[OpenAFS] Working with 3 vlservers, I purposely disabled one to test queries take longer with afs01 removed

Jeffrey Altman jaltman@secure-endpoints.com
Tue, 18 Dec 2007 09:54:45 -0500

Jerry Normandin wrote:
> Yes that is what I am seeing.  I was wondering if there was a way to
> tweek the timeout.

You can modify the Rx timeout but it will affect every operation that
the cache manager performs.

The timeout should be experienced at most once per client every two
hours when the vldb server information is refreshed.  After the vldb
server is marked down it should stay down until a probe indicates it has
come back up.

If you are seeing issues on the clients more frequently than that it
would indicate a bug.