summaryrefslogtreecommitdiffstats
path: root/xlators/cluster/ec/src/ec-common.c
Commit message (Collapse)AuthorAgeFilesLines
...
* cluster/ec: Fix invalid __fd_unref() callXavier Hernandez2016-06-131-4/+1
| | | | | | | | | | | | | | | | | | | | | | | | | __fd_unref() doesn't do any cleanup, so it cannot be called to release fd references, specially if it's the last reference. The code has been changed to avoid a call to this function. In the previous version we always tried to keep the newest fd in the ec_lock_t structure. However this is not necessary. We'll always keep one reference to an open file on the same inode. It's irrelevant if the reference is new or old. The function __fd_unref() has also been removed from fd.h to avoid being used in the future since it's useless as it's defined now. Change-Id: Ia728777fc8e464758d5ea4d3bf020f0603919039 BUG: 1344396 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/14683 NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Smoke: Gluster Build System <jenkins@build.gluster.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Ashish Pandey <aspandey@redhat.com> Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
* core: assorted typos and spelling mistakes reported by Debian lintianKaleb S KEITHLEY2016-05-181-1/+1
| | | | | | | | | | | | | | | Also missing bang (!) in #!/bin/bash in shell scripts. Change-Id: I567a4be8f0f31f6285550f243fe802895f6bc43b BUG: 1336793 Reported-by: Patrick Matthäi <pmatthaei@debian.org> Signed-off-by: Kaleb S KEITHLEY <kkeithle@redhat.com> Reviewed-on: http://review.gluster.org/14398 NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Niels de Vos <ndevos@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kotresh HR <khiremat@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
* cluster/ec: Fix issues with eager lockingXavier Hernandez2016-05-021-68/+174
| | | | | | | | | | | | | | | | | | Due to a race in timer cancellation, in some cases it was possible to unlock the lock while another concurrent fop that needed it continues execution as if it were not released. This patch also fixes an issue that caused a lock to not be released if an error was found while preparing ec_update_size_version(). Change-Id: I1344a3f5ecfc333f05a09e62653838264c9c26b1 BUG: 1331254 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/14112 Smoke: Gluster Build System <jenkins@build.gluster.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Chen Chen <chenchen@smartquerier.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
* cluster/ec: Provide an option to enable/disable eager lockAshish Pandey2016-03-151-3/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: If a fop takes lock, and completes its operation, it waits for 1 second before releasing the lock. However, If ec find any lock contention within this time period, it release the lock immediately before time expires. As we take lock on first brick, for few operations, like read, it might happen that discovery of lock contention might take long time and can degrades the performance. Solution: Provide an option to enable/disable eager lock. If eager lock is disabled, lock will be released as soon as fop completes. gluster v set <VOLUME NAME> disperse.eager-lock on gluster v set <VOLUME NAME> disperse.eager-lock off Change-Id: I000985a787eba3c190fdcd5981dfbf04e64af166 BUG: 1314649 Signed-off-by: Ashish Pandey <aspandey@redhat.com> Reviewed-on: http://review.gluster.org/13605 Smoke: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
* cluster/ec: Fix invalid config check for directoriesXavier Hernandez2016-02-291-1/+2
| | | | | | | | | | | | | | | | | | | | | | The trusted.ec.config xattr is not defined for directories. However sometimes it could be requested because the inode type of a directory can temporarily be IA_INVAL. Requesting such xattr using the xattrop fop when it doesn't exist, returns a config value full of 0's, which is invalid and caused some fops to fail. This patch filters out this case by ignoring config xattr == 0. Change-Id: Ied51c35b313ea8c3eeae27812f9bae61d3808e92 BUG: 1293223 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/13446 Smoke: Gluster Build System <jenkins@build.gluster.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Ashish Pandey <aspandey@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
* cluster/ec: Handle non-existent config xattr for non regular filesXavier Hernandez2016-02-051-23/+25
| | | | | | | | | | | | | | | | | | | | Since we now try to get the 'trusted.ec.config' xattr for inodes of type IA_INVAL (these inodes will be set to some valid type later), if that inode corresponds to a non regular file, the xattr won't exist and we will handle this as an error when it's not. This patch solves the problem by only considering errors for inodes that are already known to be regular files. Change-Id: Id72f314e209459236d75cf087fc51e09943756b4 BUG: 1293223 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/13238 Smoke: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Ashish Pandey <aspandey@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
