Red Hat Bugzilla – Bug 171034
Missing: Monitoring for local and cluster file systems in resource agents
Last modified: 2009-04-16 16:18:13 EDT
Description of problem:
The file system resource agent, fs.sh, talks about read/write tests to/from file
systems at OCF_CHECK_LEVEL 20 in its meta-data. This is not currently done, but
is necessary in order to detect critical problems like loss of connectivity to
the shared block devices on which the file system resides.
As a side note, an option to use active monitoring of access to the file system
(e.g. constant monitoring by a separate daemon) would be useful for customers
who which to detect file system access problems in a timely manner (down to 2-5
This functionality should minimally be implemented for the <fs> resource type,
but probably for <clusterfs> as well.
Fixes for the <fs> resource agent are in CVS; these need to be ported to the
<clusterfs> resource agent.
Fix in CVS.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.