Bug 765258 (GLUSTER-3526) - in nfs, socket write failures not handled properly.
Summary: in nfs, socket write failures not handled properly.
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: GLUSTER-3526
Product: GlusterFS
Classification: Community
Component: nfs
Version: 3.2.3
Hardware: x86_64
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Krishna Srinivas
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2011-09-08 05:20 UTC by Amar Tumballi
Modified: 2013-12-19 00:06 UTC (History)
4 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed:
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions: 3.2.5qa6
Embargoed:

Attachments (Terms of Use)

Description Amar Tumballi 2011-09-08 05:20:41 UTC 
Noticed that in nfs (as of release-3.2 branch, please validate this with master), if while sending a buffer if there is a error from socket, the buffer is never removed from queue, and hence will be in a loop trying to send it.
Comment 1 Anand Avati 2011-10-10 08:03:19 UTC 
CHANGE: http://review.gluster.com/400 (Change-Id: If8ba73ac57bc235a28c1e1cdc7a5e11947e7a52f) merged in release-3.2 by Vijay Bellur (vijay)
Note You need to log in before you can comment on or make changes to this bug.