summaryrefslogtreecommitdiffstats
path: root/xlators/cluster/afr/src/afr-common.c
Commit message (Collapse)AuthorAgeFilesLines
* cluster/afr: Honor default timeout of 5min for analyzing split-brain fileskarthik-us2017-11-271-1/+5
| | | | | | | | | | | | | | | | | | | | Problem: After setting split-brain-choice option to analyze the file to resolve the split brain using the command "setfattr -n replica.split-brain-choice -v "choiceX" <path-to-file>" should allow to access the file from mount for default timeout of 5mins. But the timeout was not honored and was able to access the file even after the timeout. Fix: Call the inode_invalidate() in afr_set_split_brain_choice_cbk() so that it will triger the cache invalidate after resetting the timer and the split brain choice. So the next calls to access the file will fail with EIO. Change-Id: I698cb833676b22ff3e4c6daf8b883a0958f51a64 BUG: 1514388 Signed-off-by: karthik-us <ksubrahm@redhat.com> (cherry picked from commit 933ec57ccda2c1ba5ce6f207313c3b6802e67ca3)
* afr: heal metadata in discover code pathRavishankar N2017-09-171-17/+42
| | | | | | | | | | | | | | | | | | | | | | | | | | | ****************************************************** Backport of: https://review.gluster.org/18202 Also added loc_is_nameless() to libglusterfs since the patch that introduced it in master was not backported to release-3.10. Note: 18202 is a squash of 17850 and 18187 in master. ****************************************************** During graph switch, if fuse sends nameless (gfid) lookups, afr takes the discover code path to serve it. If there are pending metadata heals, they do not happen unless an inode refresh happens as a part of discover (which is not guaranteed to happen always). This patch fixes it by attempting metadata heal as a part of discover, just like how it is done in lookup code path. Change-Id: I87c493045b9225741cad173bf3f645848697032e BUG: 1492010 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: https://review.gluster.org/18304 Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Shyamsundar Ranganathan <srangana@redhat.com>
* afr: check validity of afr_replyRavishankar N2017-09-171-9/+14
| | | | | | | | | | | | | | | | | | | | | | ...in various self-heal code paths. Originally found by Pranith in __afr_selfheal_name_impunge () Also change __afr_selfheal_assign_gfid() to send lookup only on those bricks that don't have a gfid matching that of the source. > Reviewed-on: https://review.gluster.org/18065 > Smoke: Gluster Build System <jenkins@build.gluster.org> > CentOS-regression: Gluster Build System <jenkins@build.gluster.org> (cherry picked from commit d594900dbca92c356152be65fce16f77c402117c) Change-Id: I70a2ccd750a2af92c5fc36e0eefb2b6125404b4a BUG: 1491995 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: https://review.gluster.org/18303 Smoke: Gluster Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Shyamsundar Ranganathan <srangana@redhat.com>
* afr: add errno to afr_inode_refresh_done()Ravishankar N2017-06-161-7/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Backport of https://review.gluster.org/17413 and https://review.gluster.org/17436 Problem: When parellel `rm -rf`s were being done from cifs clients, opendir might fail on some replicas with ENOENT. DHT ignores partial opendir failures in dht_fd_cbk() and winds readdirs on those replicas. Afr inode refresh (as a part of readdirp read_txn) sees in its fd context that the state of the fds is *not* AFR_FD_OPENED and bails out to afr_inode_refresh_done() without doing a refresh. When this happens, the errno is set as EIO due to lack of readable subvols, logging split-brain messages in the logs. Fix: Introduce an errno argument to afr_inode_refresh_do() to bail out with the right error value when inode refresh is not performed. Change-Id: Id88e604278abb8df47750d45258d9e6dde710600 BUG: 1457732 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: https://review.gluster.org/17516 Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/afr: Return the list of node_uuids for the subvolumekarthik-us2017-05-301-0/+49
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: AFR was returning the node uuid of the first node for every file if the replica set was healthy, which was resulting in only one node migrating all the files. Fix: With this patch AFR returns the list of node_uuids to the upper layer, so that they can decide on which node to migrate which files, resulting in improved performance. Ordering of node uuids will be maintained based on the ordering of the bricks. If a brick is down, then the node uuid for that will be set to all zeros. > Reviewed-on: https://review.gluster.org/17084 > Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> > Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> > Smoke: Gluster Build System <jenkins@build.gluster.org> > NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> > CentOS-regression: Gluster Build System <jenkins@build.gluster.org> (cherry picked from commit 0a50167c0a8f950f5a1c76442b6c9abea466200d) Change-Id: I73ee0f9898ae473584fdf487a2980d7a6db22f31 BUG: 1451561 Signed-off-by: karthik-us <ksubrahm@redhat.com> Reviewed-on: https://review.gluster.org/17337 Tested-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Raghavendra Talur <rtalur@redhat.com>
* afr: send the correct iatt values in fsync cbkRavishankar N2017-05-141-25/+43
