What we have found, is that when the OC sessions start experiencing a "alarm retrieve errors" in the console, we are getting the following errors in ems and alarmviewer-api logs.
2022-11-16 13:55:09.413 ERROR [AlarmServiceImpl.getAlarmsWithoutMatchesFilter:1143] Error in getAlarmsFromDB : com.firehunter.ump.exceptions.DataFactoryException: Cannot get a connection, pool error Timeout waiting for idle object...
Release : 20.4 up to CU7
exhausted connection pool to database