas Gorur2009-11-302-117/+303
| | | | | | | | | | | | | Error handling in afr_lookup_cbk was faulty because it did not give priority to errors such as ESTALE over ENOENT, and ENOENT over other errors. This patch fixes that, and also breaks up afr_lookup_cbk into multiple logical functions. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 205 ([ glusterfs 2.0.6rc4 ] - Hard disk failure not handled correctly) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=205
* afr: fix crash in afr_sh_data_closeAnand Avati2009-11-291-0/+3
| | | | | | | | | | | | when active_sink count is 0, the code proceeded into a dangerous loop resulting in a crash while issuing the call or in the callback afr_sh_data_setattr_cbk or afr_sh_data_flush_cbk Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* afr: handle fdctx->pre_op_done handlingAnand Avati2009-11-291-0/+1
| | | | | | | | | | | | reset pre_op_done[i] to 0 after issuing a postop in flush. this was missed during the introduction of pre_op_done[] array and was resulting in a lot of spurious self heals when spurious flushes were received Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* afr: fix logic to break out of diff/checksum loopAnand Avati2009-11-281-9/+5
| | | | | | | | | | | | When checksum fop returns error, mark for terminating the loop at the end of the iteration (when all checksum calls of that iteration return) and not immediately Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* moved unify from 'cluster/unify' to 'legacy/cluster/unify'Amar Tumballi2009-11-281-1/+1
| | | | | | | | | | | | maintaining unify is not feasible, and its design of namespace is not scalable too. hence moving it to 'legacy/' directory.. Thanks to unify, we learnt something about rename(). 'cluster/unify' has served GlusterFS well for more than 2years, and its the time to say goodbye :-( Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com>
* fixes in nufaAmar Tumballi2009-11-263-96/+92
| | | | | | | | | | | | | | | Now nufa uses dht_conf->private variable to store its contents, and doesn't pollute distribute's structures. It used to hang earlier as layout_lock was not INIT()'d. whitespace-cleanup done on the code too. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 409 (implement a switch scheduler) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=409
* switch translator addedAmar Tumballi2009-11-263-1/+1059
| | | | | | | | | | | switch translator is a wrapper around distribute to work for a pattern based scheduling. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 409 (implement a switch scheduler) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=409
* don't allow stripe with just one subvolumeAmar Tumballi2009-11-261-1/+7
| | | | | | | | | | | | stripe code assumes in many places that it will have more than one subvolume, but the strict check was not there in init, hence, if somebody configures stripe with just one subvolume, it will cause many fops with frame loss. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 340 (Kernel Compilation fail in stripe) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=340
* cluster/afr: Include "common-utils.h" instead of alloca.hVikas Gorur2009-11-261-1/+1
| | | | | | | | | | | alloca.h should be included on a platform-specific basis. Lets common-utils.h handle that. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 349 (FreeBSD compilation error (alloca.h).) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=349
* cluster/afr: Do self-heal on unopened fds.Vikas Gorur2009-11-256-26/+232
| | | | | | | | | | | | | | This patch completes the previous patch for self-heal of open fds in replicate. If an fd was never opened on a subvolume, we remember that and do the open after we've done self-heal on that fd. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Refactored the self-heal interface.Vikas Gorur2009-11-249-167/+141
| | | | | | | | | | Cleaned up the self-heal interface to callers. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Do self-heal on reopened fds.Vikas Gorur2009-11-2410-229/+699
| | | | | | | | | | | | | | | | | | | | | | | This patch brings in partial support for self-heal of open fds. The precondition is that the fd should have been opened successfully during the initial open() (or create()), and we assume that protocol/client has successfully reopened the fd when the subvolume comes back up. It works by doing an "up/down flush" (a dummy flush transaction to do post-op wherever necessary) and then triggering data self-heal on the file in the post-post-op hook of the dummy flush transaction. This ensures that any writes that come in during self-heal will wait until self-heal completes. The up/down flush is also done when a subvolume goes down, so that post-op is done on all subvolumes where pre-op was done. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Provide a post-post_op hook in the transaction.Vikas Gorur2009-11-242-6/+23
| | | | | | | | Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Refactored the data self-heal algorithm.Vikas Gorur2009-11-242-143/+248
| | | | | | | | | | | | | | | | | | Refactored the operation of the data self-heal algorithm as: * open all fd's (if fd not supplied by caller) * lock 0-0 (if lock not supplied by caller) * fxattrop, fstat (instead of lookup) ... self heal ... * unlock (if lock not supplied by caller) * close (if fd not supplied by caller). Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Hold blocking locks for data self-heal.Vikas Gorur2009-11-244-35/+62
| | | | | | | | | | | Data self-heal now holds blocking locks, and instead of locking on all subvolumes, it only locks on {data-lock-server-count} subvolumes. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Set read-child = source regardless of foreground/background ↵Vikas Gorur2009-11-241-1/+2
| | | | | | | | | | self-heal Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 170 (Auto-heal fails on files that are open()-ed/mmap()-ed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=170
* cluster/afr: Fix inode context bitmasks.Vikas Gorur2009-11-244-17/+11
| | | | | | | | | | | Set opendir_done and split_brain flags correctly in the inode context. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 249 (Self heal of a file that does not exist on the first subvolume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=249