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: afr unwinds the fsync fop with an iatt buffer from one of its children on whom fsync was successful. But that child might not be a valid read subvolume for that inode because of pending heals or because it happens to be the arbiter brick etc. Thus we end up sending the wrong iatt to mdcache which will in turn serve it to the application on a subsequent stat call as reported in the BZ. Fix: Pick a child on whom the fsync was successful *and* that is readable as indicated in the inode context. > Reviewed-on: https://review.gluster.org/17227 > CentOS-regression: Gluster Build System <jenkins@build.gluster.org> > Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> > NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> > Smoke: Gluster Build System <jenkins@build.gluster.org> (cherry picked from commit 1a8fa910ccba7aa941f673302c1ddbd7bd818e39) Change-Id: Ie8647289219cebe02dde4727e19a729b3353ebcf BUG: 1444892 RCA'ed-by: Miklós Fokin <miklos.fokin@appeartv.com> Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: https://review.gluster.org/17247 Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Raghavendra Talur <rtalur@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* afr: all children of AFR must be up to resolve s-brainRavishankar N2017-02-151-10/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: The various split-brain resolution policies (favorite-child-policy based, CLI based and mount (get/setfattr) based) attempt to resolve split-brain even when not all bricks of replica are up. This can be a problem when say in a replica 3, the only good copy is down and the other 2 bricks are up and blame each other (i.e. split-brain). We end up healing the file in such a case and allow I/O on it. Fix: A decision on whether the file is in split-brain or not must be taken only if we are able to examine the afr xattrs of *all* bricks of a given replica. Signed-off-by: Ravishankar N <ravishankar@redhat.com> > Reviewed-on: https://review.gluster.org/16476 > Smoke: Gluster Build System <jenkins@build.gluster.org> > NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> > CentOS-regression: Gluster Build System <jenkins@build.gluster.org> > Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> (cherry picked from commit 0e03336a9362e5717e561f76b0c543e5a197b31b) Change-Id: Icddb1268b380005799990f5379ef957d84639ef9 BUG: 1420982 Reviewed-on: https://review.gluster.org/16587 Tested-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Shyamsundar Ranganathan <srangana@redhat.com>
* afr: Avoid resetting event_gen when brick is always downRavishankar N2017-01-121-11/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: __afr_set_in_flight_sb_status(), which resets event_gen to zero, is called if failed_subvols[i] is non-zero for any brick. But failed_subvols[i] is true even if the brick was down *before* the transaction started. Hence say if 1 brick is down in a replica-3, every writev that comes will trigger an inode refresh because of this resetting, as seen from the no. of FSTATs in the profile info in the BZ. Fix: Reset event gen only if the brick was previously a valid read child and the FOP failed on it the first time. Also `s/afr_inode_read_subvol_reset/afr_inode_event_gen_reset` because the function only resets event gen and not the data/metadata readable. Change-Id: I603ae646cbde96995c35db77916e2ed80b602a91 BUG: 1409206 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/16309 Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* ec: Invalidations in disperse volume should not update the statPoornima G2017-01-051-2/+1
| | | | | | | | | | | | | | | | | | | | | | Issue: In disperse volume, the file is present across bricks, hence the stat from one brick doesn't carry the valid size of the file. Therefore the upcall from one brick updating the md-cache results in wrong size being updated. Fix: If the notification is cache invalidation then, indicate md-cache that the attributes is invalid. BUG: 1410375 Change-Id: Id89d2283478e70b62b435a8891fffc86d2be8cb2 Signed-off-by: Poornima G <pgurusid@redhat.com> Reviewed-on: http://review.gluster.org/16329 Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Xavier Hernandez <xhernandez@datalab.es> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: use accused matrix instead of readable matrix for deciding healsRavishankar N2016-12-261-1/+1
| | | | | | | | | | | | | | | | | | | | | | | Problem: afr_replies_interpret() used the 'readable' matrix to trigger client side heals after inode refresh. But for arbiter, readable is always zero. So when `dd` is run with a data brick down, spurious data heals are are triggered. These heals open an fd, causing eager lock to be disabled (open fd count >1) in afr transactions, leading to extra FXATTROPS Fix: Use the accused matrix (derived from interpreting the afr pending xattrs) to decide whether we can start heal or not. Change-Id: Ibbd56c9aed6026de6ec42422e60293702aaf55f9 BUG: 1408395 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/16277 NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: Ignore event_generation checks post inode refresh for write txnsRavishankar N2016-12-221-1/+1
| | | | | | | | | | | | | | | | | | | Before http://review.gluster.org/#/c/15673/, after inode refresh, we failed read txns in case of EIO or event_generation being zero. For write transactions, the check was only for EIO. 15673 re-factored the code to fail both read and write when event_generation=0. This seems to have caused a regression as explained in the BZ. This patch restores that behaviour in afr_txn_refresh_done(). Change-Id: Ib8e116506badce6f58b55827dbe403d95069d744 BUG: 1406224 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/16205 Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
