In case of performance issue this, threadDump and other threadStatistics are often useful and we want to share the ways of collecting threadDump from NAC during such time and share the same with support for analysis.
Release : 6.6, 6.7
Component : CA RELEASE AUTOMATION CORE
Collecting Thread Statistics to troubleshoot performance issue. Please collect below data (whichever applicable) and share with support while opening Technical assistance ticket.
Thread dump collection (Optional Linux/Unix)
1: SSH into the host 2: Find the NAC’s process id - ps auxww | grep catalina | grep -v UpdateService | grep -v grep 3: Run the thread dump loop: for a in $(seq 1 30); do echo dump $a/30; kill -3 NAC_PID; sleep 5; done where NAC_PID is the process id found in #2
Thread statistics via JMX
1: Navigate and login to http://NAC-IP:20203/mbean?objectname=noliocenter%3Atype%3DcentralExecutor (Please change NAC-IP with respective hostname/IP of your NAC). 2: Please invoke switchMonitorOn which will turn on monitoring 3: Now periodically invoke getActiveTasksTable and capture the result shown on page (manually) 4: Invoke switchMonitorOff when monitoring of the queue is no longer required.
1: Navigate and login to http://NAC-IP:20203/mbean?objectname=java.lang%3Atype%3DThreading (Please change NAC-IP with respective hostname/IP of your NAC). 2: Please invoke findDeadlockedThreads which will turn on monitoring 3: Capture the result shown on page (manually)