Shai Berger writes:
Do you have a DCOracle version later than 1.3.2? No, we use DCOracle 1.3.2
We use 1.3.2; we have increased the cache from 400 (the default) to 1000, to no avail. Our ZODB size, last I checked, was ~40M. I remember you mentioning that yours was less then 10M, I think; how much cache do you use? I have to check at work, but I think it was about 10 % of the objects in the ZODB with the time set to 30 minutes.
We do know what caused our zombie processes: a bug in the Oracle Java runtime library: ... Zope starts Oracle processes from Java??? No, it does not. Apparently, your Zombie processes have a different cause then ours. Our Zombies have not been Zope children but Oracle children, spawned by Oracle stored procedures implemented in Java.
Sorry, if I responded to a message in a thread without a clear understanding of the basic problem. Dieter