* cluster/afr: Fix per-txn optimistic changelog initialisationKrutika Dhananjay2016-12-121-5/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Incorrect initialisation of local->optimistic_change_log was leading to skipped pre-op and post-op even when a brick didn't participate in the txn because it was down. The result - missing granular name index resulting in some entries never getting healed. FIX: Initialise local->optimistic_change_log just before pre-op. Also fixed granular entry heal to create the granular name index in pre-op as opposed to post-op. This is to prevent loss of granular information when during an entry txn, the good (src) brick goes offline before the post-op is done. This would cause self-heal to do conservative merge (since dirty xattr is the only information available), which when granular-entry-heal is enabled, expects granular indices, the lack of which can lead to loss of data in the worst case. Change-Id: Ia3ad716d6fb1821555f02180e86e8711a79f958d BUG: 1402730 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-on: http://review.gluster.org/16075 Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* cluster/afr: Remove backward compatibility for locks with v1Pranith Kumar K2016-12-071-36/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we have cascading locks with same lk-owner there is a possibility for a deadlock to happen. One example is as follows: self-heal takes a lock in data-domain for big name with 256 chars of "aaaa...a" and starts heal in a 3-way replication when brick-0 is offline and healing from brick-1 to brick-2 is in progress. So this lock is active on brick-1 and brick-2. Now brick-0 comes online and an operation wants to take full lock and the lock is granted at brick-0 and it is waiting for lock on brick-1. As part of entry healing it takes full locks on all the available bricks and then proceeds with healing the entry. Now this lock will start waiting on brick-0 because some other operation already has a granted lock on it. This leads to a deadlock. Operation is waiting for unlock on "aaaa..." by heal where as heal is waiting for the operation to unlock on brick-0. Initially I thought this is happening because healing is trying to take a lock on all the available bricks instead of just the bricks that are participating in heal. But later realized that same kind of deadlock can happen if a brick goes down after the heal starts but comes back before it completes. So the essential problem is the cascading locks with same lk-owner which were added for backward compatibility with afr-v1 which can be safely removed now that versions with afr-v1 are already EOL. This patch removes the compatibility with v1 which requires cascading locks with same lk-owner. In the next version we can make locking-scheme option a dummy and switch completely to v2. BUG: 1401404 Change-Id: Ic9afab8260f5ff4dff5329eb0429811bcb879079 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/16024 Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* cluster/afr: Serialize conflicting locks on all subvolsPranith Kumar K2016-12-061-32/+50
| | | | | | | | | | | | | | | | | | | | | | | Problem: 1) When a blocking lock is issued and the parallel lock phase fails on all subvolumes with EAGAIN, it is not switching to serialized locking phase. 2) When quorum is enabled and locks fail partially it is better to give errno returned by brick rather than the default quorum errno. Fix: Handled this error case and changed op_errno to reflect the actual errno in case of quorum error. BUG: 1369077 Change-Id: Ifac2e4a13686e9fde601873012700966d56a7f31 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/15984 Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com>
* afr, client: More mem-leak fixes in COMPOUND fop cbkKrutika Dhananjay2016-12-041-12/+0
| | | | | | | | | | | | | | | | | | | | | | Bugs found and fixed: 1. Use correct subvolume index in pre-op-writev compound cbk 2. Prevent use-after-free of local->compound_args members in compound fops cbk in protocol/client 3. Fix xdata and xattr leaks in client_process_response 4. Fix possible leak of xdata in client_pre_writev() in test mode. 5. Free req->compound_req_array.compound_req_array_val as well after freeing its members 6. Free tmp_rsp->flock.lk_owner.lk_owner_val in LK fop. Change-Id: I15b646d7d4e0e5cd4ea3d2d6452c815cf2eaf68f BUG: 1401218 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-on: http://review.gluster.org/16020 Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* selfheal: fix memory leak on client side healing queueMateusz Slupny2016-12-021-1/+4
| | | | | | | | | | | | | Change-Id: I2beaba829710565a3246f7449a5cd21755cf5f7d BUG: 1399592 Signed-off-by: Mateusz Slupny <mateusz.slupny@appeartv.com> Reviewed-on: http://review.gluster.org/15968 Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Reviewed-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org>
* afr: fix auto-quorumJeff Darcy2016-11-281-37/+0
| | | | | | | | | | | | | | | | | | | | | | | | | (1) afr_have_quorum is dead code. It was copied to afr_has_quorum, and everything else uses that, but the original was never deleted (until now). (2) Auto-quorum should be default for any N>2. Leaving quorum disabled is BAD, but apparently deemed acceptable for N=2 because there's no real quorum in that case. For any larger number (including arbiter configurations) there is such a thing as real quorum and we should use it by default. Note that for N=3 the answers we get from "N % 2" (the old check) and "N > 2" (the new one) are the same. (3) The special case for even N in afr_has_quorum has been simplified and explained more thoroughly in a comment. Change-Id: I48b33c15093512fecf516b26dcf09afecb7ae33b Signed-off-by: Jeff Darcy <jdarcy@redhat.com> Reviewed-on: http://review.gluster.org/15873 Smoke: Gluster Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Vijay Bellur <vbellur@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: Fix the EIO that can occur in afr_inode_refresh as a resultPoornima G2016-11-281-16/+74
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | of cache invalidation(upcall). Issue: ------ When a cache invalidation is recieved as a result of changing pending xattr, the read_subvol is reset. Consider the below chain of execution: CHILD_DOWN ... afr_readv ... afr_inode_refresh ... afr_inode_read_subvol_reset <- as a result of pending xattr set by some other client GF_EVENT_UPCALL will be sent afr_refresh_done -> this results in an EIO, as the read subvol was reset by the end of the afr_inode_refresh Solution: --------- When GF_EVENT_UPCALL is recieved, instead of resetting read_subvol, set a variable need_refresh in inode_ctx, the next time some one starts a txn, along with event gen, need_rrefresh also needs to be checked. Change-Id: Ifda21a7a8039b8874215e1afa4bdf20f7d991b58 BUG: 1396952 Signed-off-by: Poornima G <pgurusid@redhat.com> Reviewed-on: http://review.gluster.org/15892 Reviewed-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: allow I/O when favorite-child-policy is enabledRavishankar N2016-11-271-3/+202
