Bug 798198 - New brick not used after add-brick
Summary: New brick not used after add-brick
Keywords:
Status: CLOSED DUPLICATE of bug 796116
Alias: None
Product: GlusterFS
Classification: Community
Component: replicate
Version: mainline
Hardware: Unspecified
OS: Linux
unspecified
high
Target Milestone: ---
Assignee: Pranith Kumar K
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2012-02-28 09:55 UTC by Kaushal
Modified: 2012-02-28 11:33 UTC (History)
1 user (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2012-02-28 11:33:03 UTC
Regression: ---
Mount Type: fuse
Documentation: ---
CRM:
Verified Versions:
Embargoed:


Attachments (Terms of Use)
Dump of mount process (8.28 KB, application/octet-stream)
2012-02-28 09:58 UTC, Kaushal
no flags Details
Dump of self-heal daemon (5.21 KB, application/octet-stream)
2012-02-28 09:59 UTC, Kaushal
no flags Details
Mount log (7.29 KB, text/x-log)
2012-02-28 10:00 UTC, Kaushal
no flags Details
Self-heal daemon log (13.50 KB, text/x-log)
2012-02-28 10:01 UTC, Kaushal
no flags Details

Description Kaushal 2012-02-28 09:55:11 UTC
Description of problem:
Adding a new brick and changing volume to a replicate type with replica 2 to a single brick volume, doesn't cause the new brick to be used. No data gets written to the new brick, already existing files/directories don't get self-healed either by find-stat on the mount-point or by self-heal daemon. The self-heal daemon does heal when "volume heal <vol> full" command is used. But even after that new data isn't written to the new brick.
Only after remounting the volume and restarting the self-heal daemon, the normal operation is restored.

How reproducible:
Always

Steps to Reproduce:
1. create, start and mount a volume with single brick
2. Do some operations on mount point
3. add new brick, #gluster volume add <vol> replica 2 <new-brick>
4. Do more operations on mount point.
  
Actual results:
Older data is not automatically self-healed to the new brick and new data is not replicated to the new brick, until self-heal daemon is restarted and volume is remounted.

Expected results:
Self-heal should happen and new data should be replicated.


Additional info:

Comment 1 Kaushal 2012-02-28 09:58:55 UTC
Created attachment 566274 [details]
Dump of mount process

Comment 2 Kaushal 2012-02-28 09:59:31 UTC
Created attachment 566275 [details]
Dump of self-heal daemon

Comment 3 Kaushal 2012-02-28 10:00:47 UTC
Created attachment 566276 [details]
Mount log

Comment 4 Kaushal 2012-02-28 10:01:16 UTC
Created attachment 566277 [details]
Self-heal daemon log

Comment 5 Kaushal 2012-02-28 11:33:03 UTC

*** This bug has been marked as a duplicate of bug 796116 ***


Note You need to log in before you can comment on or make changes to this bug.