Sporadic system load peaks on mail servers using Veritas Cluster File System version 5.0MP4HF2
Issue
-
Cluster of two nodes running RHEL5U4 and Veritas Cluster File System (5.0MP4HF2), used as a pop3/imap/smtp mail server with simultaneuous read/write access from both nodes on a shared veritas cluster filesystem.
-
The load average on both systems when operating normally is below 5.
-
At random times, the System Load average rises unexpectidly on one or both nodes, and the cause of the load spike cannot be determined.
-
Sometimes it rises on one node to something in the order of 100 or 200, while it remains low on the other node. After a few minutes it dissapears and no action is required.
-
Sometimes it rises on both nodes to something in the order of 1000. It does not go away after a few minites and the only way to fix it is to reboot both nodes simultaneously
-
-
A case has been opened with Symantec.
Environment
- Red Hat Enterprise Linux 5.4
- Veritas Cluster File System (5.0MP4HF2)
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.