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: Currently, I/O on a split-brained file fails even when the favorite-child-policy is set until the self-heal is complete. Fix: If a valid 'source' is found using the set favorite-child-policy, inspect and reset the afr pending xattrs on the 'sinks' (inside appropriate locks), refresh the inode and then proceed with the read or write transaction. The resetting itself happens in the self-heal code and hence can also happen in the client side background-heal or by the shd's index-heal in addition to the txn code path explained above. When it happens in via heal, we also add checks in undo-pending to not reset the sink xattrs again. Change-Id: Ic8c1317720cb26bd114b6fe6af4e58c73b864626 BUG: 1386188 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reported-by: Simon Turcotte-Langevin <simon.turcotte-langevin@ubisoft.com> Reviewed-on: http://review.gluster.org/15673 Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* cluster/afr: Fix bugs in [f]inodelk/[f]entrylkPranith Kumar K2016-11-261-325/+361
| | | | | | | | | | | | | | | | | | | | | | Problems: 1) Inodelk is not taking quorum into account 2) finodelk, [f]entrylk are not implemented correctly 3) By default afr doesn't go for non-blocking parallel locks. Fix: Implemented a common framework which can be used by [f]inodelk/[f]entrylk. Used quorum for the same. Change-Id: I239f13875a065298630d266941df10cfa3addc85 BUG: 1369077 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/15802 Tested-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
* afr,dht,ec: Replace GF_EVENT_CHILD_MODIFIED with event SOME_DESCENDENT_DOWN/UPPoornima G2016-11-211-22/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently these are few events related to child_up/down: GF_EVENT_CHILD_UP : Issued when any of the protocol client connects. GF_EVENT_CHILD_MODIFIED : Issued by afr/dht/ec GF_EVENT_CHILD_DOWN : Issued when any of the protocol client disconnects. These events get modified at the dht/afr/ec layers. Here is a brief on the same. DHT: - All the subvolumes reported once, and atleast one child came up, then GF_EVENT_CHILD_UP is issued - connect GF_EVENT_CHILD_UP is issued - disconnect GF_EVENT_CHILD_MODIFIED is issued - All the subvolumes disconnected, GF_EVENT_CHILD_DOWN is issued AFR: - First subvolume came up, then GF_EVENT_CHILD_UP is issued - Subsequent subvolumes coming up, results in GF_EVENT_CHILD_MODIFIED - Any of the subvolumes go down, then GF_EVENT_SOME_CHILD_DOWN is issued - Last up subvolume goes down, then GF_EVENT_CHILD_DOWN is issued Until the patch [1] introduced GF_EVENT_SOME_CHILD_UP, GF_EVENT_CHILD_MODIFIED was issued by afr/dht when any of the subvolumes go up or down. Now with md-cache changes, there is a necessity to differentiate between child up and down. Hence, introducing GF_EVENT_SOME_DESCENDENT_DOWN/UP and getting rid of GF_EVENT_CHILD_MODIFIED. [1] http://review.gluster.org/12573 Change-Id: I704140b6598f7ec705493251d2dbc4191c965a58 BUG: 1396038 Signed-off-by: Poornima G <pgurusid@redhat.com> Reviewed-on: http://review.gluster.org/15764 CentOS-regression: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: N Balachandran <nbalacha@redhat.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
* features/shard: Fill loc.pargfid too for named lookups on individual shardsKrutika Dhananjay2016-11-081-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On a sharded volume when a brick is replaced while IO is going on, named lookup on individual shards as part of read/write was failing with ENOENT on the replaced brick, and as a result AFR initiated name heal in lookup callback. But since pargfid was empty (which is what this patch attempts to fix), the resolution of the shards by protocol/server used to fail and the following pattern of logs was seen: Brick-logs: [2016-11-08 07:41:49.387127] W [MSGID: 115009] [server-resolve.c:566:server_resolve] 0-rep-server: no resolution type for (null) (LOOKUP) [2016-11-08 07:41:49.387157] E [MSGID: 115050] [server-rpc-fops.c:156:server_lookup_cbk] 0-rep-server: 91833: LOOKUP(null) (00000000-0000-0000-0000-000000000000/16d47463-ece5-4b33-9c93-470be918c0f6.82) ==> (Invalid argument) [Invalid argument] Client-logs: [2016-11-08 07:41:27.497687] W [MSGID: 114031] [client-rpc-fops.c:2930:client3_3_lookup_cbk] 2-rep-client-0: remote operation failed. Path: (null) (00000000-0000-0000-0000-000000000000) [Invalid argument] [2016-11-08 07:41:27.497755] W [MSGID: 114031] [client-rpc-fops.c:2930:client3_3_lookup_cbk] 2-rep-client-1: remote operation failed. Path: (null) (00000000-0000-0000-0000-000000000000) [Invalid argument] [2016-11-08 07:41:27.498500] W [MSGID: 114031] [client-rpc-fops.c:2930:client3_3_lookup_cbk] 2-rep-client-2: remote operation failed. Path: (null) (00000000-0000-0000-0000-000000000000) [Invalid argument] [2016-11-08 07:41:27.499680] E [MSGID: 133010] Also, this patch makes AFR by itself choose a non-NULL pargfid even if its ancestors fail to initialize all pargfid placeholders. Change-Id: I5f85b303ede135baaf92e87ec8e09941f5ded6c1 BUG: 1392445 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-on: http://review.gluster.org/15788 CentOS-regression: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org>
* md-cache, afr: Reduce the window of stale readPoornima G2016-10-201-1/+40