* cluster/ec: Get size and config for invalid inodeAshish Pandey2016-01-131-11/+20
| | | | | | | | | | | | | | | | | | | Problem: After creating an inode and before linking it to inode table, if there is a request to setattr for that file, it fails and leads to crash. Before linking inode to inode table ia_type is IA_INVAL which will casue have_size and have_config as zero. Solution: Check and get size and config if an inode is invalid Change-Id: I0c0e564940b1b9f351369a76ab14f6b4aa81f23b BUG: 1293223 Signed-off-by: Ashish Pandey <aspandey@redhat.com> Reviewed-on: http://review.gluster.org/13039 Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: NetBSD Build System <jenkins@build.gluster.org>
* cluster/ec: Mark internal fops appropriatelyXavier Hernandez2015-11-191-14/+4
| | | | | | | | | | | | | 1) Mark read fops in read-modify-write by EC as internal. 2) Handle uid/gid set/reset correctly BUG: 1282761 Change-Id: I5c1ce0cd6213367eaead5fed33aa2397c4e46df7 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/12599 Tested-by: Gluster Build System <jenkins@build.gluster.com> Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: fix bug in update_goodPranith Kumar K2015-11-111-5/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: Bricks that didn't participate in the fops are considered to be good. This is happening two fold. Examples: Case-1: 1) 2+1 volume. 'd1' directory on Brick-0 is bad. 2) readdir takes locks and lock->good_mask is '7' 3) readdir does xattrop and fop->mask is '6'. 4) because fop->expected is '1' lock->good_mask remains '7' Case-2: 1) when all the bricks are up, it does lock + xattrop before op and figures out all the bricks are good. 2) By the time second operation starts brick-0 is down. Now lock->good_mask will always have the '0' bit set as long as the operations are happening on it. because: "lock->good_mask &= ~fop->mask | fop->remaining" fop->mask doesn't have '0' th bit. 3) When it comes time to perform the final xattrop in update_size_version brick-0 comes online because of which it gives the same version to brick-0 as well thinking it has participated in all the transactions till then, even when it didn't participate in the transactions. Fix: Case-1's fix: Update lock->good_mask in ec_prepare_update_cbk with latest good/bad bricks Case-2's fix: Consider non-participating brick as bad. Change-Id: Ic01a733f8180131ded6a3cc784fcb1960758cf23 BUG: 1276989 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/12561 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Fix bad management of lock ownersXavier Hernandez2015-11-051-8/+8
| | | | | | | | | | | | | | | | | | | | Since the addition of parallel reads patch for ec, a lock can have more than one owner at the same time. The list of owners was stored inside the 'owner_list' field of each fop. The problem was with fops that required more than one lock (like rename). In this case the same field was used to add the fop to more than one list, casing an overwrite of the previous list. This has been solved moving the 'owner_list' field from ec_fop_data_t to ec_lock_link_t structure. Change-Id: I6042129f09082497b80782b5704a52c35c78f44d BUG: 1276031 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/12445 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
* cluster/ec: update version and size on good bricksAshish Pandey2015-10-281-10/+2
| | | | | | | | | | | | | | | | | | | | Problem: readdir/readdirp fops calls [f]xattrop with fop->good which contain only one brick for these operations. That causes xattrop to be failed as it requires at least "minimum" number of brick. Solution: Use lock->good_mask to call xattrop. lock->good_mask contain all the good locked bricks on which the previous write opearion was successfull. Change-Id: If1b500391aa6fca6bd863702e030957b694ab499 BUG: 1274629 Signed-off-by: Ashish Pandey <aspandey@redhat.com> Reviewed-on: http://review.gluster.org/12419 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Implement gfid-hash read-policyPranith Kumar K2015-10-091-9/+28
| | | | | | | | | | | | | | Add a policy in ec to performs reads from same bricks as long as they are good. Based on the gfid of the file/directory it determines the bricks to be considered for reading. Change-Id: Ic97b5c54c086a28b5e07a330a4fd448551b49376 BUG: 1261260 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/12133 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Allow read fops to be processed in parallelXavier Hernandez2015-08-291-160/+311
