Bug 1302948
Summary: | tar complains: <fileName>: file changed as we read it | |||
---|---|---|---|---|
Product: | [Community] GlusterFS | Reporter: | Sakshi <sabansal> | |
Component: | distribute | Assignee: | Sakshi <sabansal> | |
Status: | CLOSED CURRENTRELEASE | QA Contact: | ||
Severity: | medium | Docs Contact: | ||
Priority: | medium | |||
Version: | mainline | CC: | bkunal, george.lian, nbalacha, olim, pkarampu, rgowdapp, sabansal, smohan, spalai, storage-qa-internal | |
Target Milestone: | --- | |||
Target Release: | --- | |||
Hardware: | Unspecified | |||
OS: | Unspecified | |||
Whiteboard: | ||||
Fixed In Version: | glusterfs-3.8rc2 | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | ||
Clone Of: | 1298724 | |||
: | 1312721 1335285 (view as bug list) | Environment: | ||
Last Closed: | 2016-06-16 13:56:06 UTC | Type: | Bug | |
Regression: | --- | Mount Type: | --- | |
Documentation: | --- | CRM: | ||
Verified Versions: | Category: | --- | ||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | ||
Cloudforms Team: | --- | Target Upstream Version: | ||
Embargoed: | ||||
Bug Depends On: | ||||
Bug Blocks: | 1312721, 1335285 |
Description
Sakshi
2016-01-29 06:10:48 UTC
REVIEW: http://review.gluster.org/13352 (dht: mkdir must unwind with latest ctime) posted (#1) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13352 (dht: mkdir must unwind with latest ctime) posted (#2) for review on master by Sakshi Bansal COMMIT: http://review.gluster.org/13352 committed in master by Raghavendra G (rgowdapp) ------ commit dbf0ca53a5704c7f94257fc4223b41865458301a Author: Sakshi Bansal <sabansal> Date: Thu Feb 4 16:02:59 2016 +0530 dht: mkdir must unwind with latest ctime Currently fops like mkdir used the the ctime it gets after creating the directory entry. But setting layout also updates the ctime of a directory. Hence DHT must get the ctime after the setxattr call and unwind with the latest ctime to avoid mismatch in time seen by applications like tar. Change-Id: Iecbbe3aac5244af5da9788b48ccf299ca56b4bae BUG: 1302948 Signed-off-by: Sakshi Bansal <sabansal> Reviewed-on: http://review.gluster.org/13352 Smoke: Gluster Build System <jenkins.com> CentOS-regression: Gluster Build System <jenkins.com> NetBSD-regression: NetBSD Build System <jenkins.org> Reviewed-by: Raghavendra G <rgowdapp> REVIEW: http://review.gluster.org/13781 (dht: lock around dht_iatt_merge after setting layout) posted (#1) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13782 (NFS: do not fill attr information when returning directory entries in readdirp) posted (#1) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13781 (dht: update attr information in refresh layout to avoid stale timestamp) posted (#2) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13770 (dht: report constant directory size) posted (#3) for review on master by N Balachandran (nbalacha) REVIEW: http://review.gluster.org/13782 (NFS: do not fill attr information when returning directory entries in readdirp) posted (#2) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13770 (dht: report constant directory size) posted (#4) for review on master by N Balachandran (nbalacha) REVIEW: http://review.gluster.org/13770 (dht: report constant directory size) posted (#5) for review on master by N Balachandran (nbalacha) COMMIT: http://review.gluster.org/13770 committed in master by Vijay Bellur (vbellur) ------ commit 942d127fee13653daba49195692c5e8465f1fbff Author: Jeff Darcy <jdarcy> Date: Thu Mar 17 13:33:34 2016 -0400 dht: report constant directory size Directory size is meaningless. Every filesystem has its own unpredictable way of increasing or decreasing it, based on internal data structures and even transient conditions. Some filesystems (e.g. ext4) never decrease it at all. Others (e.g. btrfs) don't even report it. Very few programs look at it, and those that do are broken. Unfortunately, one such program is GNU tar, which will complain when it sees different values because at different times we got the value from different DHT subvolumes. To avoid