Description of problem:
Rgmanager has a per-service operational queues. In order to stop critical
operations (stop, start) from being held off for too long, there is supposed to
be a max of 1 'status' operation either in the queue or currently in progress
per service at any time. This prevents long status checks from getting queued
over and over.
For example, if you have a 500 export NFS service, it can take literally 10
minutes to check the service. This allows 10*10, or 100 status checks to be
queued up. If you try to then reconfigure the service, it could wait for up to
all of those status checks to complete prior to the reconfiguration event
Created attachment 158472 [details]
Makes the status check have a max of 1
That is, if you have 500 exports and status checks are queued every 10 seconds
(whoops!), during the course of 1 status check, 10*6 or 60 status checks can be
queued up and waiting for the first status check to complete.
This will pause reconfiguration indefinitely.
Patch in RHEL4 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.