Bug 762526 (GLUSTER-794)

Summary: Self-heal truncates files or file size remains zero
Product: [Community] GlusterFS Reporter: Chida <chida>
Component: replicateAssignee: Anand Avati <aavati>
Status: CLOSED WORKSFORME QA Contact:
Severity: medium Docs Contact:
Priority: low    
Version: 3.0.3CC: amarts, chrisw, gluster-bugs, vijay
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Chida 2010-03-31 11:38:03 UTC
when a file is created one one server when both servers are up, file will replicate successfully across both servers. 

When one server is down, a file is added to the server that is still up, and then the server the failed is brought back up. At this point the newly created file will not auto heal, nor a self heal is possible using ls -lR. It knows that the file should be there, but the file size remains 0 or is half the size of the original file.

Comment 1 Vikas Gorur 2010-04-13 15:36:37 UTC
Do you have stat-prefetch loaded? Self-heal will not be triggered by doing an ls -lR in that case. Can you do "find | xargs stat" and see if the file still doesn't heal?

Did the file start healing and stop at half the size? How big was this file? Are you sure you had enough diskspace on the backend for the whole file?

Comment 2 Amar Tumballi 2010-04-20 08:01:53 UTC
Vikas, Any progress on this?

Comment 3 Amar Tumballi 2010-10-05 06:01:12 UTC
Most of the self-heal (replicate related) bugs are now fixed with 3.1.0 branch. As we are just week behind the GA release time.. we would like you to test the particular bug in 3.1.0RC releases, and let us know if its fixed.

Comment 4 Amar Tumballi 2010-10-06 09:36:58 UTC
Avati, can you see if these bugs are valid anymore? if not valid.. please close.