summaryrefslogtreecommitdiffstats
path: root/xlators
Commit message (Collapse)AuthorAgeFilesLines
* Changes to replace flock with gf_flock across GlusterFS.Pavan Sondur2010-10-0141-156/+158
| | | | | | | | Signed-off-by: Pavan Vilas Sondur <pavan@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 865 (Add locks recovery support in GlusterFS) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=865
* cluster/pump: Initiailize pthread mutex and saved fds list.Pavan Sondur2010-10-011-0/+3
| | | | | | | | Signed-off-by: Pavan Vilas Sondur <pavan@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 865 (Add locks recovery support in GlusterFS) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=865
* remove 'gfs_id' field from all protocol xdr structuresAmar Tumballi2010-10-014-176/+86
| | | | | | | | | | | | | It was kept as a debugging requirement/placeholder. Because RPC is mostly bug free with regard to basic xdr structures, it doesn't make sense to carry this forward in release. Saves 8bytes in each request. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@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 if the table is NULL before accessing itRaghavendra Bhat2010-10-011-1/+6
| | | | | | | | Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1757 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1757
* nfs3: Revalidate inode on receiving ESTALE on lookupShehjar Tikoo2010-10-012-1/+51
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Vijay Bellur <vijay@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
* cluster/afr: Recover locks on child_up from source to sink.Pavan Sondur2010-10-016-5/+414
| | | | | | | | Signed-off-by: Pavan Vilas Sondur <pavan@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 865 (Add locks recovery support in GlusterFS) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=865
* mgmt/glusterd: Make replace-brick atomicVijay Bellur2010-10-011-18/+54
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* glusterfsd: perform init new graph if re-configure failsVijay Bellur2010-09-301-8/+27
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* protocol/client: cluster/afr: Support lock recovery and self heal.Pavan Sondur2010-09-3022-73/+2106
| | | | | | | | Signed-off-by: Pavan Vilas Sondur <pavan@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 865 (Add locks recovery support in GlusterFS) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=865
* glusterd: fix in log filename and log rotatev3.1.0qa36Amar Tumballi2010-09-301-18/+64
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1731 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1731
* mgmt/glusterd: memory leak fixesPranith K2010-09-302-0/+8
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1726 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1726
* Block add, remove and replace brick ops when rebalance is in progressshishir gowda2010-09-304-6/+52
| | | | | | | | | | Also fixing msg, changing defrag to rebalance Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* start the per brick process with 'glusterfsd' instead of 'glusterfs'Amar Tumballi2010-09-301-1/+1
| | | | | | | | | | | for users, it becomes easy to differenciate which process is exporting bricks, and which is used as a client process. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* mgmt/glusterd: remove debug messages in cleanup_and_exitPranith K2010-09-301-1/+2
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1746 (volume delete fails after volume stop) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1746
* protocol/client: handle more NULL dereferenceAmar Tumballi2010-09-301-43/+50
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1745 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1745
* protocol/server: set 'this->private' only if 'init()' is successfulAmar Tumballi2010-09-291-2/+1
| | | | | | | | | | | | | If volume file is fetched (due to HUP or volume file changes) and server init is called again, server protocol init fails with 'bind to port failed' error. If 'this->private' is set to new 'conf' before successful init, the whole process goes to corrupted state. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* replicate: remove checks which prevented self-heal when open fds were presentAnand Avati2010-09-292-16/+1
| | | | | | | | | | | | | this check is not needed anymore since the introduction of changelog piggybacking as the optimization technique instead of first-write-to-flush technique some of the self-healing issues with NFS mounts should be resolved Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* replicate: keep read_child in inode ctx as up-to-date as possibleAnand Avati2010-09-293-4/+50
| | | | | | | | | | | | | In every transaction check if the currently set read child in the inode context failed in the fop and set it to another subvol on which the latest fop has passed. This will prevent read fops landing on subvols which have witnessed a failure. Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1172 (ls -lh on NFS mount of 2-mirror replicate gives incorrect file size) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1172
* pump: skip "." and ".." entries while traversing dir treeAnand Avati2010-09-291-0/+4
| | | | | | | | Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* mgmt/glusterd: resolve uuid if nullPranith K2010-09-291-0/+5
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1734 (replace brick should accept both ipaddress and hostname for bricks) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1734
* glusterd: make sure init is not called when there is no topology changeKaushik BV2010-09-291-1/+1
| | | | | | | | Signed-off-by: Kaushik BV <kaushikbv@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1159 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1159
* mgmt/glusterd: set the error before starting checksPranith K2010-09-291-0/+1
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1734 (replace brick should accept both ipaddress and hostname for bricks) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1734
* mgmt/glusterd: volume start forcePranith K2010-09-299-274/+284
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1736 (implement volume start force) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1736
* storage/posix: prevent chmod() from getting called on symlinksAnand Avati2010-09-291-6/+25
| | | | | | | | | | | | symlinks, if at all their mode can be changed (on non-linux platforms) can only be done via lchmod(). Attempting chmod on a symlink can be disastrous. Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@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
* features/quiesce: new translatorAmar Tumballi2010-09-296-1/+1654
| | | | | | | | | | | | | | | | | 'Quiesce' is used to describe pausing or altering the state of running processes on a computer In GlusterFS context, this translator looks for CHILD_DOWN/CHILD_UP events, and will hold the requests for some time if the child node is in disconnected state current timeout is 20seconds, will be made tunable soon Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1500 (Mount point should not be in-accessible between reconnect to server) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1500
