summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* geo-rep: Fix deadlock during worker startKotresh HR2018-09-212-4/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Analysis: Monitor process spawns monitor threads (one per brick). Each monitor thread, forks worker and agent processes. Each monitor thread, while intializing, updates the monitor status file. It is synchronized using flock. The race is that, some thread can fork worker while other thread opened the status file resulting in holding the reference of fd in worker process. Cause: flock gets unlocked either by specifically unlocking it or by closing all duplicate fds referring to the file. The code was relying on fd close, hence a reference in worker/agent process by fork could cause the deadlock. Fix: 1. flock is unlocked specifically. 2. Also made sure to update status file in approriate places so that the reference is not leaked to worker/agent process. With this fix, both the deadlock and possible fd leaks is solved. Backport of: > Patch: https://review.gluster.org/20704 > BUG: bz#1614799 > Change-Id: I0d1ce93072dab07d0dbcc7e779287368cd9f093d > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1630145 Change-Id: I0d1ce93072dab07d0dbcc7e779287368cd9f093d Signed-off-by: Kotresh HR <khiremat@redhat.com>
* geo-rep/hook-script: Fix ssh/scp optionsKotresh HR2018-09-215-24/+84
| | | | | | | | | | | | | | | | | | | | Always use ssh and scp with "-oPasswordAuthentication=no" and "-oStrictHostKeyChecking=no" options. It might hang the post script otherwise leading geo-rep setup failure Also increased geo-rep timeout. Occasionally, it's taking more time to reach Active/Passive status. Especially, the first start after create. Backport of: > Patch: https://review.gluster.org/20601 > BUG: bz#1610405 > Change-Id: I9560d64dbe0edf5db73446a9fc97dda19b88d233 > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1630144 Change-Id: I9560d64dbe0edf5db73446a9fc97dda19b88d233 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* rpc: handle EAGAIN when SSL_ERROR_SYSCALL is returnedMilind Changire2018-09-211-2/+9
| | | | | | | | | | | | | | | | | Problem: A return value of ENODATA was forcibly returned in the case where SSL_get_error(r) returned SSL_ERROR_SYSCALL. Sometimes SSL_ERROR_SYSCALL is a transient error which is identified by setting errno to EAGAIN. EAGAIN is not a fatal error and indicates that the syscall needs to be retried. Solution: Bubble up the errno in case SSL_get_error(r) returns SSL_ERROR_SYSCALL and let the upper layers handle it appropriately. fixes: bz#1601356 Change-Id: I76eff278378930ee79abbf9fa267a7e77356eed6 Signed-off-by: Milind Changire <mchangir@redhat.com>
* storage/posix: Avoid log flood in posix_set_parent_ctime()Vijay Bellur2018-09-171-4/+0
| | | | | | | | | | | | | | | | | | | | posix_set_parent_ctime() unconditionally logs an error if consistent time attributes is not enabled. This log does not add any value, prints an incorrect errno & floods the log file. Hence nuking this log message in this patch. Backport of : > Patch: https://review.gluster.org/20547/ > Change-Id: I82a78f2f8ce5ab518f8cdf6d9086a97049712f75 > BUG: 1607049 > Signed-off-by: Vijay Bellur <vbellur@redhat.com> (cherry picked from commit e0df887ba044ce92e9a2822be9261d0f712b02bd) Change-Id: I82a78f2f8ce5ab518f8cdf6d9086a97049712f75 fixes: bz#1629548 Signed-off-by: Vijay Bellur <vbellur@redhat.com>
* doc: Release notes for v4.1.4v4.1.4Jiffin Tony Thottan2018-09-061-0/+41
| | | | | | Change-Id: Idfce8b9ec79303b92045e68ab98765f7e2f98940 fixes: bz#1623161 Signed-off-by: Jiffin Tony Thottan <jthottan@redhat.com>
* posix: disable open/read/write on special filesAmar Tumballi2018-09-061-0/+33