* cluster/afr: Fix double-free in opendir self-heal callback.Vikas Gorur2009-11-241-1/+4
| | | | | | | | | | | local->cont.opendir.checksum was being free'd both in the self-heal completion function and self-heal unwind. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 249 (Self heal of a file that does not exist on the first subvolume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=249
* cluster/afr: Unlock only those paths which have been locked during rename.Vikas Gorur2009-11-241-77/+142
| | | | | | | | | | | For ENTRY_RENAME_TRANSACTIONs, keep track separately whether the lower_path and the higher_path have been locked, and unlock only those which have been. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 112 (parallel deletion of files mounted by different clients on the same back-end hangs and/or does not completely delete) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=112
* cluster/afr: Fix handling of revalidate lookups.Vikas Gorur2009-11-242-24/+38
| | | | | | | | | | | | | | | | This patch does two things related to revalidate: 1) If a revalidate fails on any subvolume, the entire lookup call is failed. 2) Self-heal is not triggered on a revalidate if revalidate has failed on any subvolume. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 389 (auto-heal fails randomly and causes "Stale NFS file handle" errors) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=389
* cluster/distribute: copy the stat structure while copying dentries in ↵Raghavendra G2009-11-191-0/+2
| | | | | | | | | | | | | readdirp_cbk. - only the transformed inode number was being copied, leaving other fields of stat 0. Signed-off-by: Raghavendra G <raghavendra@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 371 (rm -rf fails on stat-prefetch.) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=371
* cluster/stripe: Wrong variable is checked after calloc.vinayak hegde2009-11-191-1/+1
| | | | | | | | Signed-off-by: Vinayak Hegde <vinayak@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 368 (setattr on a non-regular file makes glusterfs crash) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=368
* cluster/afr: Handle op_ret properly in opendir_cbkVikas Gorur2009-11-191-2/+3
| | | | | | | | | | | Change the success condition to op_ret >= 0 instead of op_ret == 0. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 249 (Self heal of a file that does not exist on the first subvolume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=249
* fixes in stripeAmar Tumballi2009-11-161-2/+3
| | | | | | | | | | | | | * fix in stripe readv, so that a read request which falls in two (or more) subvolumes, depending on stripe-size works fine. * fix in stripe_setattr_cbk, where 'postop' stat structure was returning wrong 'st_size'. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 368 (setattr on a non-regular file makes glusterfs crash) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=368
* cluster/afr: Fix self-heal loop driver termination logic.Vikas Gorur2009-11-131-7/+19
| | | | | | | | Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Ensure directory contents are in sync during opendir.Vikas Gorur2009-11-135-10/+281
| | | | | | | | | | | | | | | | | | | | | | The problem: If some files on the first subvolume disappeared without leaving a trace in the entry changelog (this can happen, for example, when an fsck has deleted files or when a hard drive is replaced), those files would never be self-healed even though they would be present on the second subvolume. This is because readdir is sent only to the first subvolume, and since the files don't appear in the directory listing, no lookup would ever be sent on them. This patch fixes this problem by doing a readdir on all the subvolumes during the first opendir on a directory inode. If a discrepancy in the contents is detected, entry self-heal in a special "force merge" mode is triggered on that directory. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 249 (Self heal of a file that does not exist on the first subvolume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=249
* cluster/afr: Refactored inode context bitmasks.Vikas Gorur2009-11-131-6/+10
| | | | | | | | | | | Defined symbolic constants for the bit masks and made 'split-brain' a single bit field in the ctx. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 249 (Self heal of a file that does not exist on the first subvolume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=249
* cluster/dht: Do readdir in dht_readdir instead of readdirpVikas Gorur2009-11-131-1/+1
| | | | | | | | | | | | The readdir call will not be used by applications (fuse will only ever send readdirp), but it will be used by afr to do directory self-heal at the time of opendir. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 249 (Self heal of a file that does not exist on the first subvolume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=249
* cluster/stripe: Initialize frame->local before checking for directory or ↵vinayak hegde2009-11-131-18/+22
| | | | | | | | | | regular files. Signed-off-by: Vinayak Hegde <vinayak@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 368 (setattr on a non-regular file makes glusterfs crash) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=368
* cluster/stripe: Set local->inode value before referring it.vinayak hegde2009-11-121-1/+7
| | | | | | | | Signed-off-by: Vinayak Hegde <vinayak@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 362 (GlusterFS crash while doing mkdir in stripe setup) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=362
* cluster/stripe: local was not being set in stripe_mknod_ifreg_setxattr_cbk()Vijay Bellur2009-11-121-0/+2
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 367 (GlusterFS crash while doing mknod in stripe setup) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=367
* Add unhashed-sticky-bit option to volume_options structureSachidananda2009-11-121-0/+3
| | | | | | | | Signed-off-by: Sachidananda <sac@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 365 (unhashed-sticky-bit is not recognized) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=365
