summaryrefslogtreecommitdiffstats
path: root/xlators
Commit message (Collapse)AuthorAgeFilesLines
* rpc-clnt: show disconnect messages in log file by defaultAnand Avati2010-12-062-9/+24
| | | | | | | | Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 875 (Implement a new protocol to provide proper backward/forward compatibility) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=875
* check whether the file is a symlink while doing utimesRaghavendra Bhat2010-12-061-1/+19
| | | | | | | | Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 881 (GlusterFS daemon hangs on replication of symlink (3.0.4)) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=881
* mount.glusterfs: use option --direct-io-mode instead of --disable-direct-io-modeRaghavendra G2010-12-031-1/+1
| | | | | | | | Signed-off-by: Raghavendra G <raghavendra@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2173 (enabling/disabling direct-io mode is not possible when glusterfs is mounted using mount command) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2173
* nfs: Start nfs process even if portmap registration failsShehjar Tikoo2010-12-031-4/+7
| | | | | | | | | | | | It helps to have the process be listed in ps ax|grep output so that user can get the path to the log from the listed command line and check what the problem is. Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2093 (volumes cannot start when one node in a replicated setup is down) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2093
* fuse-bridge: construct fa->rdev properlyAnand Avati2010-12-031-1/+2
| | | | | | | | Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2145 ([3.1.1qa9] mknod does not work properly) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2145
* stat-prefetch: changes to incorporate gfidAnand Avati2010-11-251-1/+7
| | | | | | | | Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2136 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2136
* mgmt/glusterd: Enable stat-prefetch for fuseVijay Bellur2010-11-251-9/+14
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2136 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2136
* nfs3: Prevent second lookup on a fresh entry lookupShehjar Tikoo2010-11-253-3/+11
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1756 (NFS must revalidate inode on first ESTALE on lookup) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1756
* nfsrpc: Add new macro nfs_rpcsvc_request_procnumShehjar Tikoo2010-11-251-0/+1
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1756 (NFS must revalidate inode on first ESTALE on lookup) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1756
* nfs: Export subvolumes on per-subvolume CHILD-UPShehjar Tikoo2010-11-253-13/+29
| | | | | | | | | | | | | | ..so that nfs clients can mount an UP subvolume even if other subvolumes havent come up yet. This was a problem because nfs was waiting for all children to receive CHILD-UP before exporting any subvolume. Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2093 (volumes cannot start when one node in a replicated setup is down) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2093
* nfs3: Fresh inode lookup on failed revalidationShehjar Tikoo2010-11-252-1/+47
| | | | | | | | | | | | | | | | Brings in changes that were earlier introduced in commit: f5afcc47f9f00472d6c2b3f48127e02332cd457a but reverted because the patch was buggy and caused a seg-fault due to extra inode_unrefs. It fixes that extra inode_unref and cleans up the revalidation logic. Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1756 (NFS must revalidate inode on first ESTALE on lookup) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1756
* nfs: Undo selective changes in inode revalidationShehjar Tikoo2010-11-254-92/+2
| | | | | | | | | | | Those changes were brought in by commit 161850285ff06e90f2c990989bab9513dd4a4289: "nfs: re-implement logic to perform fresh lookups when lookup revalidates fail" Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1756 (NFS must revalidate inode on first ESTALE on lookup) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1756
* mgmt/glusterd: Move .cmd_log_history to log directoryVijay Bellur2010-11-251-1/+2
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2154 (Move .cmd_log_history to log directory) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2154
* mgmt/glusterd: Temporary fix for a crash seen in replace-brickVijay Bellur2010-11-251-2/+2
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2084 ([3.1.1qa5] : replace-brick fails to migrate data when migration from same hostname) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2084
* resolver: include par_idx access within the conditional sectionAnand Avati2010-11-241-1/+1
| | | | | | | | Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* volgen: put back an amended form of check erased in 96f44835 ("do not error ↵Csaba Henk2010-11-241-0/+10
| | | | | | | | | | out if brick count is less than sub count...") Signed-off-by: Csaba Henk <csaba@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2054 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2054
* volgen: clean up 0fbf226c (... add nfs.enable-ino32 as an option ...)Csaba Henk2010-11-241-96/+46
| | | | | | | | Signed-off-by: Csaba Henk <csaba@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1972 (xcs get doesn't work with gNFS) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1972
* display the error to the user if volume stop is given on a volume which does ↵Raghavendra Bhat2010-11-241-2/+4
| | | | | | | | | | not exist Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2140 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2140
* resolver: handle corner cases of handling root inode betterAnand Avati2010-11-241-4/+27
| | | | | | | | | | | | - changes in deep resolution and simple resolution of pure paths to handle resolution of "/" - warn on pure path resolution Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* nfs: treat GF_EVENT_CHILD_CONNECTING as subvolume up statusAnand Avati2010-11-181-0/+1
