> We have a clustered environment, so I think I'll setup a cronjob to run that 
> sql line.

Running that line at any other time than an orphaned lock condition
runs the risk of damaging lock state in such a way that concurrent
execution of cleanup may occur, which is the whole purpose of
database-backed locking strategy.  I would strongly advise against
this.  We've been running this code in our production environment for
months now and have had zero problems other than redeployments, where
it's reasonable to execute that SQL statement as part of post-deploy
process.  I'd encourage investigating how the database got in this
state in the first place before attempting other workarounds.

> Just one more thing.. When this "error" occurs is it normal that CAS returns 
> a error message to the users (not even showing the authentication page)?

It's strictly informational, and it is logged at INFO to help clarify
that's it part of normal operation.  Ticket cleanup has absolutely no
impact on CAS availability, so if you're getting error messages on the
front end when that occurs, there's something else going on.

M

-- 
You are currently subscribed to [email protected] as: 
[email protected]
To unsubscribe, change settings or access archives, see 
http://www.ja-sig.org/wiki/display/JSG/cas-user

Reply via email to