I have not seen it myself but essentially a "restart" is a combination of a "stop" followed by a "start". The question then is "which of the two is failing?". If the "stop" not working causing the "start" to fail or is the "stop" working and there is a specific "start" problem you are seeing.
Can you also give more details on how the "restart" fails.
The full details are in case 00005844. I was wondering if this problem was exclusive to us, or if others have seen this problem as well.
Here is the error output if that helps...
% nexec hpux.host.com agentclt restart
agentctl: Warning - The RSCD agent is currently not running.
rscd - Copyright (C) 1996-2004, BladeLogic, Inc.
bin/rscd: Address already in use
Hi Todd. From working with Sean on this issue this morning over webex, we have narrowed the problem down a little. The stop piece of the agentctl restart is not working because it does not realize that the agent is actually running. It therefore does not attempt to stop the agent and the subsequent start piece of the restart fails with the bind exception you are seeing. We are investigating this here using output from Sean. Looks like a problem while examining the running processes on the box. Strangely, we cannot reproduce it here on HP-UX so it is going to take a little work in nailing it down. John
Message was edited by: John O'Toole
In case anyone runs into this issue and tries to follow this thread for help. We tracked down the issue to a problem with one of our libraries on 64-bit HP machines. The problem has been addressed in an upcoming release.