Adv Multi-Row Queue gets stuck trying to initiate Windows Update

[mkrautler]

I have a a couple hundred servers in an advanced multi-row queue sequence & every once in a while random servers get stuck on initiating Windows Update after a server reboots, which never seems to initiate & I will need to pause & resume the job queue on that row before it connects to Windows Update again. Tonight the entire group’s queue had to be paused & resumed before the next task in the job queue would kick off. I’m curious if it’s looking to initialize Windows Update before Windows Update on the remote machine is ready for it? I already set the maximum concurrent thread count to unlimited and was wondering if there are any other settings that enable more resources to hopefully prevent this from happening again in the future.

Any help would be greatly appreciate, thanks in advance.

[doug]

Are you using integrated security or alternate credentials?

To be clear, when you described using pause/resume on the job queue… are you saying that you literally are just clicking “Pause job queue” for the stuck row, and then a moment later “Unpause/resume job queue” for that row, and it immediately starts moving again?

Thanks.

[mkrautler]

Yes, that is what I had to do this time, other times, I have had to abort the job queue since it is holding up other row sequences & won’t proceed until that row is completed.

[doug]

Hmmm. This is weird. The reason I’m confused is because when BP is “Attempting to initiate Windows Update…” it’s in that Windows Update action thread. During that “Attempting” phase, the BP Windows Update thread is trying to put some files on the target computer via SMB. When you abort or pause a job queue, the abort or pause flag on the job queue is set, but that flag is not looked at in the middle of any action. It’s only checked in between each step/action in the job queue. So setting “pause” or “abort” while the action is sitting on “Attempting to initiate Windows Update” wouldn’t have any impact on the Windows Update thread’s status while that thread is in the middle of processing since that thread is not checking the job queue “pause” or “abort” flags. It would only be after that thread completes that the “pause” or “abort” flags would be checked by the main job queue thread, which makes it confusing that pausing or aborting could have any impact on the progress of the Windows Update action.

Are you using ‘Alternate Credentials’ for the rows/hosts that are affected? Or are you using ‘Integrated Security’ where no credentials are being specified for the row and instead that row is relying on the account that was used to launch the BatchPatch.exe?

[Author]

Posts