| | | | | | | | | | | | | | | | | | | In the file system, the responsibility w.r.to the block and char device files is related to only support for 'creating' them (using mknod(2)). Once the device files are created, the read/write syscalls for the specific devices are handled by the device driver registered for the specific major number, and depending on the minor number, it knows where to read from. Hence, we are at risk of reading contents from devices which are handled by the host kernel on server nodes. By disabling open/read/write on the device file, we would be safe with the bypass one can achieve from client side (using gfapi) Fixes: bz#1625096 Change-Id: I48c776b0af1cbd2a5240862826d3d8918601e47f Signed-off-by: Amar Tumballi <amarts@redhat.com>
* protocol: don't use allocaAmar Tumballi2018-09-062-103/+58
| | | | | | | | | | | | current implementation of alloca can cause issues when strings larger than the allocated buffer is passed to the xdr. Hence it makes sense to allow XDR decode functions to deal with memory allocations, which we can free later. Fixes: bz#1625097 Change-Id: I3a05553f5702de9575c244649ca0e5ac9abaac94 Signed-off-by: Amar Tumballi <amarts@redhat.com>
* io-stats: dump io-stats info in /var/run/glusterAmar Tumballi2018-09-062-15/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | It wouldn't make sense to allow iostats file to be written in *any* directory. While the formating makes sure we try to append io-stats-name for the file, so overwriting existing file is slim, but in any case it makes sense to restrict dumping to one directory. Below are the sample commands, and files created for the corresponding values: $ setfattr -n trusted.io-stats-dump -v file-for-dump $M0 In this case, the file would be in /var/run/gluster/file-for-dump $ setfattr -n trusted.io-stats-dump -v /dir1/dir2/file-for-dump $M0 In this case, then the dump file is in /var/run/gluster/dir1-dir2-file-for-dump Note that the value passed for this virtual xattr would be treated as a file, and even if the value has '/' in it, it would be changed to '-' for sanity. Fixes: bz#1625106 Change-Id: Id9ae6a40a190b8937c51662e6e1c2a0f6c86a0e0 Signed-off-by: Amar Tumballi <amarts@redhat.com>
* server-protocol: don't allow '../' path in 'name'Amar Tumballi2018-09-062-0/+18
| | | | | | | | | | | | | This will prevent any arbitrary file creation through glusterfs by modifying the client bits. Also check for the similar flaw inside posix too, so we prevent any changes in layers in-between. Fixes: bz#1625095 Signed-off-by: Amar Tumballi <amarts@redhat.com> Change-Id: Id9fe0ef6e86459e8ed85ab947d977f058c5ae06e
* dict: handle negative key/value length while unserializeAmar Tumballi2018-09-061-1/+2
| | | | | | Fixes: bz#1625089 Change-Id: Ie56df0da46c242846a1ba51ccb9e011af118b119 Signed-off-by: Amar Tumballi <amarts@redhat.com>
* posix: remove not supported get/set contentAmar Tumballi2018-09-064-192/+1
| | | | | | | | | | | | | | | | getting and setting a file's content using extended attribute worked great as a GET/PUT alternative when an object storage is supported on top of Gluster. But it needs application changes, and also, it skips some caching layers. It is not used over years, and not supported any more. Removing the dead code. Fixes: bz#1625102 Change-Id: Ide3b3f1f644f6ca58558bbe45561f346f96b95b7 Signed-off-by: Amar Tumballi <amarts@redhat.com>
* storage/posix: Increment trusted.pgfid in posix_mknodN Balachandran2018-08-292-4/+71
| | | | | | | | | | The value of trusted.pgfid.xx was always set to 1 in posix_mknod. This is incorrect if posix_mknod calls posix_create_link_if_gfid_exists. Change-Id: Ibe87ca6f155846b9a7c7abbfb1eb8b6a99a5eb68 fixes: bz#1623317 Signed-off-by: N Balachandran <nbalacha@redhat.com>
* doc: Release notes for v4.1.3v4.1.3ShyamsundarR2018-08-271-0/+36
| | | | | | Change-Id: Ia944a2353f0a89b61f66bde7f270489dff3793b4 fixes: bz#1607945 Signed-off-by: ShyamsundarR <srangana@redhat.com>
* gfapi : Handle the path == "" glfs_resolve_atJiffin Tony Thottan2018-08-221-7/+10
| | | | | | | | | | | | Currently there is no check for path = "" in glfs_resolve_at. So if application sends an empty path, then the function resolves into the parent inode which is incorrect. Plus modified possible of "path" with "origpath" in the same function. Change-Id: Ie5ff9ce4b771607b7dbb3fe00704fe670421792a fixes: bz#1618347 Signed-off-by: Jiffin Tony Thottan <jthottan@redhat.com> (cherry picked from commit febee007bb1a99d65300630c2a98cbb642b1c8dc)