| | | | | | | | | | | | | | Currently ec only sends a single read request at a time for a given inode. Since reads do not interfere between them, this patch allows multiple concurrent read requests to be sent in parallel. Change-Id: If853430482a71767823f39ea70ff89797019d46b BUG: 1245689 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/11742 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* cluster/ec: Fix tracking of good bricksXavier Hernandez2015-08-061-146/+49
| | | | | | | | | | | | | | | | | | | The bitmask of good and bad bricks was kept in the context of the corresponding inode or fd. This was problematic when an external process (another client or the self-heal process) did heal the bricks but no one changed the bitmaks of other clients. This patch removes the bitmask stored in the context and calculates which bricks are healthy after locking them and doing the initial xattrop. After that, it's updated using the result of each fop. Change-Id: I225e31cd219a12af4ca58871d8a4bb6f742b223c BUG: 1236065 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/11844 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Minimize usage of EIO errorXavier Hernandez2015-07-281-50/+127
| | | | | | | | | | Change-Id: I82e245615419c2006a2d1b5e94ff0908d2f5e891 BUG: 1245276 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/11741 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: NetBSD Build System <jenkins@build.gluster.org>
* cluster/ec: Handle race between unlock-timer, new lockPranith Kumar K2015-07-231-19/+14
| | | | | | | | | | | | | | | | | | | | | | | | Problem: New lock could come at the time timer is on the way to unlock. This was leading to crash in timer thread because thread executing new lock can free up the timer_link->fop and then timer thread will try to access structures already freed. Fix: If the timer event is fired, set lock->release to true and wait for unlock to complete. Thanks to Xavi and Bhaskar for helping in confirming that this race is the RC. Thanks to Kritika for pointing out and explaining how Avati's patch can be used to fix this bug. Change-Id: I45fa5470bbc1f03b5f3d133e26d1e0ab24303378 BUG: 1243187 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11670 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: NetBSD Build System <jenkins@build.gluster.org>
* cluster/ec: wind readlink on good subvol(s)Pranith Kumar K2015-07-141-6/+10
| | | | | | | | | | BUG: 1232172 Change-Id: I3a56e487840d86147dd85bf5fbe79b165eae289f Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11589 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* cluster/ec: Prevent data corruptionsPranith Kumar K2015-07-141-3/+20
| | | | | | | | | | | | | | - On lock reuse preserve 'healing' bits - Don't set ctx->size outside locks in healing code - Allow xattrop internal fops also on the fop->mask. Change-Id: I6b76da5d7ebe367d8f3552cbf9fd18e556f2a171 BUG: 1232678 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11640 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Don't read from bricks that are healingPranith Kumar K2015-07-091-1/+1
| | | | | | | | | | BUG: 1232678 Change-Id: I35503039e4723cf7f33d6797f0ba90dd0aca130b Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11580 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Fix use after free bugPranith Kumar K2015-07-071-0/+8
| | | | | | | | | | | | | | | In ec_lock() there is a chance that ec_resume is called on fop even before ec_sleep. This can result in refs == 0 for fop leading to use after free in this function when it calls ec_sleep so do ec_sleep at start and ec_resume at end of this function. Change-Id: I879b2667bf71eaa56be1b53b5bdc91b7bb56c650 BUG: 1240284 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11558 Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: Gluster Build System <jenkins@build.gluster.com> Tested-by: NetBSD Build System <jenkins@build.gluster.org>
* cluster/ec: Don't read from bad subvolsPranith Kumar K2015-07-061-18/+23
| | | | | | | | | | Change-Id: Ic22813371faca4e8198c9b0b20518e68d275f3c1 BUG: 1232678 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11531 Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: NetBSD Build System <jenkins@build.gluster.org> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* ec: Porting messages to new logging frameworkNandaja Varma2015-06-261-38/+65
| | | | | | | | | Change-Id: Ia05ae750a245a37d48978e5f37b52f4fb0507a8c BUG: 1194640 Signed-off-by: Nandaja Varma <nandaja.varma@gmail.com> Reviewed-on: http://review.gluster.org/10465 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: wind fops on good subvols for access/readdir[p]Pranith Kumar K2015-06-261-6/+22
| | | | | | | | | Change-Id: I1e629a6adc803c4b7164a5a7a81ee5cb1d0e139c BUG: 1232172 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11246 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Avoid parallel executions of the same state machineXavier Hernandez2015-06-181-11/+13
