Slow Time in Server Automation Tasks

Hi Folks, I’m preparing my tasks for patch with Server Automation, and I can see it’s so slow. I made some tests in servers and the time obtained is not good. Looking for the bottleneck… I take a server fully patched and send the Automation Task in different times, enabling/disabling the Prefetch downloads option, scheduling for start after a couple of hours…and practically obtain the same time.

I can resume the steps and time taken:

I believe than the only step who really needs a time for execution would be the Baseline, but not the rests.

Additionally, when I want to monitor the steps, they only show several minutes after was completion…

It is normal?

I want to thank you for your attention!

Hi hectorio,

Thanks for posting us your findings.

So I would probably ask for two additional pieces of information:
(1) How many endpoints are you targeting with these steps (tasks)?
(2) Are the times you’re seeing here different to running the tasks as standalone actions? (i.e. not using SA, but instead just taking action on the tasks in the IEM console). The time I would be interested in here is the time taken for ALL targets to return a Not Relevant, Fixed or Failed result.

The way SA operates is that all targeted endpoints for a step must return a status that indicates there is no further work to perform on that endpoint (e.g. Not Relevant, Fixed, Failed, etc.) before the next step is processed. If this is not desired behavior, then you can optionally set a timeout for any step so that it only will run as long as you’re prepared to allow. The plan then moves on to the next step (in the case of setting a timeout, you’ll probably want to set the ‘continue-on-fail’ option here, as endpoints that do not complete a step within the allocated time are seen as failures in the SA world. You can also optionally exclude such endpoints from the remaining steps in the plan if you so wish).

If some targeted endpoints are particularly slow, then this can delay the overall elapsed time for the plan step, since as I mentioned, it will continue to wait for those to complete the step unless a timeout is specified. To optimize with this, it can be useful to group slow endpoints together and run a separate plan that targets those. Both plans will run at the same time, but it just means that the faster endpoints will complete quicker.

On your last monitoring question, there is a blue refresh button in the top right corner of the Automation Plan Status Dashboard, I suspect you just need to click that every so often to get the most up-to-date status.

Please let us know your findings on the two questions above.

In addition, you can check the Client logs (C:\Program Files (x86)\BigFix Enterprise\BES Client__BESData__Global\Logs) on the targetted server to see exactly what is happening.

Hi Paul, sorry for delay in the answer and sorry for re-open this old conversation…

I am working with automation plans sending one task for one server (like the initial example), and the delay occurs in different servers, and in others, it runs like a charm.

In other example, in this server, i send a task for restart
custom task 01: Normally takes two minutes for complete, but in the example we can see it take almost one hour
restart endpoint: Server was online after five minu

Looking for the log, I can found than the actions only was executed after a manually restart for bes client…
log_sa.pdf (16.6 KB)

Bigfix server: 9.1 patch 3.

Thanks for your help!

Can the server that is having trouble receive the UDP messages from it’s Relay (or your BigFix Server if no Relay)? No firewall or communication issues?

If you right click the computer in the Console and choose “Send Refresh” does the client report back in a few minutes with a new Last Report time?

Also, if you watch the client logs (Baretail is great for this) on the problematic server while sending the “Send Refresh” command, do you see a line “ForceRefresh command received.” shortly afterward?

1 Like

Thanks Tim! I found some servers was working with a relay with udp communication problems… Moving the clients to the correct relay, and reinstalling bes client when the move does not work… seems to be work the automation plans…

Many Thanks for your help and for you quick reply!

Regards