* Bash integration script should namespace variablesMark Mielke2018-08-211-20/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the original submitted script, it looks like there was effort put into namespacing all global variables. However a few mistakes remained. GLUSTER_TOP_SUBOPTIONSx were defined, but TOP_SUBOPTIONSx were referenced. This was likely an unrecognized defect in the original code submission? These are now corrected to refer to GLUSTER_TOP_SUBOPTIONSx. FINAL_LIST, LIST, and TOP were leaked into all Bash shells and used by the command completion functions. The most problematic of these was TOP, which was declared with "-i" making it an integer. This cause other code which used TOP to define a path to fail like this: $ bash $ TOP=/abc bash: /abc: syntax error: operand expected (error token is "/abc") These are now qualified as GLUSTER_FINAL_LIST, GLUSTER_LIST, and GLUSTER_TOP to reduce impact on scripts that might choose to use these extremely common variable names. > Change-Id: Ic96eda8efd1f3238bbade6c6ddb69118e8d82158 > Signed-off-by: Mark Mielke <mark.mielke@gmail.com> (cherry picked from commit 89545e745e4075845c18078be67a31dea93a4e88) Change-Id: Ic96eda8efd1f3238bbade6c6ddb69118e8d82158 Fixes: bz#1425326 Signed-off-by: Mark Mielke <mark.mielke@gmail.com>
* dht: Delete MDS internal xattr from dict in dht_getxattr_cbkMohit Agrawal2018-08-162-31/+4
| | | | | | | | | | | | | | | | | | Problem: At the time of fetching xattr to heal xattr by afr it is not able to fetch xattr because posix_getxattr has a check to ignore if xattr name is MDS Solution: To ignore same xattr update a check in dht_getxattr_cbk instead of having a check in posix_getxattr Backport of: > BUG: 1584098 > Change-Id: I86cd2b2ee08488cb6c12f407694219d57c5361dc > Signed-off-by: Mohit Agrawal <moagrawa@redhat.com> Change-Id: I86cd2b2ee08488cb6c12f407694219d57c5361dc fixes: bz#1611116 Signed-off-by: Mohit Agrawal <moagrawa@redhat.com>
* geo-rep : fix possible crashSunny Kumar2018-08-161-2/+5
| | | | | | | | | | | | | | | | | Problem : In 'glusterd_verify_slave' while tokenizing error message we call 'strtok_r' and store return value in 'tmp' which can be NULL. We are passing this 'tmp' as 1st argument to 'strcmp' which will lead to segmentation fault. Solution : before calling 'strcmp' we should NULL check 'tmp'. Backport of: > Change-Id: Ifd3864b904afe6cd09d9e5a4b55c6d0578e22b9d > BUG: 1602121 > Signed-off-by: Sunny Kumar <sunkumar@redhat.com> Change-Id: Ifd3864b904afe6cd09d9e5a4b55c6d0578e22b9d fixes: bz#1611115 Signed-off-by: Sunny Kumar <sunkumar@redhat.com>
* geo-rep: Fix issues with gfid conflict handlingKotresh HR2018-08-163-64/+169
| | | | | | | | | | | | | | | | | | | | | | | | | | | | 1. MKDIR/RMDIR is recorded on all bricks. So if one brick succeeds creating it, other bricks should ignore it. But this was not happening. The fix rename of directories in hybrid crawl, was trying to rename the directory to itself and in the process crashing with ENOENT if the directory is removed. 2. If file is created, deleted and a directory is created with same name, it was failing to sync. Again the issue is around the fix for rename of directories in hybrid crawl. Fixed the same. If the same case was done with hardlink present for the file, it was failing. This patch fixes that too. Backport of: > BUG: 1598884 > Change-Id: I6f3bca44e194e415a3d4de3b9d03cc8976439284 > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1611114 Change-Id: I6f3bca44e194e415a3d4de3b9d03cc8976439284 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* geo-rep: Fix symlink rename syncing issueKotresh HR2018-08-161-2/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: Geo-rep sometimes fails to sync the rename of symlink if the I/O is as follows 1. touch file1 2. ln -s "./file1" sym_400 3. mv sym_400 renamed_sym_400 4. mkdir sym_400 The file 'renamed_sym_400' failed to sync to slave Cause: Assume there are three distribute subvolume (brick1, brick2, brick3). The changelogs are recorded as follows for above I/O pattern. Note that the MKDIR is recorded on all bricks. 