such problems, just report a constant value. Change-Id: Id64ce917c75b5f7ff50cb55b6e997f3b3556e7e3 BUG: 1302948 Original-author: Shyam <srangana> Signed-off-by: Jeff Darcy <jdarcy> Signed-off-by: N Balachandran <nbalacha> Reviewed-on: http://review.gluster.org/13770 Smoke: Gluster Build System <jenkins.com> NetBSD-regression: NetBSD Build System <jenkins.org> CentOS-regression: Gluster Build System <jenkins.com> Reviewed-by: Shyamsundar Ranganathan <srangana> Reviewed-by: Vijay Bellur <vbellur> COMMIT: http://review.gluster.org/13781 committed in master by Raghavendra G (rgowdapp) ------ commit f14e800cd9bfe34822e1530f71f233bfe4fb04c0 Author: Sakshi Bansal <sabansal> Date: Sat Mar 19 10:15:24 2016 +0530 dht: update attr information in refresh layout to avoid stale timestamp Consider the scenario where an mkdir has just created the directory but has not healed it yet. A parallel lookup on this entry will find anomalies and trigger a selfheal which will sample the ctime of the directory after the mkdir phase. Meanwhile the mkdir has completed setting the layout and updated the ctime. The selfheal then sees the layout to be healed and returns with the ctime it got after the mkdir phase which has now become stale. However if the lookup happens to unwind before the mkdir then the inode associated with lookup will get linked in the inode table which has the stale ctime. To avoid this selfheal must do an iatt_merge in refresh layout to get the latest timestamp irrespective of whether it needs to heal the layout or not. Change-Id: I3634c3978bcc1710705f44b48f3876601682d33e BUG: 1302948 Signed-off-by: Sakshi Bansal <sabansal> Reviewed-on: http://review.gluster.org/13781 Smoke: Gluster Build System <jenkins.com> Reviewed-by: Raghavendra G <rgowdapp> NetBSD-regression: NetBSD Build System <jenkins.org> CentOS-regression: Gluster Build System <jenkins.com> REVIEW: http://review.gluster.org/13827 (cluster/afr: Fix read-child selection in entry create fop) posted (#3) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13827 (cluster/afr: Fix read-child selection in entry create fop) posted (#4) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13826 (debug/trace: Print {acm}times as integers) posted (#2) for review on master by Krutika Dhananjay (kdhananj) REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#4) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#5) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13826 (debug/trace: Print {acm}times as integers) posted (#4) for review on master by Krutika Dhananjay (kdhananj) COMMIT: http://review.gluster.org/13827 committed in master by Pranith Kumar Karampuri (pkarampu) ------ commit 2d0c2a9e15a581950a966841866f1f18a8888440 Author: Pranith Kumar K <pkarampu> Date: Fri Mar 25 06:04:23 2016 +0530 cluster/afr: Fix read-child selection in entry create fop When an entry is being created the inode is yet to be linked so args must be filled with gfid and ia_type for it to give consistent iatt. Also handle Dht sending fops on inode not yet linked. BUG: 1302948 Change-Id: I6969cacb437cad02f66716f3bf8ec004ffe7c691 Signed-off-by: Pranith Kumar K <pkarampu> Reviewed-on: http://review.gluster.org/13827 Reviewed-by: Krutika Dhananjay <kdhananj> Smoke: Gluster Build System <jenkins.com> Reviewed-by: Anuradha Talur <atalur> NetBSD-regression: NetBSD Build System <jenkins.org> CentOS-regression: Gluster Build System <jenkins.com> REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#6) for review on master by Pranith Kumar Karampuri (pkarampu) This bug was accidentally moved from POST to MODIFIED via an error in automation, please see mmccune with any questions REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#7) for review on master by Pranith Kumar Karampuri (pkarampu) COMMIT: http://review.gluster.org/13826 committed in master by Jeff Darcy (jdarcy) ------ commit 516db811a3bc1b3e778c383fc18aa5eb56342f77 Author: Krutika Dhananjay <kdhananj> Date: Thu Mar 24 20:43:05 2016 +0530 debug/trace: Print {acm}times as integers This helps in debugging issues with