wadie
Well Known Member
Hi list,
I have a client that is bound to go-live 1 Feb with a new installation (E1 9.0, TR 8.98.2, Red Stack, all windows servers)
I have a strange happening the past few weeks during CRP. Some of my CallObject kernels get 'stuck' at 6% CPU. This is the third time I catch some of them 'in the act'. First time was because a package deployment failed. In their logs, I can see this:
392/4636SYS
ispatch Fri Jan 15 01:28:22.937000Jdekdisp.c2010KNT0000099 - JDEK_DispatchCallObjectMessage: Pooled Connection: 06330008 timed out!
392/4636SYSispatch Fri Jan 15 01:28:22.937001Jdekdisp.c2034KNT0000199 - JDEK_DispatchCallObjectMessage: Pooled Connection: 06330008 still in use, can not be freed!
Since there are other log messages behind that of users using the kernel, I think it's not dead, but the hogging of CPU power worries me. Only thing I can do to stop the behaviour is kill it in task manager and remove the zombie. (after this, the failed package deployment was successful BTW)
In the logs, the time of the messages changes (morning, evening, night) so I don't think any user actions provoke this.
Has anyone else seen this? Probably a setup issue somewhere... (OAS? Oracle DB? JDE.INI? JDBJ?)
If anyone can point me in the right direction, feel free to do so
I have a client that is bound to go-live 1 Feb with a new installation (E1 9.0, TR 8.98.2, Red Stack, all windows servers)
I have a strange happening the past few weeks during CRP. Some of my CallObject kernels get 'stuck' at 6% CPU. This is the third time I catch some of them 'in the act'. First time was because a package deployment failed. In their logs, I can see this:
392/4636SYS
ispatch Fri Jan 15 01:28:22.937000Jdekdisp.c2010KNT0000099 - JDEK_DispatchCallObjectMessage: Pooled Connection: 06330008 timed out!392/4636SYS
ispatch Fri Jan 15 01:28:22.937001Jdekdisp.c2034KNT0000199 - JDEK_DispatchCallObjectMessage: Pooled Connection: 06330008 still in use, can not be freed!Since there are other log messages behind that of users using the kernel, I think it's not dead, but the hogging of CPU power worries me. Only thing I can do to stop the behaviour is kill it in task manager and remove the zombie. (after this, the failed package deployment was successful BTW)
In the logs, the time of the messages changes (morning, evening, night) so I don't think any user actions provoke this.
Has anyone else seen this? Probably a setup issue somewhere... (OAS? Oracle DB? JDE.INI? JDBJ?)
If anyone can point me in the right direction, feel free to do so