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: Consider a replica setup, where one mount writes data to a file and the other mount reads the file. In afr, read operations are not transaction based, a brick(read subvolume) is chosen as a part of lookup or other operations, read is always wound only to the read subvolume, even if there was write from a different client that failed on this brick. This stale read continues until there is a lookup or any write operation from the mount point. Currently, this is not a major issue, as a lookup is issued before every read and it will switch the read subvolume to a correct one. But with the plan of increasing md-cache timeout to 600s, the stale read problem will be more pronounced, i.e. stale read can continue for 600s(or more if cascaded with readdirp), as there will be no lookups. Solution: Afr doesn't have any built-in solution for stale read(without affecting the performance). The solution that came up, was to use upcall. When a file on any brick is marked bad for the first time, upcall sends a notification to all the clients that had recently accessed the file. The solution has 2 parts: - Identifying when a file is marked bad, on any of the bricks, for the first time - Client side actions on recieving the notifications Identifying when a file is marked bad on any of the bricks for the first time: ----------------------------------------------------------------------------- The idea is to track xattrop in upcall. xattrop currently comes with 2 afr xattrs - afr dirty bit and afr pending xattrs. Dirty xattr is set to 1 before every write, and is unset if write succeeds. In certain scenarios, dirty xattr can be 0 and still the file could be bad copy. Hence do not track dirty xattr. Pending xattr is set on the good copy, indicating the other bricks that have bad copy. It is still not as simple as, notifying when any of the pending xattrs change. It could lead to flood of notifcations, in case the other brick is completely down or consistantly failing. Hence it is important to notify only once, the first time a good copy is marked bad. Client side actions on recieving pending xattr change, notification: -------------------------------------------------------------------- md-cache will invalidate the cache of that file, so that further lookup is passed down to afr and hence update the read subvolume. Invalidating only in md-cache is not enough, consider the folling oder of opertaions: - pending xattr invalidation - invalidate md-cache - readdirp on the bad read subvolume - fill md-cache - lookup (served from md-cache) - read - wound to the old read subvol. Hence, along with invalidating md-cache, it is very important to reset the read subvolume for that file, in afr. Design Credit: Anuradha Talur, Ravishankar N 1. xattrop doesn't carry info saying post op/pre op. 2. Pre xattrop will have 0 value for all pending xattrs, the cbk of pre xattrop carries the on-disk xattr value. Non zero indicated healing is required. 3. Post xattrop will have non zero value for any of the pending xattrs, if the fop failed on any of the bricks. Change-Id: I469cbc111714c433984fe1c922be2ef113c25804 BUG: 1211863 Signed-off-by: Poornima G <pgurusid@redhat.com> Reviewed-on: http://review.gluster.org/15398 Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* cluster/afr: Prevent dict_set() on NULL dictPranith Kumar K2016-10-151-1/+2
| | | | | | | | | | | | | | | | | In afr lookup when NULL dict is received in lookup, afr is supposed to set all the xattrs it requires in a new dict it creates, but for 'link-count' it is trying to set to the dict that is passed in lookup which can be NULL sometimes. This is leading to error logs. Fixed the same in this patch. BUG: 1385104 Change-Id: I679af89cfc410cbc35557ae0691763a05eb5ed0e Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/15646 Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com>
* afr: Implement IPC fopPoornima G2016-09-291-0/+120
| | | | | | | | | | | | | | | | | | | | | Currently ipc() is not implemented in afr. md-cache and upcall uses ipc to register the list of xattrs, [1] for more details. For the ipc op GF_IPC_TARGET_UPCALL, it has to be wound to all the replica subvolumes. ipc() is failed when any of the subvolumes fails with other than ENOTCONN or all of the subvolumes are down. [1] http://review.gluster.org/#/c/15002/ Change-Id: I0f651330eafda64e4d922043fe53bd0014536247 BUG: 1211863 Signed-off-by: Poornima G <pgurusid@redhat.com> Reviewed-on: http://review.gluster.org/15378 Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: Ignore gluster internal (virtual) xattrs in metadata heal checkRavishankar N2016-09-261-6/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | Problem: In arbiter configuration, posix-xlator in the arbiter brick always sets the GF_CONTENT_KEY in the response dict with a value 0. If the file size on the data bricks is more than quick-read's max-file-size (64kb default), those bricks don't set the key. Because of this difference in the no. of dict elements, afr triggers metadata heal in lookup code path, in turn leading to extra lookups+inodelks. Fix: Changed afr dict comparison logic to ignore all virtual xattrs and the on-disk ones that we should not be healing. Also removed is_virtual_xattr() function. The original callers to this function (upcall) don't seem to need it anymore. Change-Id: I05730bdd39d8fb0b9a49a5fc9c0bb01f0d3bb308 BUG: 1378684 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/15548 NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Smoke: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: add replication eventsRavishankar N2016-09-061-2/+14
| | | | | | | | | | | | | | | | | | | | | Added the following events for the eventing framework: "EVENT_AFR_QUORUM_MET", --> Sent when quorum is met. "EVENT_AFR_QUORUM_FAIL" -->Sent when quorum is lost. "EVENT_AFR_SUBVOL_UP" -->Sent when afr witnesses the first up subvolume. "EVENT_AFR_SUBVOLS_DOWN"-->Sent when all children of an afr subvol are down. "EVENT_AFR_SPLIT_BRAIN" -->Sent when self-heal detects split-brain in heal path (not read/write path). Change-Id: I937c61ca1ce78b5922ade73c7bfa3051df59c513 BUG: 1371485 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/15349 Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* afr: Consume compound fops in afr transactionAnuradha Talur2016-09-011-0/+55
