On Sep 22, 2009, at 12:44, Scott Ferguson wrote:

> On Sep 22, 2009, at 11:51 AM, Rob Lockstone wrote:
>
>> I've never seen this before and can't find this string in the Resin
>> (non-Pro) source. This happened on several of our server this  
>> morning.
>> But I've never seen it happen before in months of running this
>> particular configuration.
>>
>> Environment:
>>
>> Windows Server 2003
>> Resin Pro 3.1.9 - 100 Server License
>> JDK 1.6_13
>>
>> What would cause this?
>
> Certain types of deadlock or a true JVM or thread freeze.  It's
> detected by the <ping> service.

Thanks, Scott. The fact that it was a result of the ping service got  
me on the right path.

Rob

>
> Resin has an timer thread in Alarm that returns the internal current
> time for Resin.  If that thread freezes, Resin's internal time won't
> be updated and will become out of sync with the
> System.currentTimeMillis().  If that discrepancy gets very large
> (15m), Resin will force an exit.  Basically, it's an indication that
> something is very, very wrong with the JVM.
>
> Since I just updated the shutdown logic for 4.0.2, it might be
> possible for Resin to give a thread dump before exiting in that case.
> Previously, Resin just does a halt() because the JVM might be in such
> a bad state that any additional work might create new problems.
>
> -- Scott



_______________________________________________
resin-interest mailing list
resin-interest@caucho.com
http://maillist.caucho.com/mailman/listinfo/resin-interest

Reply via email to