I increased the log-level to debug, and found that exactly at the time
the thread counts increase (and become more inconsistent), there was the
message:
2002-09-23 18:13:10,015 DEBUG
[org.jboss.resource.connectionmanager.IdleRemover] run: IdleRemover
notifying pools, interval: 450000
2002-09-23 18:13:10,015 DEBUG
[org.jboss.resource.adapter.jdbc.local.LocalManagedConnectionFactory]
Using properties: {user=sa, password=}
(last line repeats 10 times)
This might have something to do with our own JMSProvider, which uses
JDBC as its persistence layer.
Cause it has its own XA implementation it uses a db-connection not
enlisted under under the tx-mgr.
This non-transacted db-connection comes out of a NoTxDatasource which I
created via the XSLSubDeployer.
I suspect that I did something wrong (too much copy/paste?) when
modifying the XSL-code from LocalTX.
(This is the modification/addition I have commited to cvs some weeks ago).
There is no thread which our JMSProvider directly creates at that moment
in time, so I think I triggered a bug
in the jvm (non-started threads, see previous posting) or a questionable
use of such threads by the pool
or by the jdbc-driver.
I experience this with oracle-thin _and_ hypersonic which provides
evidence against a jdbc issue.
This might be a job for an expert in jca and pooling issues (David?).
Thank you,
Michael Bartmann
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development