| | | | | | | | | | | | | Change-Id: Ib06ece3cce1b10d28d6d2953da28444f5c2457ad BUG: 1290304 Signed-off-by: Anuradha Talur <atalur@redhat.com> Reviewed-on: http://review.gluster.org/15014 Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* cluster/afr: Give option to do consistent-ioPranith Kumar K2016-08-221-32/+161
| | | | | | | | | | | | | | | | | | | | | | | Problem: When tiering/rebalance does migrations and afr with 2-way replica is in picture, migration can read stale data if the source brick goes down and writes to the destination. After this deletion of the file leads to permanent loss of data after migration. Fix: Rebalance/tiering should migrate only when the data is definitely not stale. So introduce an option in afr called consistent-io which will be enabled in migration daemons. BUG: 1306398 Change-Id: I750f65091cc70a3ed4bf3c12f83d0949af43920a Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/13425 Reviewed-by: Anuradha Talur <atalur@redhat.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* cluster/afr: Prevent split-brain when bricks are brought off and on in ↵Krutika Dhananjay2016-08-221-6/+119
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | cyclic order When the bricks are brought offline and then online in cyclic order while writes are in progress on a file, thanks to inode refresh in write txns, AFR will mostly fail the write attempt when the only good copy is offline. However, there is still a remote possibility that the file will run into split-brain if the brick that has the lone good copy goes offline *after* the inode refresh but *before* the write txn completes (I call it in-flight split-brain in the patch for ease of reference), requiring intervention from admin to resolve the split-brain before the IO can resume normally on the file. To get around this, the patch does the following things: i) retains the dirty xattrs on the file ii) avoids marking the last of the good copies as bad (or accused) in case it is the one to go down during the course of a write. iii) fails that particular write with the appropriate errno. This way, we still have one good copy left despite the split-brain situation which when it is back online, will be chosen as source to do the heal. Change-Id: I9ca634b026ac830b172bac076437cc3bf1ae7d8a BUG: 1363721 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-on: http://review.gluster.org/15080 Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> Reviewed-by: Oleksandr Natalenko <oleksandr@natalenko.name> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: some coverity fixesRavishankar N2016-07-261-92/+136
| | | | | | | | | | | | | | | Thanks to Krutika for a cleaner way to track inode refs in afr_set_split_brain_choice(). Change-Id: I2d968d05b815ad764b7e3f8aa9ad95a792b3c1df BUG: 1355604 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/14895 Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.org> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
* cluster/afr: Unwind with xdata in inode-write fopsPranith Kumar K2016-06-011-1/+1
| | | | | | | | | | | | | | | When there is a failure afr was not unwinding xdata to xlators above. xdata need not be NULL on failures. So it is important to send it to parent xlators. Change-Id: Ic36aac10a79fa91121961932dd1920cb1c2c3a4c BUG: 1340623 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/14567 Smoke: Gluster Build System <jenkins@build.gluster.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/afr: Unwind xdata_rsp even in case of failuresPranith Kumar K2016-05-301-8/+21
| | | | | | | | | | | | | | | | | | | DHT expects GF_PREOP_CHECK_FAILED to be present in xdata_rsp in case of mkdir failures because of stale layout. But AFR was unwinding null xdata_rsp in case of failures. This was leading to mkdir failures just after remove-brick. Unwind the xdata_rsp in case of failures to make sure the response from brick reaches dht. BUG: 1340623 Change-Id: Idd3f7b95730e8ea987b608e892011ff190e181d1 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/14553 NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anuradha Talur <atalur@redhat.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
* cluster/afr: Do not inode_link in afrPranith Kumar K2016-05-201-5/+1
| | | | | | | | | | | | | | | | | | Race is explained at https://bugzilla.redhat.com/show_bug.cgi?id=1337405#c0 This patch also handles performing of self-heal with shd-pid. Also performs the healing with this->itable's inode rather than main itable. BUG: 1337405 Change-Id: Id657a6623b71998b027b1dff6af5bbdf8cab09c9 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/14422 Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
* cluster/afr: Refresh inode for inode-write fops in needPranith Kumar K2016-05-191-10/+21
| | | | | | | | | | | | | | | | | | | | | | | | | Problem: If a named fresh-lookup is done on an loc and the fop fails on one of the bricks or not sent on one of the bricks, but by the time response comes to afr, if the brick is up, 'can_interpret' will be set to false in afr_lookup_done(), this will lead to inode-ctx for that inode to be not set, this can lead to EIO in case of a transaction as it depends on 'readable' array to be available by that point. Fix: Refresh inode for inode-write fops for the ctx to be set if it is not already done at the time of named fresh-lookup or if the file is in split-brain where we need to perform one more refresh before failing the fop to check if the file is still in split-brain or not. BUG: 1336612 Change-Id: I5c50b62c8de06129b8516039f7c252e5008c47a5 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/14368 Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
