Skip to main content

i found bug is caused by locking timeout

6 replies [Last post]
nijie8
Offline
Joined: 2005-06-09
Points: 0

hi:
i test mobicents with sipp and got below error.
activitycontext locking timeout cause system failure. Is this really a bottleneck ?

2005-07-25 16:03:16,936 246630 ERROR [org.jboss.cache.lock.IdentityLock] (Thread-1022 EventTypeID[javax.sip.message.Request.INVITE#javax.sip#1.1], #16:) read lock for /services:JAIN SIP Proxy Service-NIST-1.0 could not be acquired by :1361 after 15000 ms. Locks: Read lock owners: [:1355]
Write lock owner: null
, lock info: read owners=[:1355] (org.jboss.cache.lock.LockStrategySerializable@c07125)

Reply viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
mranga
Offline
Joined: 2003-06-06
Points: 0

Can you give it a try now. This problem has likely been resolved.

Ranga

nijie8
Offline
Joined: 2005-06-09
Points: 0

it seems the problem is gone,but in fact the performance is still not so good.

nijie8
Offline
Joined: 2005-06-09
Points: 0

i get 3 caps at the begin time. then after lots of call it drop to very low . i think there might be some memory leaks.

fram
Offline
Joined: 2004-05-13
Points: 0

I think you are right there are memory leaks, I had the same result during my performance tests. Ranga and I are going to work on fix some of those today.
Thank you for your feedback,
Francesco

ivelin
Offline
Joined: 2003-07-13
Points: 0

Looks like a bottleneck or deadlock. Can you create a new issue and outline the steps to reproduce the problem?
https://mobicents.dev.java.net/servlets/ProjectIssues

Ivelin

mranga
Offline
Joined: 2003-06-06
Points: 0

> Looks like a bottleneck or deadlock. Can you create a
> new issue and outline the steps to reproduce the
> problem?
> https://mobicents.dev.java.net/servlets/ProjectIssues
>
> Ivelin

This is not necessarily an error. We are setup to rollback and try redelivering the event. However, it would be helpful if we put an additional pair of eyes on these kinds of issues. Its been just Francesco and I hacking on this for a while now and I'd like to pull in some other review. Do take a look at it and let us have your substantive comments if you have some time.

I need to detail this further however, note that Serializablity does not imply a serializable consistency model for the cache. We are working on loosening up the cache consistency while still presenting a serializable consistency model for the data structures of the SLEE.

Ranga.