JBOSS 3.2.3 with Linux Red hat 9 and java 1.4.2...

[mhemling]

I am noticing that under peak loads that the server begins throwing exceptions and my queue receivers keep losing their connections. My code tries to reconnect. After a while everthing freezes and there are Out of memory errors. The same exact code seems to work fine under jboss 3.2.1.

Can someone please explain export LD_ASSUME_KERNEL=2.4.1.

What does this do? Are there known problems with Redhat 9??

thanks

[genman]

Several things are possible:

1. JDK+Linux kernel combo broken. Not likely but not impossible. I use RedHat 9 and I would suggest downloading the JDK 1.4.1 from: http://www.blackdown.org/ which is compiled for RedHat 9.

2. JBoss 3.2.3 broken. Could be, but I haven't noticed on Solaris. More likely is that you haven't ported the same configuration from JBoss 3.2.1 to 3.2.3. The cache or persistence configuration might not be the same.

3. Something else changed.

You need to indentify all the variables that have changed, post the Exception trace and logs, and provide example code for someone to reproduce.

[adrian.brock]

Redhat's backport of NPTL to 2.4 for Redhat 9
doesn't work correctly with java.

I find using the old threading model (activated using the export above)
fixes the problems.

Regards,
Adrian

[mwaltz1]

When you refer not "work[ing] correctly with java" do you mean the Sun vm in particular or all vms? The prior post mentions Blackdown -- has there been a better experience with Blackdown?

Thx.

[genman]

I haven't seen problems with Redhat 9 and blackdown, using JMS and Oracle and lots of MDBs and threads for the last 4-5 months. But your milage may vary.