Hello,

From time to time we are getting deadlock exceptions. Any help is appreciated.

Thanks,
Alex

Hibernate version:
3.1.1

c3p0 version:
0.9.0.4

c3p0 properties:
hibernate.c3p0.maxPoolSize=20
hibernate.c3p0.minPoolSize=2
hibernate.c3p0.maxStatements=100
hibernate.c3p0.acquireIncrement=5
hibernate.c3p0.maxIdleTime=5000

Full stack trace of any exception that occurs:
2006-02-12 12:34:23,671 [Thread-3] WARN ThreadPoolAsynchronousRunner$DeadlockDetector - com.mchange.v2.async.ThreadPoolAsynchronousRunner$DeadlockDetector@430440 -- APPARENT DEADLOCK!!! Creating emergency threads for unassigned pending tasks!

2006-02-12 12:34:23,671 [Thread-3] WARN ThreadPoolAsynchronousRunner$DeadlockDetector - com.mchange.v2.async.ThreadPoolAsynchronousRunner$DeadlockDetector@430440 -- APPARENT DEADLOCK!!! Complete Status: [num_managed_threads: 3, num_active: 3; activeTasks: com.mchange.v2.c3p0.stmt.GooGooStatementCache$1@19ed6fd (com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread-#2), com.mchange.v2.c3p0.stmt.GooGooStatementCache$1@134861b (com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread-#1), com.mchange.v2.c3p0.stmt.GooGooStatementCache$1@15067a4 (com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread-#0); pendingTasks: com.mchange.v2.c3p0.stmt.GooGooStatementCache$3@1c48148, com.mchange.v2.resourcepool.BasicResourcePool$6@3db66, com.mchange.v2.resourcepool.BasicResourcePool$6@139089f]

Name and version of the database you are using:
MsSql 2000 with DataDirect JDBC driver.

Can you get a Thread dump ? Both the threads listed are called "Thread-3" and look like the same thread.

Usualy in a deadlock we are looking for 2 or more threads that have taken opposing locks and waiting for each other, just bad design by not taking multiple locks in the same order at runtime.

If you are using the Sun JVM from Unix maybe a "kill -QUIT pid" will make the JVM dump its thread state to stdout/stderr and from that you can post the useful parts of that as well.

Hi,

Please specify the version of C3P0 you are using.

It is written in description - 0.9.0.4

Alex

Hi,

Correct, I didn't notice in your first post.

Try using these settings.

c3p0.acquireRetryDelay=1000
c3p0.acquireRetryAttempts=60
c3p0.breakAfterAcquireFailure=false
c3p0.acquire_increment=1
c3p0.idle_test_period=100
c3p0.minPoolSize=3
c3p0.maxPoolSize=20
c3p0.maxStatements=0
c3p0.maxStatementsPerConnection=0
c3p0.timeout=100

Also paste the code of your hibernate factory. I had the same problem, now I'm just trying to remember what was the solution.

Sounds like the problem is in number of pooled statements.

After setting c3p0.maxStatements=600 I do not see exceptions.

Alex

Deadlocks should never happen, thats a design error. So if you can confirm that

c3p0.maxStatements=<low_value>

can cause deadlocks, I'm sure the c3p0 maintainer might be interested to hear.

I do confirm that when this setting is set to 100. We experienced a deadlock. Unfortunately the system is quite complex and I cannot post a tiny piece of code that reproduce this.

Alex

I had a similar deadlock issue and it appeared to be a bug in c3p0 0.9.0.4.

From the c3p0 change log for release c3p0-0.9.1-pre2

-- Fixed a deadlock which could occur if a pool is closed at the same time as
one of its Connections is closed. One thread gets the pool's lock, then tries
to get a PooledConnections lock to close it. An ordinary client closes a
Connection, which acquires the PooledConnection's lock, and then tries to
check back in to the pool, requiring the pool's lock. Broke the deadlock by
making the destruction of resources on pool close asynchronous, by a Thread
without the resource pool's lock.. [Many thanks to Roger Westerlund for
sending a full thread-stack dump capturing this deadlock!]

so I updated teh c3p0 version and the issue went away.