| | | | | | | | | | | | | | | GF_EVENT_CHILD_CONNECTING is sent as a status from underlying subvolumes after a sufficient margin of time (tcp connect() timeout) if the subvolumes are not reachable. This should avoid the problem of NFS not being reachable if any volume is down Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2093 (volumes cannot start when one node in a replicated setup is down) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2093
* features/locks: Send prelock unlock only if it is not grantable and is a ↵Pavan Sondur2010-11-181-6/+11
| | | | | | | | | | blocking lock call. Signed-off-by: Pavan Vilas Sondur <pavan@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1017 (Locking deadlock when upgrading lock) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1017
* Remove spurious inode_ref call on parent dir in fuse_create_cbkshishir gowda2010-11-181-1/+0
| | | | | | | | | | | The additional ref caused leaks. The ref is taken in __dentry_create called from inode_link Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2117 ([glusterfs-3.1.1qa7]: memleak in glusterfsd) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2117
* error out if the data-self-heal-window-size is less than 1Raghavendra Bhat2010-11-181-2/+2
| | | | | | | | Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2120 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2120
* mgmt/glusterd: Avoid creating multiple destination brickinfo during ↵Vijay Bellur2010-11-181-5/+33
| | | | | | | | | | replace-brick Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2084 ([3.1.1qa5] : replace-brick fails to migrate data when migration from same hostname) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2084
* cluster/pump: send PARENT_UP to dst_child after init is successfulAmar Tumballi2010-11-181-0/+6
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2122 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2122
* protocol/client: skip notify if query portmap is successfulshishir gowda2010-11-181-1/+2
| | | | | | | | | | Mount point does not hang, though stale handle error on / till the downed brick is online again Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1905 (Mounting volume is not working when any one server is down) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1905
* protocol/client: use the new rpc_clnt initialization methodsAnand Avati2010-11-161-12/+2
| | | | | | | | | | | | | This way of initializing an rpc_clnt object by splitting into a passive rpc_clnt_new and a delayed rpc_clnt_start between which registration of callback programs and notify methods are performed provides a race-free way to avoid the missing of connect notifications Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2078 (Volume Migration is not working) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2078
* nfsrpc: Fail auth-unix if more than 16 gids foundShehjar Tikoo2010-11-161-0/+5
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2045 (Write permission denied for non-primary group membership) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2045
* Stripe I/O exits when graph changesshishir gowda2010-11-151-0/+1
| | | | | | | | | | Adding event CONNECTING to be handled like CHILDUP in notify. Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2111 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2111
* protocol/client: fix ping timeout checks in reconfigure()Amar Tumballi2010-11-151-4/+4
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2104 ([3.1.1qa5]: network.ping-timeout set failed) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2104
* fuse: dentry create time inode resolution fixesAmar Tumballi2010-11-151-20/+30
| | | | | | | | | | | | treat successful 'inode' resolution as stale entries in case of entry creation fops, instead of continueing with an error log with old entry. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1240 (DVM: after graph change, inodes should resolve to new inode-table) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1240
* Solaris: df -h returned invalid sizeshishir gowda2010-11-151-2/+2
| | | | | | | | | | | | The reason seems to have been multiplying fields with f_bsize, while they are in f_frsize units. On linux both f_bize and f_frbsize seems to have been the same Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1021 (NFS: df -h shows incorrect out put) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1021
* Remove symlnk of log dir to /etc/glusterdshishir gowda2010-11-151-2/+2
| | | | | | | | | | | Hence forth the default log dir will be /usr/local/var/log/glusterfs Missed out on log locate changes Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2109 (Remove symbolic link of log directory from /etc/glusterd) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2109
* Remove symlnk of default log dir to /etc/glusterdshishir gowda2010-11-152-13/+4
| | | | | | | | | | Hence forth the default log dir will be /usr/local/var/log/glusterfs Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2109 (Remove symbolic link of log directory from /etc/glusterd) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2109
* glusterd rebalance: update the status in failure casesAmar Tumballi2010-11-151-2/+9
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1999 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1999
* Revert "nfs: Use only 38467 as default port instead of 3846[567]"Shehjar Tikoo2010-11-153-4/+3
| | | | | | | | | | | | | | | This reverts commit be5c02a81c19336a6b922b1e1f28293c90955e7f. By default, continue to register the three original port numbers so that an upgrade to future version from 3.1 release does not break mount requests against portmap, which may have old port numbers registered. Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1743 (XenServer is not compatible with GlusterNFS) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1743