* cluster/afr: In fop structure afr_readdirp is not correctly registered.vinayak hegde2009-11-051-1/+1
| | | | | | | | Signed-off-by: Vinayak Hegde <vinayak@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 360 (All fop fails when stat-prefetch is loaded on afr.) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=360
* cluster/afr: Bail out of the self-heal driver loops as soon as possibleVikas Gorur2009-11-031-4/+4
| | | | | | | | | | | | Don't wait for the next recursive call to sh_{full,diff}_loop_driver to decide that we've reached the end of file, as the frame could have been destroyed by that time (if subvolumes are posix). Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Don't try to self-heal if there are locks heldVikas Gorur2009-10-302-1/+18
| | | | | | | | | | | If the inodelk_count or entrylk_count is positive on a file/directory, don't try to do self-heal on it. Signed-off-by: Vikas Gorur <vikas@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 326 ([2.0.8rc9] Spurious self-heal) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=326
* cluster/afr: Move deleted files to /.trash in entry self-heal.Vikas Gorur2009-10-294-33/+226
| | | | | | | | | | | | | | | | | If entry self-heal determines that a file/directory should be deleted from a subvolume, move that entry to a directory called "/.trash" on that subvolume. This is for two reasons: 1) It limits the damage that can be done by a "wrong" entry self-heal. 2) It solves the problem of a to-be-deleted directory not being empty. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 227 (replicate selfheal does not remove directory with contents in it) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=227
* cluster/afr: Set local->fd in fsync.Vikas Gorur2009-10-281-0/+1
| | | | | | | Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 145 (NFSv3 related additions to 2.1 task list) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=145
* cluster/afr: Remove struct's for [f]chown, [f]chmod, utimens from afr_local_t.Vikas Gorur2009-10-271-37/+0
| | | | | | | Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 146 (Add setattr FOP) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=146
* cluster/afr: NFS-friendly logic changesVikas Gorur2009-10-276-73/+250
| | | | | | | Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 145 (NFSv3 related additions to 2.1 task list) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=145
* cluster/afr: Pick the data self-heal algorithm dynamically.Vikas Gorur2009-10-262-13/+66
| | | | | | | Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Check the target of symlink's in entry self-heal.Vikas Gorur2009-10-263-12/+208
| | | | | | | | | | During entry self-heal, make sure not only that a symlink exists on all subvolumes, but also that their targets match. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 193 (symlink contents not self-healed by replicate) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=193
* cluster/afr: Set read-subvolume to the source while self-heal is in progress.Vikas Gorur2009-10-261-0/+1
| | | | | | | | | | | | Just before the lookup is unwound during background data self-heal, the read subvolume is set to the self-heal source subvol so that read operations on the file work correctly, and don't have to wait for the self-heal to complete. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Do self-heal in the background.Vikas Gorur2009-10-265-11/+163
| | | | | | | | | | | | | | This patch introduces a new option "background-self-heal-count", with a default value of 16. This means that upto {background-self-heal-count} number of files/directories will be healed in the background at any given time. If such number of self-heals are already in progress, further self-heals take place in the foreground. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Pipeline the "diff" data self-heal read-write loop.Vikas Gorur2009-10-232-110/+267
| | | | | | | | | | Start upto "data-self-heal-window-size" instances of the read-write loop of the "diff" data self-heal algorithm simultaneously. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Pipeline the "full" data self-heal read-write loop.Vikas Gorur2009-10-234-55/+243
| | | | | | | | | | | | | Start upto "data-self-heal-window-size" instances of the read-write loop of the "full" data self-heal algorithm simultaneously. Add a new option "data-self-heal-window-size" with range [1-1024], and a default value of 16. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 320 (Improve self-heal performance) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=320
* cluster/afr: Prevent spurious entry self-heal.Vikas Gorur2009-10-231-2/+13
| | | | | | | | | | | | | | If the initial lookup shows that 'pending' is positive, then self-heal will hold a lock and do a lookup again. This lookup might show that 'pending' is zero everywhere. However, entry self-heal used to consider this as a case of 'no sources' and try to merge the directories. This patch checks for that case and does not do the merge. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 326 ([2.0.8rc9] Spurious self-heal) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=326
* cluster/afr: entry self-heal: Store parent_loc on the heap for ↵Vikas Gorur2009-10-221-7/+12
| | | | | | | | | impunge_parent_setattr. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 137 (Parent directory mtime not reset after a create in self-heal) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=137
* cluster/afr: entry self-heal: Set parent directory's attributes using a new ↵Vikas Gorur2009-10-221-24/+20
| | | | | | | | | | | | | | | | | | | | | frame. There was a race condition in assuming that afr_sh_entry_impunge_parent_setattr_cbk will always return before impunge_xattrop_cbk and impunge_setattr_cbk. This patch fixes two additional problems: 1) Building the parent_loc from impunge_local->loc after STACK_WIND to impunge_xattrop_cbk has happened. In a simple afr-posix configuration the stack will have been destroyed by the time building of parent_loc is attempted. 2) parent_loc built in impunge_newfile_cbk was not being loc_wipe'd. Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 137 (Parent directory mtime not reset after a create in self-heal) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=137