1. brick1: ------- CREATE file1 SYMLINK sym_400 RENAME sym_400 renamed_sym_400 MKDIR sym_400 2. brick2: ------- MKDIR sym_400 3. brick3: ------- MKDIR sym_400 The operations on 'brick1' should be processed sequentially. But since MKDIR is recorded on all the bricks, The brick 'brick2/brick3' processed MKDIR first before 'brick1' causing out of order syncing and created directory sym_400 first. Now 'brick1' processed it's changelog. CREATE file1 -> succeeds SYMLINK sym_400 -> No longer present in master. Ignored RENAME sym_400 renamed_sym_400 While processing RENAME, if source('sym_400') doesn't present, destination('renamed_sym_400') is created. But geo-rep stats the name 'sym_400' to confirm source file's presence. In this race, since source name 'sym_400' is present as directory, it doesn't create destination. Hence RENAME is ignored. Fix: The fix is not rely only on stat of source name during RENAME. It should stat the name and if the name is present, gfid should be same. Only then it can conclude the presence of source. Backport of: > BUG: 1600405 > Change-Id: I9fbec4f13ca6a182798a7f81b356fe2003aff969 > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1611113 Change-Id: I9fbec4f13ca6a182798a7f81b356fe2003aff969 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* geo-rep: Fix geo-rep for older versions of unshareKotresh HR2018-08-163-7/+42
| | | | | | | | | | | | | | | | | | | Geo-rep mounts are private to worker. It uses mount namespace using unshare command to achieve the same. Well, the unshare command has to support '--propagation' option. So geo-rep breaks on the systems with older unshare version. The patch makes it fall back to lazy umount behaviour if the unshare does not support propagation option. Backport of: > BUG: 1589782 > Change-Id: Ia614f068aede288d63ac62fea4461b1865066054 > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1611111 Change-Id: Ia614f068aede288d63ac62fea4461b1865066054 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* glusterd: memory leak in geo-rep statusSanju Rakonde2018-08-161-2/+6
| | | | | | | | | | | Backport of: > BUG: 1580352 > Change-Id: I9648e73090f5a2edbac663a6fb49acdb702cdc49 > Signed-off-by: Sanju Rakonde <srakonde@redhat.com> fixes: bz#1611110 Change-Id: I9648e73090f5a2edbac663a6fb49acdb702cdc49 Signed-off-by: Sanju Rakonde <srakonde@redhat.com>
* geo-rep/scheduler: Fix crashKotresh HR2018-08-161-35/+35
| | | | | | | | | | | | | | | | | | | Fix crash where session_name is referenced before assignment. Well, this is a corner case where the geo-rep session exists and the status output doesn't show any rows. This might happen when glusterd is down or when the system is in inconsistent state w.r.t glusterd. Backpor of: > BUG: 1576179 > Change-Id: Iec1557e01b35068041b4b3c1aacee2bfa0e05873 > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1611108 Change-Id: Iec1557e01b35068041b4b3c1aacee2bfa0e05873 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* glusterd/geo-rep: Fix glusterd crashKotresh HR2018-08-151-1/+1
| | | | | | | | | | | | | | | | | Using strdump instead of gf_strdup crashes during free if mempool is being used. gf_free checks the magic number in the header which will not be taken care if strdup is used. Backport of: > BUG: 1576392 > Change-Id: Iab36496554b838a036af9d863e3f5fd07fd9780e > Signed-off-by: Kotresh HR <khiremat@redhat.com> fixes: bz#1611106 Change-Id: Iab36496554b838a036af9d863e3f5fd07fd9780e Signed-off-by: Kotresh HR <khiremat@redhat.com>