* nfs: opendir/closedir for every readdirAnand Avati2010-11-152-14/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Revert "nfs3: Unref & unbind dir fd with inode lock on EOF" This reverts commit 4e6fb304ce41acbaf7c9ba67c06bf443e65082e8. The above commit (which unbinds fds at EOF) does not fix the original bug (1619) because a readdir from a second app could have already started before the readdir_cbk of the first app's readdir reaches NFS code. Hence the race still exists. Performing extra unrefs when EOF is received is not a reliable way of detecting that a client has performed a closedir (and to close the fd ourselves). Neither is interpreting a 0 cookies a new opendir. Clients can always use telldir/seekdir and hit EOFs twice. Due to the way NFS3 protocol is designed, it is just not possible for the server to reliably detect opendirs/closedirs performed by the client and map the corresponding readdirs to the same dir fd on the server side. The only reliable way of fixing this is to perform opendir/closedir at the cost of performance. Any optimization towards keeping dir fds open attempting to map them with application's opendir/closedir will either result in fd leaks or extra fd unrefs. Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2061 (NFS server crashes in readdir_fstat_cbk due to extra fd unref) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2061
* cluster/replicate: Fix memory leak in afr_fd_ctx_cleanup.Raghavendra G2010-11-151-0/+3
| | | | | | | | Signed-off-by: Raghavendra G <raghavendra@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1438 (memory leaks) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1438
* check for dict also while setting the port for source brick while doing ↵Raghavendra Bhat2010-11-141-1/+1
| | | | | | | | | | replace brick Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2084 ([3.1.1qa5] : replace-brick fails to migrate data when migration from same hostname) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2084
* mgmt/glusterd: fixes for uninterrupted replace-brick with nfsVijay Bellur2010-11-141-8/+26
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2084 ([3.1.1qa5] : replace-brick fails to migrate data when migration from same hostname) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2084
* cluster/pump: Reset saved path upon pump completionVijay Bellur2010-11-141-0/+2
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2084 ([3.1.1qa5] : replace-brick fails to migrate data when migration from same hostname) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2084
* performance/quick-read: handle the errors in qr_loc_fill properly when an ↵Raghavendra G2010-11-131-2/+24
| | | | | | | | | | | | fop needs to do open before continuing. - without need_unwind set, 'goto out' will result in infinite loop. Signed-off-by: Raghavendra G <raghavendra@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2075 (inode_from_path silently consumes 99% CPU causing nfs client disconnects) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2075
* mgmt/glusterd: "peer probe new-hostname" should replace old-hostname of friendPranith K2010-11-136-86/+78
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1995 (Gluster Peer probe command keeps ip address instead of server name in the destination server) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1995
* stripe: allow lookup on an entry if other than first subvolume is downAmar Tumballi2010-11-131-1/+1
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2099 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2099
* nfsrpc: Change log levels for RPC program search messagesShehjar Tikoo2010-11-131-3/+3
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2072 (NFS server crash in __nfs_rpcsvc_program_actor) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2072
* distribute: handle a case of 'local->key' being free'd twice.Amar Tumballi2010-11-131-1/+0
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2097 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2097
* rpcsvc: Fix crash in program search after portmap registration failureShehjar Tikoo2010-11-102-16/+29
| | | | | | | | | | | | | Failure to register with portmap was leaving the registered programs list in inconsistent state. This was causing a crash. portmap registration can fail if there is an NFS server already running. Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2072 (NFS server crash in __nfs_rpcsvc_program_actor) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2072
* protocol/server: fix log level of message in server-resolver.cAnand Avati2010-11-091-1/+1
| | | | | | | | Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 1092 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1092
* replicate: optimistic changelogAnand Avati2010-11-093-13/+101
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The standard way of maintaining changelog in replicate has been to write out pending flags and to unset the pending flag post the actual operation. This new optimization kicks in only when all subvolumes are up. The optimization is that, during pre-op, no changelog is written for METADATA and ENTRY/RENAME operations. If during the operation nothing failed, no changelog is updated in post-op either. If however, something does fail during an operation, then, pending flags get written during post op pointing only towards the failed nodes. DATA transactions continue to work the way they are. If one subvolume is down, pending flags are written in pre-op changelog itself as before. The impact of this optimization is only in the case when both servers die or the client dies while the 'FOP' stage of the transaction is in progress. By nature of METADATA and ENTRY operations, detecting a mismatch later is not dependent on the presence of changelog. Changelog only determines the direction in which self-heal happens for these types of transactions. For the direction too this optimization does not have a major impact because in the cases of failure (both servers dieing or client dieing) the final state (direction of self-heal) would be arbitrary anyways as the syscall wouldn't have completed. Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2068 (performance enhancements) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2068
* client: Return correct pre/post iatt for dst during renameShehjar Tikoo2010-11-091-1/+1
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Anand V. Avati <avati@dev.gluster.com> BUG: 2050 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2050
generated by cgit (git 2.34.1) at 2024-11-15 09:13:04 +0000