thanks for reading this urgent thread.
Our client just migrated to production in a high availability setup (2 load balanced servers), our standard procedure was used: a tar of the whole test environment webMethods6 folder (this has worked for all others up to now).
When starting the prod server, there were problems with the import in TN, however we resolved this, and everything seemed to work fine.
All of a sudden, the JVM crashed, nearly at the same time on both machines, with the returncode 134. Since then, this has been happening (it seems) at random.
It looks like this happens just after IS starts cleaning connections in the repository.
In the high availability test environment, everything has been working fine for months, and still is.
Also, the production servers host another IS, which has been working for nearly a year now, without any similar problems.
HW: Sun Fire 480
OS: Solaris 5.8
Kernel: Generic_108528 Patch 23
MW: MQ Series 5.3
DB: Oracle 8i
Has anybody had the same or similar problem? Any ideas?
Also, does anybody know how to clean the repo somehow?
I have been told that deleting repo4 was no problem, and if you do the same for repo2, you will mereley need to recreate the MQ handlers?
Can someone confirm this?
Any suggestions are welcome!!