1 Reply Latest reply on Nov 8, 2006 11:20 AM by Jim Collins

    agentinfo timeout

      I am running into the problem where a server will become hosed up and although the server is pingable, all resource consuming services (ssh, bl agent, etc) hang with any attempted connection. This is caused by a number of things like max filehandles consumed or all swap space consumed. By the way, this is on RHEL 3. What I do is run agentinfo across the whole server environment using the output to dynamically update the IS_ONLINE property. When agentinfo gets to the hosed-up server it hangs indefinitely. Is there any way around this? Possibly a timeout value for agentinfo or a timeout wrapper that someone has come up with for agentinfo.

        • 1. Re: agentinfo timeout



          Given the age of your post you likely found a way around the lack of timeouts for NSH commands and related tools (agentinfo, etc.)... However for those who have not, here is a shell solution I've recently used at Putnam as part of a test to alert us when the RSCD agent on our App server(s) drops into a "reachable by unresponsive" state which would result in "agentinfo" test hanging forever.


          The shell logic to deal with a command which may never return could easily be modified for your case and any other case where a network shell command is potentially applied against a broken RSCD which may result in the command hanging forever.


          - Jim Collins