| | | | | | | | | | | | | | In very rare circumstances it was possible that a subfop started by another fop could finish fast enough to cause that two or more instances of the same state machine be executing at the same time. Change-Id: I319924a18bd3f88115e751a66f8f4560435e0e0e BUG: 1233258 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/11317 Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Prevent Null dereference in dht-renamePranith Kumar K2015-06-121-1/+1
| | | | | | | | | | Change-Id: I3059f3b577f550c92fb77c6b6b44defd0584cd2e BUG: 1230647 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11178 Tested-by: Gluster Build System <jenkins@build.gluster.com> Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* cluster/ec: Wind unlock fops at all costPranith Kumar K2015-06-101-3/+24
| | | | | | | | | | | | | | | | | Problem: While files are being created if more than redundancy number of bricks go down, then unlock for these fops do not go to the bricks. This will lead to stale locks leading to hangs. Fix: Wind unlock fops at all costs. Change-Id: I50a87e8b4d6d2dde5bf7405b82e3aeecd95ad00e BUG: 1220348 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11152 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Prevent double unwindPranith Kumar K2015-06-081-4/+2
| | | | | | | | | | | | | | | | | | | | | | Problem: 1) ec_access/ec_readlink_/ec_readdir[p] _cbks are trying to recover only from ENOTCONN. 2) When the fop succeeds it unwinds right away. But when its ec_fop_manager resumes, if the number of bricks that are up is less than ec->fragments, the the state machine will resume with -EC_STATE_REPORT which unwinds again. This will lead to crashes. Fix: - If fop fails retry on other subvols, as ESTALE/ENOENT/EBADFD etc are also recoverable. - unwind success/failure in _cbks Change-Id: I2cac3c2f9669a4e6160f1ff4abc39f0299303222 BUG: 1228952 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11111 Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* Changing log level to DEBUG in case of ENOENTAshish Pandey2015-06-081-2/+3
| | | | | | | | | Change-Id: I264e47ca679d8b57cd8c80306c07514e826f92d8 BUG: 1193388 Signed-off-by: Ashish Pandey <aspandey@redhat.com> Reviewed-on: http://review.gluster.org/10784 Tested-by: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* cluster/ec: Don't handle EC_XATTR_DIRTY in responsePranith Kumar K2015-06-061-24/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | Problem: ec_update_size_version expects all the keys it did xattrop with to come in response so that it can set the values again in ec_update_size_version_done. But EC_XATTR_DIRTY is not combined so the value won't be present in the response. So ctx->post/pre_dirty are not updated in ec_update_size_version_done. So these values are still non-zero. When ec_unlock_now is called as part of flush's unlock phase it again tries to perform same xattrop for EC_XATTR_DIRTY. But ec_update_size_version is not expected to be called in unlock phase of flush because ec_flush_size_version should have reset everything to zero and unlock is never invoked from ec_update_size_version_done for flush/fsync/fsyncdir. This leads to stale lock which leads to hang. Fix: EC_XATTR_DIRTY is removed in ex_xattrop_cbk and is never combined with other answers. So remove handling of this in the response. Change-Id: If0ea3efec3235a6e312465d8838585fbe752c7ea BUG: 1227654 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11078 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* cluster/ec: Forced unlock when lock contention is detectedXavier Hernandez2015-05-271-685/+786
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | EC uses an eager lock mechanism to optimize multiple read/write requests on the same entry or inode. This increases performance but can have adverse results when other clients try to access the same entry/inode. To solve this, this patch adds a functionality to detect when this happens and force an earlier release to not block other clients. The method consists on requesting GF_GLUSTERFS_INODELK_COUNT and GF_GLUSTERFS_ENTRYLK_COUNT for all fops that take a lock. When this count is greater than one, the lock is marked to be released. All fops already waiting for this lock will be executed normally before releasing the lock, but new requests that also require it will be blocked and restarted after the lock has been released and reacquired again. Another problem was that some operations did correctly lock the parent of an entry when needed, but got the size and version xattrs from the entry instead of the parent. This patch solves this problem by binding all queries of size and version to each lock and replacing all entrylk calls by inodelk ones to remove concurrent updates on directory metadata. This also allows rename to correctly update source and destination directories. Change-Id: I2df0b22bc6f407d49f3cbf0733b0720015bacfbd BUG: 1165041 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/10852 Tested-by: NetBSD Build System Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Fix use after free crashPranith Kumar K2015-05-211-14/+0