* cluster/afr: Handle non-zero source in heal-info decisionPranith Kumar K2016-05-121-2/+2
| | | | | | | | | | | | | | | | | | | | | | | Problem: Spurious entries are reported in heal info when the mount is on second/third brick of the replica pair because local-child is given preference in selecting source. The code is supposed to suggest the file needs heal if the (source < 0) (failure code path), but instead it is written as if any non-zero value is considered failure. Fix: Treat +ve source as success case BUG: 1335429 Change-Id: I1be7f9defef2ae03be7eec8d7d49bf34adeca82c Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/14302 Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-by: Anuradha Talur <atalur@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Ravishankar N <ravishankar@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
* cluster/afr: Entry self-heal performance enhancementsKrutika Dhananjay2016-04-291-2/+11
| | | | | | | | | | | Change-Id: I52da41dff5619492b656c2217f4716a6cdadebe0 BUG: 1269461 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-on: http://review.gluster.org/12442 Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> 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>
* afr: propagate child up event after timeoutRavishankar N2016-04-271-96/+184
| | | | | | | | | | | | | | | | | | | | | Problem: During mount, afr waits for response from all its children before notifying the parent xlator. In a 1x2 replica volume , if one of the nodes is down, the mount will hang for more than a minute until child down is received from the client xlator for that node. Fix: When parent up is received by afr, start a 10 second timer. In the timer call back, if we receive a successful child up from atleast one brick, propagate the event to the parent xlator. Change-Id: I31e57c8802c1a03a4a5d581ee4ab82f3a9c8799d BUG: 1054694 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/11113 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>
* cluster/afr: Fix spurious entries in heal infoPranith Kumar K2016-04-201-10/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: Locking schemes in afr-v1 were locking the directory/file completely during self-heal. Newer schemes of locking don't require Full directory, file locking. But afr-v2 still has compatibility code to work-well with older clients, where in entry-self-heal it takes a lock on a special 256 character name which can't be created on the fs. Similarly for data self-heal there used to be a lock on (LLONG_MAX-2, 1). Old locking scheme requires heal info to take sh-domain locks before examining heal-state. If it doesn't take sh-domain locks, then there is a possibility of heal-info hanging till self-heal completes because of compatibility locks. But the problem with heal-info taking sh-domain locks is that if two heal-info or shd, heal-info try to inspect heal state in parallel using trylocks on sh-domain, there is a possibility that both of them assuming a heal is in progress. This was leading to spurious entries being shown in heal-info. Fix: As long as there is afr-v1 way of locking, we can't fix this problem with simple solutions. If we know that the cluster is running newer versions of locking schemes, in those cases we can give accurate information in heal-info. So introduce a new option called 'locking-scheme' which if it is 'granular' will give correct information in heal-info. Not only that, Extra network hops for taking compatibility locks, sh-domain locks in heal info will not be necessary anymore. Thus it improves performance. BUG: 1322850 Change-Id: Ia563c5f096b5922009ff0ec1c42d969d55d827a3 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/13873 Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Ashish Pandey <aspandey@redhat.com> Reviewed-by: Anuradha Talur <atalur@redhat.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
* cluster/afr: Don't lookup/forget inodesPranith Kumar K2016-03-311-5/+2
| | | | | | | | | | | | | | | | | | | | | | | | Problem: All inodes that are looked-up are always forgotten without fail in afr removing the benefits of them being in lru. This same code can cause crashes if between inode_lookup, inode_forget in afr if the top xlator does inode_forget(0). Fix: Don't use lookup/forget in afr. No benefits are there at the moment for keeping this code. It is impossible to prevent top xlators to do inode_forget(0). Found similar instances in ec and removed them even though those code paths are not going to be executed in any place other than heal-daemon. BUG: 1321554 Change-Id: Ia4cb236178f7f129cc898d53f0bbd26f494a2a8d Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/13834 Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anuradha Talur <atalur@redhat.com>
* afr : Enable auto heal when replica count increasesAnuradha Talur2016-03-211-0/+11
| | | | | | | | | | | | | | | | | | | | | | | This patch is part two change to prevent data loss in a replicate volume on doing a add-brick operation. Problem: After doing add-brick, there is a chance that self heal might happen from the newly added brick rather than the source brick, leading to data loss. Solution: Mark pending changelogs on afr children for the new afr-child so that heal is performed in the correct direction. Change-Id: I11871e55eef3593aec874f92214a2d97da229b17 BUG: 1276203 Signed-off-by: Anuradha Talur <atalur@redhat.com> Reviewed-on: http://review.gluster.org/12454 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/afr: Choose local child as source if possiblePranith Kumar K2016-03-111-0/+1
| | | | | | | | | | | | | | | | | It is better to choose local brick as source if possible to prevent over the wire read thus saving on bandwidth. Also changed code to not attempt data-heal if 'source' is selected as arbiter. Change-Id: I9a328d0198422280b13a30ab99545370a301dfea BUG: 1314150 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/13585 Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Tested-by: Krutika Dhananjay <kdhananj@redhat.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
