Another randomly dying scheduler

[Paul Gardner]

Bob,

Another case of scheduler and ACP just straight up dying. This one at Thacher. The scheduler engine log is full of errors ive never seen before. I think some may be somewhat similar to some of the other systems that have had issues. Here is the scenario...

1. fresh startup of ACP and scheduler
2. enable dispatcher
3. system runs startup
4. at some point along the way, not sure when since I was off doing other stuff, everything just shuts down. Both Scheduler and ACP programs full stop, no longer open.

User reports it crashing when logged onto the web server. I have not been able to fully test that theory to see if it makes any difference. I was logged on to the web browser each time it happened. So far it has been repeatable.

Paul

[Paul Gardner]

it is running Windows 10 build 1809

[Timothy Brothers]

Paul - So, older build of W10... could be some of those smaller KB or .Net updates that have been recent? I had updates suspended since September, then once I let mine through on build 1903, Scheduler simulator started working again. Still a lot more testing to run through though. Tim

[Bob Denny]

Yes this was a broken update that got through, I've seen it before. There were some broken updates that got out around August as far as I can tell, and seemingly they were replaced by later ones that fixed the problems.

Did this "just start happening" or has Thacher ben shut down for a period and maybe when fired up or rebooted one or more Windows updates were applied? Did Swift not get my message about stopping updates? It went to all of the ACP customers' technical contacts. He is listed for Thacher.

Right now I have a Windows 10 1903 system running with daily updates, and I ran an exhaustive 4 hour test yesterday. All is well. There are two ways to go: (1) If you have an image backup of the running system, restore it and shut off Windows Updates, or (2) keep applying updates to it until there are no more to apply. If you see any optional "rollup" updates apply them as well.

This applies to all of your systems like Soka. I'm going to be sending out another letter as soon as we see the next feature update replacement for the failed/pulled 1909 one. Go back and check the "stop updates" on all of your systems so it extends through November. I did it to GBO through 11/17.

[Paul Gardner]

1903 has been stable so I will likely bring it up to that level and see how it goes.

[Bob Denny]

Do all of the updates. There may be a bad one from over the Summer. I have applied all of them up to last night and tested the heck out of it. Keep checking for updates over and over until you get “no more”.

[Bob Denny]

I want to close this thread. What is the status?

[Paul Gardner]

You can close it. All stable on the western front. 8 systems running smoothly last night.

[Bob Denny]

Excellent. Thanks Paul!!