Bug 762052 (GLUSTER-320) - Improve self-heal performance
Summary: Improve self-heal performance
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: GLUSTER-320
Product: GlusterFS
Classification: Community
Component: replicate
Version: mainline
Hardware: All
OS: Linux
low
low
Target Milestone: ---
Assignee: Vikas Gorur
QA Contact:
URL:
Whiteboard:
: GLUSTER-457 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2009-10-15 09:11 UTC by Vikas Gorur
Modified: 2010-01-20 14:30 UTC (History)
3 users (show)

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

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

Description Vikas Gorur 2009-10-15 09:11:43 UTC 
Pipelining the read/write loops of the data self-heal algorithms should increase performance dramatically.
Comment 1 Anand Avati 2009-10-23 11:03:14 UTC 
PATCH: http://patches.gluster.com/patch/1913 in master (cluster/afr: Pipeline the "full" data self-heal read-write loop.)
Comment 2 Anand Avati 2009-10-23 12:29:28 UTC 
PATCH: http://patches.gluster.com/patch/1991 in master (cluster/afr: Pipeline the "diff" data self-heal read-write loop.)
Comment 3 Anand Avati 2009-10-26 11:09:08 UTC 
PATCH: http://patches.gluster.com/patch/1992 in master (cluster/afr: Do self-heal in the background.)
Comment 4 Anand Avati 2009-10-26 11:09:12 UTC 
PATCH: http://patches.gluster.com/patch/1995 in master (cluster/afr: Set read-subvolume to the source while self-heal is in progress.)
Comment 5 Anand Avati 2009-10-26 11:09:34 UTC 
PATCH: http://patches.gluster.com/patch/2004 in master (cluster/afr: Pick the data self-heal algorithm dynamically.)
Comment 6 Anand Avati 2009-11-03 13:16:57 UTC 
PATCH: http://patches.gluster.com/patch/2119 in master (cluster/afr: Bail out of the self-heal driver loops as soon as possible)
Comment 7 Anand Avati 2009-11-13 07:13:49 UTC 
PATCH: http://patches.gluster.com/patch/2189 in master (cluster/afr: Fix self-heal loop driver termination logic.)
Comment 8 Anand Avati 2009-11-29 02:30:13 UTC 
PATCH: http://patches.gluster.com/patch/2408 in master (afr: fix logic to break out of diff/checksum loop)
Comment 9 Anand Avati 2009-12-01 11:30:03 UTC 
PATCH: http://patches.gluster.com/patch/2454 in master (cluster/afr: Set the self-heal "source" as read subvolume even when not doing self-heal.)
Comment 10 Anand Avati 2009-12-01 11:30:07 UTC 
PATCH: http://patches.gluster.com/patch/2453 in master (cluster/afr: Set file size properly before unwinding from self-heal.)
Comment 11 Anand Avati 2009-12-01 11:30:10 UTC 
PATCH: http://patches.gluster.com/patch/2455 in master (cluster/afr: Don't do memcpy of afr_local_t in afr_local_copy.)
Comment 12 Anand Avati 2009-12-04 12:06:58 UTC 
PATCH: http://patches.gluster.com/patch/2554 in master (posix_rchecksum() - fix leak of read buffer)
Comment 13 Anand Avati 2009-12-05 05:49:28 UTC 
PATCH: http://patches.gluster.com/patch/2571 in master (cluster/afr: Fix inode_ref's for local->cont.lookup.inode)
Comment 14 Anand Avati 2009-12-28 09:39:23 UTC 
PATCH: http://patches.gluster.com/patch/2635 in master (cluster/afr: Report number of blocks healed during diff self-heal.)
Comment 15 Anand Avati 2009-12-28 09:39:27 UTC 
PATCH: http://patches.gluster.com/patch/2640 in master (cluster/afr: Allocate diff algorithm loop_state structures only once.)
Comment 16 Anand Avati 2009-12-28 09:39:30 UTC 
PATCH: http://patches.gluster.com/patch/2637 in master (storage/posix: Zero out the checksum array in rchecksum.)
Comment 17 Vikas Gorur 2010-01-12 06:10:37 UTC 
*** Bug 457 has been marked as a duplicate of this bug. ***
Note You need to log in before you can comment on or make changes to this bug.