Elastic Search stores threads for long periods of time
Pega Elastic Search stores threads for long periods of time which causes application issues when searching for new cases. The overall thread pool operates at 90% capacity on Pega servers.
"WebContainer : 5" J9VMThread:0x0000000042FBB400, j9thread_t:0x000001002BD1C7A0, java/lang/Thread:0x000000064751A2C8, state:P, prio=5
(java/lang/Thread getId:0x486, isDaemon:true)
(native thread ID:0x7D40053, native priority:0x5, native policy:UNKNOWN, vmstate:P, vm thread flags:0x01020001)
CPU usage total: 1895.813914000 secs, user: 1850.702885000 secs, system: 45.111029000 secs
Parked on: com/pega/elasticsearch/common/util/concurrent/BaseFuture[email protected]0x000000065E9102E0 Owned by: <unknown>
"PegaRULES-Search[riverClusterService#updateTask][T#1]" J9VMThread:0x0000000042F9CD00, j9thread_t:0x0000010023EDCEA0, java/lang/Thread:0x0000000639E7BD50, state:P, prio=5
(java/lang/Thread getId:0x206, isDaemon:true)
(native thread ID:0x3FF00F5, native priority:0x5, native policy:UNKNOWN, vmstate:P, vm thread flags:0x01020001)
CPU usage total: 0.000985000 secs, user: 0.000259000 secs, system: 0.000726000 secs
Parked on: java/util/concurrent/locks/AbstractQueuedSynchronizer$ConditionObject@0x0000000639F0ED08 Owned by: <unknown>
Steps to Reproduce
A defect or configuration issue in the operating environment
Low availability of file handles that process indexing requests affected all Hazelcast communication.
ResolutionMake the following change to the operating environment:
Set the ULIMIT to 65535 and refer to IBM Websphere ULIMIT recommendations for further explanations.