* geo-rep: Fix upgrade issueKotresh HR2018-08-151-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Cause and Analysis: The last synced changelog for entry operations is marked in current version to avoid re-processing of already processed entry operations in a batch during crash/restart of geo-rep. This was not present in previous versoins. The marker is maintained in the dictionary with the key 'last_synced_entry' and dictionary is persisted into status file. So upgrading to current version in which the marker is present was failing with KeyError. Solution: Load the dictionary with default keys first which contains all the keys including latest ones and then load the values from status file instead of doing otherwise. Backport of: > BUG: 1575490 > Change-Id: Ic654e6f9a3c97f616761f1362f890352a2186fb4 > Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit 23c1385b5f6f6103e820d15ecfe1df31940fdb45) fixes: bz#1611104 Change-Id: Ic654e6f9a3c97f616761f1362f890352a2186fb4 Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit 23c1385b5f6f6103e820d15ecfe1df31940fdb45)
* storage/posix: Fix excessive logging in WRITE fop pathKrutika Dhananjay2018-08-101-0/+3
| | | | | | | | | | | | | | | Backport of: https://review.gluster.org/#/c/glusterfs/+/20250 I was running some write-intensive tests on my volume, and in a matter of 2 hrs, the 50GB space in my root partition was exhausted. On inspecting further, figured that excessive logging in bricks was the cause - specifically in posix write when posix_check_internal_writes() does dict_get() without a NULL-check on xdata. Change-Id: I89de57a3a90ca5c375e5b9477801a9e5ff018bbf fixes: bz#1596686 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> (cherry picked from commit 81701e4d92ae7b1d97e5bc955703719f2e9e773a)
* rpm: do not build glusterfs-resource-agents on el6Niels de Vos2018-08-011-0/+4
| | | | | | | | | | | | | | | | glusterfs-resource-agents depends on glusterfs-server and this is not available on el6 with the current Gluster releases. It is not possible to install glusterfs-resource-agents on el6 without running into dependency problems, so do not build the sub-packae at all. > Change-Id: Ibe08ad3a1b7882559b4e445603d0508b9282b755 > Fixes: bz#1609551 > Signed-off-by: Niels de Vos <ndevos@redhat.com> (cherry picked from commit f354af3a56604c61e09e29c13fe8fb632c1013a8) Change-Id: Ibe08ad3a1b7882559b4e445603d0508b9282b755 Fixes: bz#1609550 Signed-off-by: Niels de Vos <ndevos@redhat.com>
* doc: Added release notes for 4.1.2v4.1.2ShyamsundarR2018-07-241-0/+32
| | | | | | Fixes: bz#1594908 Change-Id: I93db801c1292b8f57e35157ce2b503fefa6cb2f2 Signed-off-by: ShyamsundarR <srangana@redhat.com>
* glusterd: _is_prefix should handle 0-length pathsKaushal M2018-07-241-0/+9
| | | | | | | | | If one of the paths given to _is_prefix is 0-length, then it is not a prefix of the other. Hence, _is_prefix should return false. Change-Id: I54aa577a64a58940ec91872d0d74dc19cff9106d fixes: bz#1599785 Signed-off-by: Kaushal M <kaushal@redhat.com>
* doc: Correct release notes to remove py3 supportShyamsundarR2018-07-241-6/+1
| | | | | | | Updates: bz#1594908 Change-Id: I6cefb3b6abc1b38981da4959f1be24814be6888c Signed-off-by: ShyamsundarR <srangana@redhat.com>
* geo-rep/scheduler: Fix EBUSY trace backKotresh HR2018-07-231-5/+20
| | | | | | | | | | | | | | | | | | Fix the trace back during temporary mount cleanup. Temporary mount is done to touch the root required for checkpoint to complete. Backport of: > Patch: https://review.gluster.org/#/c/20476/ > BUG: 1598977 > Change-Id: I97fea538e92c4ef0747747e981ef98499504e336 > Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit 271510b09d0613887f983b1e3dad5c55cdc6cd07) fixes: bz#1600941 Change-Id: I97fea538e92c4ef0747747e981ef98499504e336 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* afr: switch lk_owner only when pre-op succeedsRavishankar N2018-07-232-15/+10
| | | | | | | | | | | | | | | | | | | | Problem: In a disk full scenario, we take a failure path in afr_transaction_perform_fop() and go to unlock phase. But we change the lk-owner before that, causing unlock to fail. When mount issues another fop that takes locks on that file, it hangs. Fix: Change lk-owner only when we are about to perform the fop phase. Also fix the same issue for arbiters when afr_txn_arbitrate_fop() fails the fop. Also removed the DISK_SPACE_CHECK_AND_GOTO in posix_xattrop. Otherwise truncate to zero will fail pre-op phase with ENOSPC when the user is actually trying to freee up space. Change-Id: Ic4c8a596b4cdf4a7fc189bf00b561113cf114353 fixes: bz#1603056 Signed-off-by: Ravishankar N <ravishankar@redhat.com> (cherry picked from commit ec0d7d77de3e4bd485a4fa2e53c9137e25c71ce7)
