Production instance going down repeatedly
Server is repeatedly going down due to Out Of Memory issues.
Outof Memory exceptions in the logs
Steps to Reproduce
The system is woefully low on memory. The 7.1 Installation Guide specifies “PRPC web applications require a minimum of 2GB of memory to operate.” For production systems, Pega recommendation a 4GB minimum for max heap size.
Evidence of problems in the alert log:
Pega0028 – (most prevalent alert in the alert log) system unable to recover heap space, occurs 200 times in node1 alert log. This is further indication of the shortage of memory.
Pega0005 – database interactions over 500ms, these are fairly high in number as well. This would help drive overall slowness, but wouldn’t necessarily contribute to the memory issues experienced.
Pega0035 – clipboard lists >10,000 - “The number of elements in this clipboard list property has exceeded WARN level. Maximum size: 10000 Property reference: temp1.pxResults”
Recommended increase the Min and Max heap settings to 4GB. This will bring system inline with standard recommended heap sizes and alleviate the OutOfMemory issues.
0% found this useful