Bug 762058 (GLUSTER-326) - [2.0.8rc9] Spurious self-heal
Summary: [2.0.8rc9] Spurious self-heal
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: GLUSTER-326
Product: GlusterFS
Classification: Community
Component: replicate
Version: 2.0.7
Hardware: All
OS: Linux
low
high
Target Milestone: ---
Assignee: Vikas Gorur
QA Contact:
URL:
Whiteboard:
: GLUSTER-330 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2009-10-20 07:51 UTC by Anush Shetty
Modified: 2009-12-07 13:31 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed:
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:

Attachments (Terms of Use)
Add an attachment (proposed patch, testcase, etc.)

Description Anush Shetty 2009-10-20 07:51:33 UTC 
Self-heal is triggered even when there is no disconnection

Client-log

[2009-10-20 03:19:14] D [inode.c:572:__inode_replace] fuse/inode: inode(787820) replaced (787820
[2009-10-20 03:19:14] D [inode.c:572:__inode_replace] fuse/inode: inode(14943397) replaced (14943397
[2009-10-20 03:19:14] D [inode.c:572:__inode_replace] fuse/inode: inode(787808) replaced (787808
[2009-10-20 03:19:16] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate2: self-healing directory /clients/client0/~dmtmp/WORDPRO from subvolume remote4 to 1 other
[2009-10-20 03:19:16] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate1: self-healing directory /clients/client1/~dmtmp/WORDPRO from subvolume remote2 to 1 other
[2009-10-20 03:19:18] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate2: self-healing directory /clients/client8/~dmtmp/WORDPRO from subvolume remote4 to 1 other
[2009-10-20 03:19:19] D [inode.c:572:__inode_replace] fuse/inode: inode(722356) replaced (722356
[2009-10-20 03:19:19] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate2: self-healing directory /clients/client6/~dmtmp/WORDPRO from subvolume remote4 to 1 other
[2009-10-20 03:19:19] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate2: self-healing directory /clients/client15/~dmtmp/WORDPRO from subvolume remote3 to 1 other
[2009-10-20 03:19:19] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate2: self-healing directory /clients/client1/~dmtmp/WORDPRO from subvolume remote4 to 1 other
[2009-10-20 03:19:19] D [afr-self-heal-entry.c:1880:afr_sh_entry_sync_prepare] replicate2: self-healing directory /clients/client8/~dmtmp/WORDPRO from subvolume remote4 to 1 other
[2009-10-20 03:19:19] D [inode.c:572:__inode_replace] fuse/inode: inode(787856) replaced (787856
Comment 1 Anush Shetty 2009-10-21 01:06:12 UTC 
This bug seems to be much older. It exists even in 2.0.4 . The dbench client dir doesn't delete itself on completion
Comment 2 Vikas Gorur 2009-10-23 09:10:40 UTC 
*** Bug 330 has been marked as a duplicate of this bug. ***
Comment 3 Anand Avati 2009-10-23 09:54:56 UTC 
PATCH: http://patches.gluster.com/patch/1982 in master (cluster/afr: Prevent spurious entry self-heal.)
Comment 4 Anand Avati 2009-10-23 09:55:00 UTC 
PATCH: http://patches.gluster.com/patch/1983 in release-2.0 (cluster/afr: Prevent spurious entry self-heal.)
Comment 5 Anand Avati 2009-10-29 13:45:40 UTC 
PATCH: http://patches.gluster.com/patch/2064 in master (storage/posix: Refactor xattrop)
Comment 6 Anand Avati 2009-10-29 13:45:44 UTC 
PATCH: http://patches.gluster.com/patch/2065 in master (storage/posix: Serialize do_xattrop.)
Comment 7 Anand Avati 2009-10-29 16:02:49 UTC 
PATCH: http://patches.gluster.com/patch/2066 in release-2.0 (storage/posix: Refactor xattrop)
Comment 8 Anand Avati 2009-10-29 16:02:53 UTC 
PATCH: http://patches.gluster.com/patch/2067 in release-2.0 (storage/posix: Serialize do_xattrop.)
Comment 9 Anand Avati 2009-10-30 08:11:00 UTC 
PATCH: http://patches.gluster.com/patch/2087 in master (cluster/afr: Don't try to self-heal if there are locks held)
Comment 10 Anand Avati 2009-10-30 08:31:02 UTC 
PATCH: http://patches.gluster.com/patch/2090 in release-2.0 (cluster/afr: Don't try to self-heal if there are locks held)
Comment 11 Anand Avati 2009-12-03 07:21:31 UTC 
PATCH: http://patches.gluster.com/patch/2516 in master (cluster/afr: Don't mark the fop as failed if rmdir returns ENOTEMPTY.)
Comment 12 Anand Avati 2009-12-07 10:31:32 UTC 
PATCH: http://patches.gluster.com/patch/2598 in master (cluster/afr: Hold inode lock on 0-infinity for a flush transaction.)
Comment 13 Anand Avati 2009-12-07 10:31:35 UTC 
PATCH: http://patches.gluster.com/patch/2599 in master (cluster/afr: Don't do entry self-heal in the background.)
Note You need to log in before you can comment on or make changes to this bug.