* posix: check before removing stale symlinkRavishankar N2018-07-231-4/+8
| | | | | | | | | | | | | | | BZ 1564071 complains of directories with missing gfid symlinks and corresponding "Found stale gfid handle" messages in the logs. Hence add a check to see if the symlink points to an actual directory before removing it. Note: Removing stale symlinks was added via commit 3e9a9c029fac359477fb26d9cc7803749ba038b2 Change-Id: I5d91fab8e5f3a621a9ecad4a1f9c898a3c2d346a Updates: bz#1603099 Signed-off-by: Ravishankar N <ravishankar@redhat.com> (cherry picked from commit 9ab218be5e69b9f71fe4eea9ca8d114b78cafd25)
* cluster/afr: Prevent execution of code after call_count decrementingPranith Kumar K2018-07-101-7/+8
| | | | | | | | | | | | | | | Problem: When call_count is decremented by one thread, another thread can go ahead with the operation leading to undefined behavior for the thread executing statements after decrementing call count. Fix: Do the operations necessary before decrementing call count. fixes bz#1599629 Change-Id: Icc90cd92ac16e5fbdfe534d9f0a61312943393fe Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> (cherry picked from commit 03f1f5bdc46076178f1afdf8e2a76c5b973fe11f)
* afr: heal gfids when file is not present on all bricksRavishankar N2018-07-096-12/+179
| | | | | | | | | | | | commit 20fa80057eb430fd72b4fa31b9b65598b8ec1265 introduced a regression wherein if a file is present in only 1 brick of replica *and* doesn't have a gfid associated with it, it doesn't get healed upon the next lookup from the client. Fix it. Change-Id: I7d1111dcb45b1b8b8340a7d02558f05df70aa599 fixes: bz#1597117 Signed-off-by: Ravishankar N <ravishankar@redhat.com> (cherry picked from commit eb472d82a083883335bc494b87ea175ac43471ff)
* cluster/afr: Make sure lk-owner is assigned at the time of lockPranith Kumar K2018-07-041-2/+1
| | | | | | | | | | | | | | | Problem: In the new eager-lock implementation lk-owner is assigned after the 'local' is added to the eager-lock list, so there exists a possibility of lock being sent even before lk-owner is assigned. Fix: Make sure to assign lk-owner before adding local to eager-lock list fixes bz#1598193 Change-Id: I26d1b7bcf3e8b22531f1dc0b952cae2d92889ef2 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> (cherry picked from commit c6f93e422855f656d3a86461a8458f37ad0103eb)
* posix/ctime: Fix differential ctime duing entry operationsKotresh HR2018-07-021-51/+60
| | | | | | | | | | | | | | | | | | | | | | | | We should not be relying on backend file's time attributes to load the initial ctime time attribute structure. This is incorrect as each replica set would have witnessed the file creation at different times. For new file creation, ctime, atime and mtime should be same, hence initiate the ctime structure with the time from the frame. But for the files which were created before ctime feature is enabled, this is not accurate but still fine as the times would get eventually accurate. Backport of: > Patch: https://review.gluster.org/#/c/20281/ > BUG: 1592275 > Change-Id: I206a469c83ee7b26da2fe096ae7bf8ff5986ad67 > Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit 841991130c94e3fcf4076917be6da9ce90406932) fixes: bz#1593537 Change-Id: I206a469c83ee7b26da2fe096ae7bf8ff5986ad67 Signed-off-by: Kotresh HR <khiremat@redhat.com>
* afr: don't update readables if inode refresh failed on all childrenRavishankar N2018-07-024-32/+50