| | | | | | | | | | | | | | | | | ec_heal creates ec_fop_data but doesn't run ec_manager. ec_fop_data_allocate adds this fop to ec->pending_fops, because ec_manager is not run on this heal fop it is never removed from ec->pending_fops. When it is accessed after free it leads to crash. It is better to not to add HEAL fops to ec->pending_fops because we don't want graph switch to hang the mount because of a BIG file/directory heal. BUG: 1188145 Change-Id: I8abdc92f06e0563192300ca4abca3909efcca9c3 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/10868 Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Raghavendra Bhat <raghavendra@redhat.com>
* cluster/ec: Correctly cleanup delayed locksXavier Hernandez2015-05-201-17/+57
| | | | | | | | | | | | | | | | | When a delayed lock is pending, a graph switch doesn't correctly terminate it. This means that the update of version and size xattrs is lost, causing EIO errors. This patch handles GF_EVENT_PARENT_DOWN event to correctly finish pending udpdates before completing the graph switch. Change-Id: I394f3b8d41df8d83cdd36636aeb62330f30a66d5 BUG: 1188145 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/10787 Tested-by: NetBSD Build System Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Handle lookup failures while op in progressPranith Kumar K2015-05-191-7/+9
| | | | | | | | | | Change-Id: Ia1834ec23d5de615526d4d4e4d2e32aff155b7f7 BUG: 1211962 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/10806 Tested-by: NetBSD Build System Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
* ec: Fix failures with missing filesXavier Hernandez2015-05-091-19/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | When a file does not exist on a brick but it does on others, there could be problems trying to access it because there was some loc_t structures with null 'pargfid' but 'name' was set. This forced inode resolution based on <pargfid>/name instead of <gfid> which would be the correct one. To solve this problem, 'name' is always set to NULL when 'pargfid' is not present. Another problem was caused by an incorrect management of errors while doing incremental locking. The only allowed error during an incremental locking was ENOTCONN, but missing files on a brick can be returned as ESTALE. This caused an EIO on the operation. This patch doesn't care of errors during an incremental locking. At the end of the operation it will check if there are enough successfully locked bricks to continue or not. Change-Id: I9360ebf8d819d219cea2d173c09bd37679a6f15a BUG: 1176062 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9407 Tested-by: NetBSD Build System Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Change meaning of trusted.ec.dirtyPranith Kumar K2015-05-081-14/+88
| | | | | | | | | | | | | | | | - With this change, the xattr will represent if the file needs to be healed or not. It will have different values for data/entry and metadata changes. - inode ref leaks and dict_set_dynstr related leaks fixed - Added support for trylock/lock based on heal-cmd execution or not in data heal. - Made fixes to pass regression runs Change-Id: I9d8def4c2badde18a76b7898816fecfac113737a BUG: 1215265 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/10385 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* cluster/ec: add separate versions for data/entry, metadataAshish Pandey2015-05-061-16/+31
| | | | | | | | | | | | | | | | | | | Adding 64 bits in "version" key of extended attributes. First 64 bits (Left) represents Data version. Last 64 bits (right) represents Meta Data version. Note: 3.7 and 3.6 version ec can't co-exist with this change because xattrop in 3.6 will fail with ERANGE as the buffer passed to it will be '8' bytes where as the value will be 16 bytes in 3.7. Where as 3.7 version clients can work with old version files. For upgrades we need to tell users to complete heals and then upgrade BUG: 1215265 Change-Id: Ib85114680cb7e75b8371c984d9f7b6401c1ffb93 Signed-off-by: Ashish Pandey <aspandey@redhat.com> Reviewed-on: http://review.gluster.org/10312 Tested-by: Gluster Build System <jenkins@build.gluster.com> Tested-by: NetBSD Build System Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/ec: Use fd instead of loc for get_size_versionAshish Pandey2015-05-041-51/+57
| | | | | | | | | Change-Id: Ia7d43cb3b222db34ecb0e35424f1766715ed8e6a BUG: 1188242 Signed-off-by: Ashish Pandey <aspandey@redhat.com> Reviewed-on: http://review.gluster.org/10176 Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* cluster/ec: Perform inode-write on healing subvolsPranith Kumar K2015-04-241-16/+44
