Bug 433950

Summary: conga hangs when probing storage on xen cluster nodes.
Product: Red Hat Enterprise Linux 5 Reporter: Stuart Sears <ssears>
Component: congaAssignee: Jim Parsons <jparsons>
Status: CLOSED DUPLICATE QA Contact:
Severity: medium Docs Contact:
Priority: high    
Version: 5.1CC: cluster-maint
Target Milestone: rc   
Target Release: ---   
Hardware: i386   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2008-03-25 01:41:49 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Stuart Sears 2008-02-22 10:27:14 UTC
Description of problem:
conga hangs probing storage on xen cluster nodes.

Version-Release number of selected component (if applicable):
0.10.0-6.el5

How reproducible:
every time!

Steps to Reproduce:
1. create a xen cluster, where the nodes have a primary disk called xvda
2. install conga (and the appropriate additional packages)
3. use the luci interface to probe storage on one of the cluster nodes.
  
Actual results:
a large number of parted processes on the cluster node, which hang while trying
to probe the storage devices on that node.

Expected results:
luci displays the storage devices attached to that particular node.

Additional info:
This seems to be related to the device names used by the nodes. It may actually
be a parted problem (or the way we use parted for this). If the disks used by
the xen machines are renamed to 'hd*' (or, I assume sd* - didn't test this) the
storage probe returns immediately and works properly.

Comment 1 Stuart Sears 2008-02-22 10:33:02 UTC
After some more searching, it may well be #366971 causing this. Seems likely.

Comment 2 Ryan McCabe 2008-03-25 01:41:49 UTC
A bug in parted is causing the hang, but conga should handle that kind of
behavior better. Duping into the RFE to support multiple concurrent tasks on a
busy entity, if possible.

*** This bug has been marked as a duplicate of 238935 ***