Transit - Locks (java.util.concurrent.locks.ReentrantLock)

95 views
Skip to first unread message

Andy Jarrett

unread,
Oct 31, 2013, 6:25:53 AM10/31/13
to fusion...@googlegroups.com
This morning, all of a sudden, I started seeing Lock issues stating that it would not be killed. 

The problem seems to be happening around the query to a MSSQL DB, does anyone know what this actually is or how to better understand the problem?


Kind regards,

Andy

David Stockton

unread,
Oct 31, 2013, 8:07:12 AM10/31/13
to fusion...@googlegroups.com
Hello Andy,

Thank you for your email.

This is FusionReactor refusing to kill a thread that holds an "ownable synchronizer". Java has the ability to lock on objects, not just threads. The thread trying to be killed holds one of these and therefore killing it could result in another (waiting) thread being stuck forever. To attempt to protect against this, FusionReactor will refuse to kill such a thread.

We have a technote giving further details you may wish to read - http://www.fusion-reactor.com/support/kb/FRS-310/

It's likely to be a DB connection related object the thread holds. Therefore you probably want to tackle the slow query problem. You could do so by using FusionReactor to see what query is running and your MS SQL tooling to find out why that query runs slowly (eg bad indexes/execution plan, etc).

If you'd like some assistance from our consulting team to investigate further, please let us know at sa...@cfconsultant.com

Best regards,
David Stockton
Fusion Team
Reply all
Reply to author
Forward
0 new messages