| | | | | | | | | | | | | | | | | | If the version numbers do not match, then writes are performed only on at least N-R bricks which have same version. But if we want to do healing of files which are constantly modified we need to allow writes on subvols that are undergoing heal. Data healing will mark 62nd bit while the heal is going on. When the data transaction sees that this bit is set it needs to perform the fop on that subvol irrespective of whether the versions match or do not match. Fop is considered successful only if N-R non-healing bricks succeed. Change-Id: I69a17582df397aaf6e8ca4b5e746c7ca802cbbde BUG: 1215265 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/10372 Tested-by: NetBSD Build System Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* Avoid conflict between contrib/uuid and system uuidEmmanuel Dreyfus2015-04-041-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | glusterfs relies on Linux uuid implementation, which API is incompatible with most other systems's uuid. As a result, libglusterfs has to embed contrib/uuid, which is the Linux implementation, on non Linux systems. This implementation is incompatible with systtem's built in, but the symbols have the same names. Usually this is not a problem because when we link with -lglusterfs, libc's symbols are trumped. However there is a problem when a program not linked with -lglusterfs will dlopen() glusterfs component. In such a case, libc's uuid implementation is already loaded in the calling program, and it will be used instead of libglusterfs's implementation, causing crashes. A possible workaround is to use pre-load libglusterfs in the calling program (using LD_PRELOAD on NetBSD for instance), but such a mechanism is not portable, nor is it flexible. A much better approach is to rename libglusterfs's uuid_* functions to gf_uuid_* to avoid any possible conflict. This is what this change attempts. BUG: 1206587 Change-Id: I9ccd3e13afed1c7fc18508e92c7beb0f5d49f31a Signed-off-by: Emmanuel Dreyfus <manu@netbsd.org> Reviewed-on: http://review.gluster.org/10017 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Niels de Vos <ndevos@redhat.com>
* cluster/ec: Use fd when appropriate for updating size/versionPranith Kumar K2015-03-271-3/+9
| | | | | | | | | | Change-Id: I5d3aca101c8cdda406d31d06c40404fa6a2b7170 BUG: 1192378 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/9995 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kaleb KEITHLEY <kkeithle@redhat.com> Reviewed-by: Dan Lambright <dlambrig@redhat.com>
* ec: Add trusted.ec.dirty xattrXavier Hernandez2015-02-231-74/+209
| | | | | | | | | | | | | | | | | This xattr will be incremented before each data modifying operation and decremented after it. This will add the possibility to detect partially updated writes and refuse them on reads. It will also be useful for interacting with index xlator and have a way to heal dispersed files from the self-heal daemon. Change-Id: Ie644a8dd074ae0f254c809c5863bdb030be5486a BUG: 1190581 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9607 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* ec: Fix posix compliance failuresXavier Hernandez2015-01-281-23/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch solves some problems that caused dispersed volumes to not pass posix smoke tests: * Problems in open/create with O_WRONLY Opening files with -w- permissions using O_WRONLY returned an EACCES error because internally O_WRONLY was replaced with O_RDWR. * Problems with entrylk on renames. When source and destination were the same, ec tried to acquire the same entrylk twice, causing a deadlock. * Overwrite of a variable when reordering locks. On a rename, if the second lock needed to be placed at the beggining of the list, the 'lock' variable was overwritten and later its timer was cancelled, cancelling the incorrect one. * Handle O_TRUNC in open. When O_TRUNC was received in an open call, it was blindly propagated to child subvolumes. This caused a discrepancy between real file size and the size stored into trusted.ec.size xattr. This has been solved by removing O_TRUNC from open and later calling ftruncate. Change-Id: I20c3d6e1c11be314be86879be54b728e01013798 BUG: 1161886 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9420 Reviewed-by: Dan Lambright <dlambrig@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* ec: Fix return errors when not enough bricksXavier Hernandez2014-12-051-12/+24
| | | | | | | | | | | | | | | | | | | | | Changes introduced by this patch: * Fix an incorrect error propagation when the state of the life cycle of a fop returns an error. * Fix incorrect unlocking of failed locks. * Return ENOTCONN if there aren't enough bricks online. * In readdir(p) check that the fd has been successfully open by a previous opendir. Change-Id: Ib44f25a1297849ebcbab839332f3b6359f275ebe BUG: 1162805 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9098 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* ec: Fix mutex related coverity scan issuesXavier Hernandez2014-12-041-2/+14