| | | | | | | | | | | | | | | | | | | | | | | | | Problem: If inode refresh failed on all children of afr due to ENOENT (say file migrated by dht), it resets the readables to zero. Any inflight txn which then later comes on the inode fails with EIO because no readable children present for the inode. Fix: Don't update readables when inode refresh fails on *all* children of afr. In that way any inflight txns will either proceed with its own inode refresh if needed and fail it with the right errno or use the old value of readables and continue with the txn. Also, add quorum checks to the beginning of afr_transaction(). Otherwise, we seem to be winding the lock and checking for quorum only in pre-op pahse. Note: This should ideally fix BZ 1329505 since the stop gap fix for it is has been reverted at https://review.gluster.org/#/c/20028. Change-Id: Ia638c092d8d12dc27afb3cdad133394845061319 updates: bz#1597116 Signed-off-by: Ravishankar N <ravishankar@redhat.com> (cherry picked from commit 0f13eed0c1fa74cefed486538b02e0c8a8708456)
* ctime: Fix self heal of symlink in EC volumeKotresh HR2018-07-026-6/+71
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Since IEEE Std 1003.1-2001 does not require any association of file times with symbolic links, there is no requirement that file times be updated by readlink() states [1]. stat on symlink file was generating a readlink fop on one of the subvolumes of ec set which in turn updates atime on that subvolume. This causes mdata xattr to be different across ec set and hence self heal fails. So based on [1], atime is no longer updated by readlink fop. [1] http://pubs.opengroup.org/onlinepubs/009695399/functions/readlink.html Backport of: > Patch: https://review.gluster.org/#/c/20311/ > BUG: 1592509 > Change-Id: I08bd3ca3bdb222bd18160b1aa58fc2f7630c8083 > Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit c097a7894d458e33a41f6db6092677108ef30fec) fixes: bz#1593536 Change-Id: I08bd3ca3bdb222bd18160b1aa58fc2f7630c8083 Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit c097a7894d458e33a41f6db6092677108ef30fec)
* storage/posix: Fix posix_symlinks_match()Pranith Kumar K2018-07-022-8/+23
| | | | | | | | | | | 1) snprintf into linkname_expected should happen with PATH_MAX 2) comparison should happen with linkname_actual with complete string linkname_expected fixes bz#1595524 Change-Id: Ic3b3c362dc6c69c046b9a13e031989be47ecff14 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> (cherry picked from commit 3099d3e6ba81d3e1abf37385b13aabf5837b9c5e)
* glusterfsd: Do not process GLUSTERD_BRICK_XLATOR_OP if graph is not readyRavishankar N2018-07-022-1/+9
| | | | | | | | | | | | | | | | | | | | | Problem: If glustershd gets restarted by glusterd due to node reboot/volume start force/ or any thing that changes shd graph (add/remove brick), and index heal is launched via CLI, there can be a chance that shd receives this IPC before the graph is fully active. Thus when it accesses glusterfsd_ctx->active, it crashes. Fix: Since glusterd does not really wait for the daemons it spawned to be fully initialized and can send the request as soon as rpc initialization has succeeded, we just handle it at shd. If glusterfs_graph_activate() is not yet done in shd but glusterd sends GD_OP_HEAL_VOLUME to shd, we fail the request. Change-Id: If6cc07bc5455c4ba03458a36c28b63664496b17d fixes: bz#1597229 Signed-off-by: Ravishankar N <ravishankar@redhat.com> (cherry picked from commit 2f9e555f79ce69668950693ed2e09cfafd2b7ec1)
* doc: Added release notes for 4.1.1 releasev4.1.1ShyamsundarR2018-06-251-0/+26
| | | | | | | fixes: bz#1590442 Change-Id: I4144e79314a1ee927751946f26e7d79d337c819c Signed-off-by: ShyamsundarR <srangana@redhat.com>
* glusterfs: access trusted peer group via remote-host commandMohit Agrawal2018-06-251-5/+0
| | | | | | | | | | | | | | | | Problem: In SSL environment the user is able to access volume via remote-host command without adding node in a trusted pool Solution: Change the list of rpc program in glusterd.c at the time of initialization while SSL is enabled > Change-Id: I987e433b639e68ad17b77b6452df1e22dbe0f199 > cherry picked from commit 234d611160840899bcfd5ab1c17a6253673d38ed BUG: 1593525 fixes: bz#1593525 Change-Id: Ice4eda3d8104a4d5641de3cffd7249e46080d48f Signed-off-by: Mohit Agrawal <moagrawa@redhat.com>