* replicate: clear pre_op_done/piggyback values after open fd self-healAnand Avati2010-09-291-2/+11
| | | | | | | | Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* replicate: use new style (un)locking in self heal of missing entriesAnand Avati2010-09-291-62/+5
| | | | | | | | Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 960 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=960
* replicate: replace first-write-to-flush optimizationAnand V. Avati2010-09-296-594/+422
| | | | | | | | | | | | use a changelog piggybacking optimization instead of first-write-to-flush optimization and do other cleanups (removal of post-post-op hook etc.) Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* whitespace cleanupAnand V. Avati2010-09-294-1928/+1937
| | | | | | | | | Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Anand V. Avati <avati@amp.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* mgmt/glusterd: brick info get should compare uuid instead of hostnamePranith K2010-09-281-3/+6
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1734 (replace brick should accept both ipaddress and hostname for bricks) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1734
* client-handshake: prevent NULL dereferenceAmar Tumballi2010-09-281-11/+33
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1720 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1720
* resolver: set safer conditionals during path creationAnand Avati2010-09-281-1/+1
| | | | | | | | | | | | when client would send just the path as part of a transaction, path reconstruction would create parent directory's path instead of actual path at the time of resolution. Signed-off-by: Anand V. Avati <avati@blackhole.gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 935 (Directories change mode from 0755 to 0644 automatically) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=935
* mgmt/glusterd: Fix incorrect state machine transitionv3.1.0qa35Vijay Bellur2010-09-281-1/+1
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* mgmt/glusterd: compute cksum upon volume retrievalVijay Bellur2010-09-281-7/+12
| | | | | | | | Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 971 (dynamic volume management) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=971
* Changing readdir calls to readdirp calls in DHTshishir gowda2010-09-282-21/+12
| | | | | | | | | | | | Changing readdir calls to readdirp calls to handle layout issues seen in dht. Also, changing posix readdir call to do stat after dir read for readdirp. Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1690 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1690
* nfs: Free inodes layer fop locals to fix memory leakShehjar Tikoo2010-09-281-10/+18
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1259 (Memory leaks to 5.7g usage in gnfs) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1259
* distribute: Propagate -1 op_ret on failed fopShehjar Tikoo2010-09-281-0/+1
| | | | | | | | Signed-off-by: Shehjar Tikoo <shehjart@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1717 (dht_attr_cbk does not propagate op_ret on failed fop causing nfs crash) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1717
* do not allow remove brick if the total bricks given is less than the sub_countRaghavendra Bhat2010-09-281-6/+18
| | | | | | | | Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1713 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1713
* access-control: report possible errors in graph during 'init()'Amar Tumballi2010-09-281-1/+17
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1712 (Error is not logged for a dangling volume) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1712
* glusterd: logs is a symlink to default var log dirAmar Tumballi2010-09-283-4/+5
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1728 (glusterd log file location) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1728
* unref the dict allocated in set_volume, remove_brickRaghavendra Bhat2010-09-281-0/+7
| | | | | | | | Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1726 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1726
* Starting/stoppin glusterd changes brick pairsshishir gowda2010-09-281-15/+64
| | | | | | | | | | | | | the problem was the pair information was not maintained anywhere. When glusterd re-starts, it reads the brick dir entries and populates the brick list. adding brick info to volume info file to fix this issue Signed-off-by: shishir gowda <shishirng@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1714 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1714
* mgmt/glusterd: prevent detach of peer with its bricks existing in clusterPranith K2010-09-285-110/+234
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1681 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1681
* distribute: don't check for inode's gfid in _cbkv3.1.0qa34Amar Tumballi2010-09-271-2/+1
| | | | | | | | | | | | * in dht_lookup_dir_cbk(), at the entry itself we are dereferencing local->inode, which can be NULL in many cases. Hence no need to check/dereference that variable. Just check 'local->gfid'. Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1696 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1696
* replace brick fixesPavan Sondur2010-09-276-222/+392
| | | | | | | | | | | | | | | | | Reviewed ok. Just like commit, maybe we can move 'pause' also to the commit op of replace brick since it can be performed only when replace brick has been started. I'll send in a patch for that soon. ----- Original Message ----- From: "Vijay Bellur" <vijay@gluster.com> To: glusterfs@dev.gluster.com Sent: Monday, September 27, 2010 7:59:25 PM Subject: [PATCH BUG:1235] replace brick fixes Signed-off-by: Vijay Bellur <vijay@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1235 (Bug for all pump/migrate commits) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1235
* io-threads: fix NULL dereference in 'reconfigure()'Amar Tumballi2010-09-271-6/+5
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1715 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1715
* mgmt/glusterd: delay probe till connection happensPranith K2010-09-279-178/+224
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1607 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1607
* access-control: handle a case of wrong error handlingAmar Tumballi2010-09-271-1/+1
| | | | | | | | | | | in rename_src_stat_cbk(), we were checking the return value as '-1', but the function can return '-EFAULT' (ie, -14 in few cases). Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1691 () URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1691
* fix a build warning about stack overflowAmar Tumballi2010-09-271-12/+12
| | | | | | | | Signed-off-by: Amar Tumballi <amar@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1710 (build warning) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1710
* mgmt/glusterd: perform post operations of volume setPranith K2010-09-271-5/+12
| | | | | | | | Signed-off-by: Pranith Kumar K <pranithk@gluster.com> Signed-off-by: Vijay Bellur <vijay@dev.gluster.com> BUG: 1708 (update store, re-compute checksum, restart nfs for volume set) URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1708