mtime, ctime etc. Change-Id: I4a0335f17c6b40e6c627658e91f43989c3efca22 BUG: 1302948 Signed-off-by: Krutika Dhananjay <kdhananj> Reviewed-on: http://review.gluster.org/13826 Smoke: Gluster Build System <jenkins.com> Reviewed-by: Niels de Vos <ndevos> CentOS-regression: Gluster Build System <jenkins.com> NetBSD-regression: NetBSD Build System <jenkins.org> REVIEW: http://review.gluster.org/13782 (NFS: new option nfs.dirplus-stat added) posted (#3) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13782 (NFS: new option nfs.rdirplus added) posted (#4) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13782 (NFS: new option nfs.rdirplus added) posted (#5) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13782 (NFS: new option nfs.rdirplus added) posted (#6) for review on master by Sakshi Bansal REVIEW: http://review.gluster.org/13782 (NFS: new option nfs.rdirplus added) posted (#7) for review on master by Sakshi Bansal COMMIT: http://review.gluster.org/13782 committed in master by Niels de Vos (ndevos) ------ commit 50b93f72f9a3f5bfd1d610801aecb06823adaa63 Author: Sakshi Bansal <sabansal> Date: Thu Mar 31 15:02:03 2016 +0530 NFS: new option nfs.rdirplus added When this option is 'disabled', NFS falls back to standard readdir instead of readdirp Change-Id: Icaaf4da6533bee56160d4a81e42bb60f7d341945 BUG: 1302948 Signed-off-by: Sakshi Bansal <sabansal> Reviewed-on: http://review.gluster.org/13782 Smoke: Gluster Build System <jenkins.com> NetBSD-regression: NetBSD Build System <jenkins.org> CentOS-regression: Gluster Build System <jenkins.com> Reviewed-by: Niels de Vos <ndevos> REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#8) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#9) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#10) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#11) for review on master by Pranith Kumar Karampuri (pkarampu) REVIEW: http://review.gluster.org/13785 (cluster/afr: Don't let NFS cache stat after writes) posted (#12) for review on master by Pranith Kumar Karampuri (pkarampu) COMMIT: http://review.gluster.org/13785 committed in master by Pranith Kumar Karampuri (pkarampu) ------ commit 4c4624c9bad2edf27128cb122c64f15d7d63bbc8 Author: Pranith Kumar K <pkarampu> Date: Sat Mar 19 11:40:26 2016 +0530 cluster/afr: Don't let NFS cache stat after writes Problem: Afr does post-ops after write but the stat buffer it unwinds is at the time of write, so if nfs client caches this, it will see different ctime when it does stat on it after post-op is done. From NFS client's perspective it thinks the file is changed. Tar which depends on this to be correct keeps giving 'file changed as we read it' warning. If Afr instead has to choose to unwind after post-op, eager-lock, delayed-post-op will have to be disabled which will lead to bad performance for all write usecases. Fix: Don't let client cache stat after write. Change-Id: Ic6062acc6e5cdd97a9c83c56bd529ec83cee8a23 BUG: 1302948 Signed-off-by: Pranith Kumar K <pkarampu> Signed-off-by: Anuradha Talur <atalur> Reviewed-on: http://review.gluster.org/13785 Smoke: Gluster Build System <jenkins.com> NetBSD-regression: NetBSD Build System <jenkins.org> CentOS-regression: Gluster Build System <jenkins.com> Reviewed-by: Niels de Vos <ndevos> Are there any more patches sent for this issue. If not please move the status to MODIFIED This bug is getting closed because a release has been made available that should address the reported issue. In case the problem is still not fixed with glusterfs-3.8.0, please open a new bug report. glusterfs-3.8.0 has been announced on the Gluster mailinglists [1], packages for several distributions should become available in the near future. Keep an eye on the Gluster Users mailinglist [2] and the update infrastructure for your distribution. [1] http://blog.gluster.org/2016/06/glusterfs-3-8-released/ [2] http://thread.gmane.org/gmane.comp.file-systems.gluster.user I suppose a new issue maybe introduced by this fix, the new issue I has report in https://bugzilla.redhat.com/show_bug.cgi?id=1531457 |