* gcron: create the lockfile if it is missingNiels de Vos2018-06-201-1/+1
| | | | | | | | | | The lockfile for the job may not exist yet. If that is the case, it should be created upon the first time it is accessed. Change-Id: I4da2b3ecdb79cc63ed82cc7bfa026c8f08d4d043 Fixes: bz#1590195 Signed-off-by: Niels de Vos <ndevos@redhat.com> (cherry picked from commit 7005b1a336e483ec150c2f924a618dcfe197db0b)
* gcron: catch OSError as well as IOErrorNiels de Vos2018-06-181-2/+2
| | | | | | | | | | | In case os.open() fails because the file does not exist, an OSError is raised. To prevent the script to abort uncleanly, catch the OSError in addition to the IOError. Change-Id: I48e5b23e17d63639cc33db51b4229249a9887880 Fixes: bz#1590195 Signed-off-by: Niels de Vos <ndevos@redhat.com> (cherry picked from commit 26b52694feb04c98e6c9436bcd4e23e1687f0237)
* storage/posix: Handle ENOSPC correctly in zero_fillPranith Kumar K2018-06-143-1/+121
| | | | | | | Change-Id: Icc521d86cc510f88b67d334b346095713899087a fixes: bz#1591185 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> (cherry picked from commit 6ef91480f9e75f63100585bfd19694deb0c2457b)
* doc: Updated final release notes for 4.1.0v4.1.0ShyamsundarR2018-06-121-9/+167
| | | | | | | Fixes: bz#1575386 Change-Id: Iad79e16d76a55a2c3fc46629f79fec677bad597f Signed-off-by: ShyamsundarR <srangana@redhat.com>
* protocol/server: Fix xdata leak in seek fopPranith Kumar K2018-06-121-2/+1
| | | | | | | Change-Id: I6125283ed22c04564f0b77bb7a50579a83e02eb0 fixes: bz#1590128 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> (cherry picked from commit fd5b48ea0afd907deb08604415bee14ab65f378b)
* posix/ctime: Fix fops racing in updating mtime/atimeKotresh HR2018-06-081-11/+31
| | | | | | | | | | | | | | | | | | | | | | | | In distributed systems, there could be races with fops updating mtime/atime which could result in different mtime/atime for same file. So updating them only if time is greater than the existing makes sure, only the highest time is retained. If the mtime/atime update comes from the explicit utime syscall, it is allowed to set to previous time. Thanks Xavi for helping in rooting the issue. Backport of: > Patch: https://review.gluster.org/#/c/20120/ > BUG: 1584981 > Change-Id: If1230a75b96d7f9a828795189fcc699049e7826e > Signed-off-by: Kotresh HR <khiremat@redhat.com> (cherry picked from commit a6f0e7a4f1ca203762cae2ed5e426b52124c74dc) fixes: bz#1585894 Change-Id: If1230a75b96d7f9a828795189fcc699049e7826e Signed-off-by: Kotresh HR <khiremat@redhat.com>
* features/shard: Fix missing unlock in shard_fsync_shards_cbk()Vijay Bellur2018-06-061-0/+1
| | | | | | | | | | | | | Backport of: > Change-Id: I745a98e957cf3c6ba69247fcf6b58dd05cf59c3c > Signed-off-by: Vijay Bellur <vbellur@redhat.com> > (cherry picked from commit b21f742f96d46b4adfa87281dd9a2e48fea8d031) > BUG: 789278 Change-Id: I745a98e957cf3c6ba69247fcf6b58dd05cf59c3c fixes: bz#1587908 Signed-off-by: Vijay Bellur <vbellur@redhat.com> (cherry picked from commit b21f742f96d46b4adfa87281dd9a2e48fea8d031)
* protocol/client: Don't send fops till SETVOLUME is completev4.1.0rc0Raghavendra G2018-05-312-3/+9
| | | | | | | | | | | | | | | | | | An earlier commit set conf->connected just after rpc layer sends RPC_CLNT_CONNECT event. However, success of socket level connection connection doesn't indicate brick stack is ready to receive fops, as an handshake has to be done b/w client and server after RPC_CLNT_CONNECT event. Any fop sent to brick in the window between, * protocol/client receiving RPC_CLNT_CONNECT event * protocol/client receiving a successful setvolume response can end up accessing an uninitialized brick stack. So, set conf->connected only after a successful SETVOLUME. Change-Id: I139a03d2da6b0d95a0d68391fcf54b00e749decf fixes: bz#1584633 Signed-off-by: Raghavendra G <rgowdapp@redhat.com> (cherry picked from commit 430e4872db43447c8a15b638760d23c188257139)
generated by cgit (git 2.34.1) at 2024-11-12 23:19:40 +0000