Category: Database

by A red alert was issued by production monitoring system saying that several managed servers are in SHUTDOWN state and the Admin Server console as well echoed the same.

I did not observe any abnormalities in the Admin Server thread dump, however, could see threads with following exact stack trace, equivalent to the number of Managed Servers that the Admin Server is showing as ‘RUNNING’. These threads are spawn per Managed Server, and the number of such threads confirmed that the Admin Server is no longer interested in the ‘crashed’ Managed Server.

Admin server indicated that the managed servers are in SHUTDOWN state was due to the default thread pool exhaustion in managed server – caused by the WebLogic bug BUG:27724050.

Related Articles