3 Replies Latest reply on May 4, 2011 4:43 PM by Bill Robinson

    Windows Patch Analysis job timeout

    Jim Campbell

      We're encountering a problem where Windows Patch Analysis jobs hang on a small number of servers and never complete.  Finding these is difficult as they show up with a green checkmark in the logs.  Aborting the analysis job doesn't work - it seems to buffer analysis results and flush them 100 at a time.  Anything 'buffered' at the time the job is aborted gets lost and we can't manually create remediation jobs from these results (only the sets of 100 that were flushed previously).  I.e. if we have 380 results and one or two servers hang, aborting the job causes us to lose the last 80 results.  If the job actually completes on all targets whether they fail or not, the buffer gets flushed and we get all of the results.  We can eventually fix this by combing through the logs for each server to find the one that is still running and turn off its rscd agent to force it to fail, but I'm assuming there should be some way to get it to automatically abort on a target after a certain period of time.


      I've tried setting the JOB_PART_TIMEOUT and this does not work.  The job keeps running on the problematic targets after this value is exceeded.  Is there any way to get these targets to time out or to get all of the results after aborting the job as a whole?