Hi all …
My question is in two parts. The questions I have pertain to an issue with the “Run Capacity Scan & Upload Results” action that is part of ILMT.
We have several actions that run this task set to a schedule broken up by operating system (one for Windows, one for Linux, etc). The scan is scheduled to run every 30 minutes, the start time being staggered over the first 10 minutes. If an error occurs, three reattempts will be made 30 minutes apart, afterwhich no more tries will be made. The scheduled scans have been running fine for months.
Today we happened to notice a high percentage of runs that have the status of < error >. When we looked at “show action info”, the error message at the top stated that there was a syntax error in the action. It was clear that no statement in the ActionScript was executed. How there can be a syntax error in the action that’s been running fine for months is beyond me. We did not make any changes to it whatsoever. When we checked the BigFix client log on an affected machine, it said the following:
At 21:36:57 -0400 - actionsite (http://computer.company.com:52311/cgi-bin/bfgather.exe/actionsite)
Relevant - Run Capacity Scan and Upload Results - Linux Policy (fixlet:254632)
At 21:36:58 -0400 -
ActionLogMessage: (action:254632) Action signature verified for Downloads
ActionLogMessage: (action:254632) Action temporal distribution - delay for 0 seconds.
ActionLogMessage: (action:254632) Parameter command error. Parameter may not already be defined.
ActionLogMessage: (action:254632) ending action
At 21:37:18 -0400 -
Report posted successfully
Can anyone explain why this has occurred? The version of BigFix is 9.1.1117.0; there are plans to upgrade but they are not immediate. And what does “Action temporal distribution - delay for 0 seconds.” mean?
Second, when the status becomes < error >, execution stops. It appears that whatever you specified in the schedule regarding how many times to try after an error occurs is ignored when a syntax error occurs, so anything that reported an < error > is dead in the water with no restart possible. When we look at “show action info”, we see when the scheduled action last ran and that it encountered this “syntax error”. We figured that this might be a one-time hiccup, so we picked two affected machines and ran a one-time capacity scan, which completed successfully. So, fine, I have a workaround - get a list of the affected machines that reported < error >, and run the capacity scan on them one time. But what about the schedule? Will the scheduled action see that a run has now had a successful completion and pick up where it left off, or do I need to stop the scheduled action, delete it and recreate a new one so it takes effect on all the computers all over again?
I am cross-posting this to the ILMT forum as well since I don’t know if this is a BigFix problem or a strange content-authoring problem of an IBM-supplied ILMT action.
–Mark