POA incorrectly reporting disk limit is reached
I started to receive e-mails from the one of our Post Office Agents stating that the disk limit was reached. This caused the Disk Space Threshold To Trigger a reduce Action twice over the past 3 days. It only recovered 3-7 mb because the weekly maintenance just finished running.
The problem is not that the POA is running low on disk space, because the 'Available Disk Space' reported by the POA WebConsole is 30 GB, which is 25% of the total disk.
The Default POA Disk Check Event has its trigger set to 2048MB. We monitor this system with cacti, nagios, and GW monitor. GWMON reported the disk below its threshold once but didn't show it on the trend report's graph for the poaAvailDiskSpace.
This GW 7.0.3 (03/09/2008) linux POA only recently started to report low disk space after it was moved from local storage to an EMC iscsi san. This was our first GW linux po. It's using ext3 because that was the general consensus of the GW experts I spoke with at BrainShare that year. I plan to move it to either Reiser or NSS in the near future.
Only one other agent, an MTA using reiserfs on an Equallogic SAN, reports this same incorrect information through GWMON. All the other GW agents, WebAccess and GWIAs are also running on linux using reiserfs on Equallogic.
Any ideas?
|
Threaded Mode