* afr: misc performance improvementsRavishankar N2016-03-081-28/+46
| | | | | | | | | | | | | | | | | | | | | 1. In afr_getxattr_cbk, consider the errno value before blindly launching an inode refresh and a subsequent retry on other children. 2. We want to accuse small files only when we know for sure that there is no IO happening on that inode. Otherwise, the ia_sizes obtained in the post-inode-refresh replies may mismatch due to a race between inode-refresh and ongoing writes, causing spurious heal launches. Change-Id: Ife180f4fa5e584808c1077aacdc2423897675d33 BUG: 1309462 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/13595 Smoke: Gluster Build System <jenkins@build.gluster.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: do not set arbiter as a readable subvol in inode contextRavishankar N2016-03-041-0/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Problem: If afr_lookup_done() or afr_read_subvol_select_by_policy() chooses the arbiter brick to serve the stat() data, file size will be reported as zero from the mount, despite other data bricks being available. This can break programs like tar which use the stat info to decide how much to read. Fix: In the inode-context, mark arbiter as a non-readable subvol for both data and metadata. It it to be noted that by making this fix, we are *not* going to serve metadata FOPS anymore from the arbiter brick despite the brick storing the metadata. It makes sense to do this because the ever increasing over-loaded FOPs (getxattr returning stat data etc.) and compound FOPS in gluster will otherwise make it difficult to add checks in code to handle corner cases. Change-Id: Ic60b25d77fd05e0897481b7fcb3716d4f2101001 BUG: 1310171 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reported-by: Mat Clayton <mat@mixcloud.com> Reviewed-on: http://review.gluster.org/13539 Reviewed-by: Anuradha Talur <atalur@redhat.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.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/afr: Don't delete gfid-req from lookup requestPranith Kumar K2016-03-021-3/+7
| | | | | | | | | | | | | | | | | | | | | | | | Problem: Afr does dict_ref of the xattr_req that comes to it and deletes "gfid-req" key. Dht uses same dict to send lookup to other subvolumes. So in case of directories and more than 1 dht subvolumes, second subvolume till the last subvolume won't get a lookup request with "gfid-req". So gfid reset never happens on the directories in distributed replicate subvolume for 2nd till last subvolumes. Fix: Make a copy of lookup xattr request. Also fixed replies_wipe possibly resetting gfid to NULL gfid BUG: 1312816 Change-Id: Ic16260e5a4664837d069c1dc05b9e96ca05bda88 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/13545 Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
* afr: Add throttled background client-side healsRavishankar N2016-03-011-48/+41
| | | | | | | | | | | | | | | | | | | | | | | | | | | If a heal is needed after inode refresh (lookup, read_txn), launch it in the background instead of blocking the fop (that triggered refresh) until the heal happens. afr_replies_interpret() is modified such that the heal is launched only if atleast one sink brick is up. Max. no of heals that can happen in parallel is configurable via the 'background-self-heal-count' volume option. Any number greater than that is put in a wait queue whose length is configurable via 'heal-wait-queue-leng' volume option. If the wait queue is also full, further heals will be ignored. Default values: background-self-heal-count=8, heal-wait-queue-leng=128 Change-Id: I1d4a52814cdfd43d90591b6d2ad7b6219937ce70 BUG: 1297172 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/13207 Smoke: Gluster Build System <jenkins@build.gluster.com> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
* cluster/afr: Fix heal-info slow response while IO is in progressKrutika Dhananjay2016-02-031-1/+16
| | | | | | | | | | | | | | | | | | Now heal-info does an open() on the file being examined so that the client at some point sees open-fd count being > 1 and releases the eager-lock so that heal-info doesn't remain blocked forever until IO completes. Change-Id: Icc478098e2bc7234408728b54d8185102b3540dc BUG: 1297695 Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-on: http://review.gluster.org/13326 Reviewed-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
* cluster/afr: Move remaining gf_logs to gf_msgsPranith Kumar K2016-01-271-5/+8
| | | | | | | | | | | | Change-Id: I48d9e5313bd3ccf9fe26c90a7051a8a174d75c49 BUG: 1296818 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/13195 Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* afr : Check if dict is valid in afr_replies_interpret()Ravishankar N2016-01-211-1/+2
| | | | | | | | | | | | | | | | | posix_mkdir does not send response xdata. So even though replies are valid, the response xdata dict is NULL. Check if dict is non-null in afr_replies_interpret before doing dict_get Signed-off-by: Ravishankar N <ravishankar@redhat.com> Change-Id: If543d68d8bfd2433519105839d5be106076cc276 BUG: 1294053 Reviewed-on: http://review.gluster.org/13185 Tested-by: Ravishankar N <ravishankar@redhat.com> Smoke: Gluster Build System <jenkins@build.gluster.com> NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org> CentOS-regression: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* afr: Fix excessive logging in afr_accuse_smallfiles()Ravishankar N2015-12-281-1/+2
| | | | | | | | | | | | | Commit 2b7226f9d3470d8fe4c98c1fddb06e7f641e364d did not check for the validity of a dict before doing a dict_get. Fix that. Change-Id: Ie21f4da19256b17196f242cd8fd5bb76b0a69c1e BUG: 1294053 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/13077 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>
generated by cgit (git 2.34.1) at 2024-05-27 23:14:08 +0000