| | | | | | | | | | | | | | | | | | This patch solves 3 issues detected by coverity scan: CID1241484 Data race condition CID1241486 Data race condition CID1256173 Thread deadlock With this patch, inode lock is never acquired inside a region locked with fop->lock. Change-Id: I35c4633efd1b68b9f72b42661fa7c728b1f52c6a BUG: 1170254 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9230 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Dan Lambright <dlambrig@redhat.com>
* ec: Fix self-healing issues.Xavier Hernandez2014-12-041-45/+51
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Three problems have been detected: 1. Self healing is executed in background, allowing the fop that detected the problem to continue without blocks nor delays. While this is quite interesting to avoid unnecessary delays, it can cause spurious failures of self-heal because it may try to recover a file inside a directory that a previous self-heal has not recovered yet, causing the file self-heal to fail. 2. When a partial self-heal is being executed on a directory, if a full self-heal is attempted, it won't be executed because another self-heal is already in process, so the directory won't be fully repaired. 3. Information contained in loc's of some fop's is not enough to do a complete self-heal. To solve these problems, I've made some changes: * Improved ec_loc_from_loc() to add all available information to a loc. * Before healing an entry, it's parent is checked and partially healed if necessary to avoid failures. * All heal requests received for the same inode while another self-heal is being processed are queued. When the first heal completes, all pending requests are answered using the results of the first heal (without full execution), unless the first heal was a partial heal. In this case all partial heals are answered, and the first full heal is processed normally. * An special virtual xattr (not physically stored on bricks) named 'trusted.ec.heal' has been created to allow synchronous self-heal of files. Now, the recommended way to heal an entire volume is this: find <mount> -d -exec getfattr -h -n trusted.ec.heal {} \; Some minor changes: * ec_loc_prepare() has been renamed to ec_loc_update(). * All loc management functions return 0 on success and -1 on error. * Do not delay fop unlocks if heal is needed. * Added basic ec xattrs initially on create, mkdir and mknod fops. * Some coding style changes Change-Id: I2a5fd9c57349a153710880d6ac4b1fa0c1475985 BUG: 1161588 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9072 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Dan Lambright <dlambrig@redhat.com>
* ec: Change licenseXavier Hernandez2014-12-031-16/+6
| | | | | | | | | | Change-Id: Iae90ade2421898417b53dec0417a610cf306c44b BUG: 1168167 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9201 Reviewed-by: Kaleb KEITHLEY <kkeithle@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* ec: Avoid self-heal on directories on (f)stat callsXavier Hernandez2014-11-151-1/+2
| | | | | | | | | | | | | | | | | | | To avoid inconsistent directory listings, a full self-heal cannot happen on a directory until all its contents have been healed. This is controlled by a manual command using getfattr recursively and in post-order. While navigating the directories, sometimes an (f)stat fop can be sent. This fop caused a full self-heal of the directory. This patch makes that (f)stat only initiates a partial self-heal. Change-Id: I0a92bda8f4f9e43c1acbceab2d7926944a8a4d9a BUG: 1163760 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/9117 Reviewed-by: Dan Lambright <dlambrig@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* ec: Fix rebalance issuesXavier Hernandez2014-10-271-110/+182
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some issues in ec xlator made that rebalance didn't complete successfully and generated some warnings and errors in the log. The most critical error was a race condition that caused false corruption detection when two specific operations were executed sequentially and they shared the same lock. This explains the problem: 1. A setxattr is issued. 2. setxattr: ec locks the inode before updating the xattr. 3. setxattr: The xattr is updated. 4. setxattr: Upper xlator is notified that the operation completed. 5. setxattr: A background task is initiated to update the version of the file. 6. A stat is issued on the same file. 7. stat: Since the lock is already acquired, it's reused. 