Hello Community,
We recently upgraded from Campus solution image 6(PT 8.55 in RHEL 5) to image 31(PT 8.60.10 in RHEL 8.9). Since we went live, we are seeing significant performance drop during peak hours as the RAM growing all the way to 164 GB and hung the VM. Our infra, 64 core and 164 GB RAM in RHEL 8.9. this server has 4 LARGE app server domains(only).
[PSAPPSRV]
Min Instances=20
Max Instances=20
Service Timeout=300
Recycle Count=10000
ProcessRestartMemoryLimit=0
Allowed Consec Service Failures=2
Max Fetch Size=50000
Auto Select Prompts=1
Tuxedo Queue Size=65535
The scenario we are facing is, when we start the services, the RAM comes to 12GB and over the period(4 days to a week) it grows to 160GB and requires a reboot to come back to normal stage.
I have checked with Redhat support and they conclude either increase RAM for all this PSAPPSRV instance or check with oracle support.
My experience with previous PT(8.55) and RHEL 5 was, RAM used to goes up, down and show variations. but this is not the same case with upgraded env.