8. stat: A lookup is issued to determine version and size information of the file. At this point, operations 5 and 8 can interfere. This can make that lookup sees different information on each brick, determining that some bricks are corrupted and incorrectly excluding them from the operation and initiating a self-heal. In some cases this false detection combined with self-heal could lead to invalid updates of the trusted.ec.size xattr, leaving the file smaller than it should be. This only happens if the first operation does not perform a lookup, because chained operations reuse the information returned by the previous one, avoiding this kind of problems. To solve this, now the background update is executed atomically with the posterior unlock. This avoids some reuses of the lock while updating. However this reduces performance because the window in which new requests can reuse the lock is much smaller now. This has been alleviated by using the same technique implemented in AFR (i.e. waiting some time before releasing the lock). Some minor changes also introduced in this patch: * Bug in management of 'trusted.glusterfs.pathinfo' that was writing beyond the allocated space. * Uninitialized variable. * trusted.ec.config was not created for regular files created with mknod. * An invalid state was used in access fop. Change-Id: Idfaf69578ed04dbac97a62710326729715b9b395 BUG: 1152902 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/8947 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* ec: Fix self-heal issuesXavier Hernandez2014-10-211-140/+131
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: Doing an 'ls' of a directory that has been modified while one of the bricks was down, sometimes returns the old directory contents. Cause: Directories are not marked when they are modified as files are. The ec xlator balances requests amongst available and healthy bricks. Since there is no way to detect that a directory is out of date in one of the bricks, it is used from time to time to return the directory contents. Solution: Basically the solution consists in use versioning information also for directories, however some additional changes have been necessary. Changes: * Use directory versioning: This required to lock full directory instead of a single entry for all requests that add or remove entries from it. This is needed to allow atomic version update. This affects the following fops: create, mkdir, mknod, link, symlink, rename, unlink, rmdir Another side effect is that opendir requires to do a previous lookup to get versioning information and discard out of date bricks for subsequent readdir(p) calls. * Restrict directory self-heal: Till now, when one discrepancy was found in lookup, a self-heal was automatically started. This caused the versioning information of a bad directory to be healed instantly, making the original problem to reapear again. To solve this, when a missing directory is detected in one or more bricks on lookup or opendir fops, only a partial self-heal is performed on it. A partial self-heal basically creates the directory but does not restore any additional information. This avoids that an 'ls' could repair the directory and cause the problem to happen again. With this change, output of 'ls' is always consistent. However, since the directory has been created in the brick, this allows any other operation on it (create new files, for example) to succeed on all bricks and not add additional work to the self-heal process. To force a self-heal of a directory, any other operation must be done on it. For example a getxattr. With these changes, the correct healing procedure that would avoid inconsistent directory browsing consists on a post-order traversal of directoriesi being healed. This way, the directory contents will be healed before healing the directory itslef. * Additional changes to fix self-heal errors - Don't use fop->fd to decide between fd/loc. open, opendir and create have an fd, but the correct data is in loc. - Fix incorrect management of bad bricks per inode/fd. - Fix incorrect selection of fop's target bricks when there are bad bricks involved. - Improved ec_loc_parent() to always return a parent loc as complete as possible. Change-Id: Iaf3df174d7857da57d4a87b4a8740a7048b366ad BUG: 1149726 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/8916 Reviewed-by: Dan Lambright <dlambrig@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* ec: Fix 32 bits issues on file size calculationXavier Hernandez2014-10-201-4/+2
| | | | | | | | | | | | | | Some additional 32 bits issues have been added by a recent patch. This patch solves it. Change-Id: Ice81032fbe8e36e5ccad19a781b7876891993906 BUG: 1146903 Signed-off-by: Xavier Hernandez <xhernandez@datalab.es> Reviewed-on: http://review.gluster.org/8882 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Emmanuel Dreyfus <manu@netbsd.org> Tested-by: Emmanuel Dreyfus <manu@netbsd.org> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
generated by cgit (git 2.34.1) at 2024-05-28 07:08:36 +0000