Page 1 of 1

Pause/Unpause doesn't seem to work in the docker image?

Posted: Mon Feb 01, 2021 3:28 am
by aolives
Long story short, I have been trying to programmatically pause and unpause my clients via another machine.
Came across the fah control client and figured no problem, telnet in, pause or unpause, no problem.
Spun up a clean instance of the Official Folding@home GPU Container on my local machine, telnetted in, paused

Code: Select all

02:51:32:FS00:Paused
02:51:32:FS00:Shutting core down
02:51:32:WU00:FS00:0xa7:Caught signal SIGINT(2) on PID 19
02:51:32:WU00:FS00:0xa7:Exiting, please wait. . .
02:51:37:WU00:FS00:0xa7:Folding@home Core Shutdown: INTERRUPTED
It then saves the config and then gives

Code: Select all

02:52:57:WARNING:FS00:Killing WU00
until I shut it down.

Anyone else seeing this happen whenever they try to pause or am I doing something wrong?
Thank you

PS: If I try to unpause despite the above weirdness the machine does not resume folding.

Re: Pause/Unpause doesn't seem to work in the docker image?

Posted: Thu Feb 04, 2021 8:33 pm
by aolives
:e?:

Re: Pause/Unpause doesn't seem to work in the docker image?

Posted: Thu Feb 04, 2021 9:20 pm
by bruce
I can't speak for the Docker image, but the log from the conventional client would read much like that. FAHCore_a7 should have been terminated, leaving WU00 in a state that would allow it to resume from the previous checkpoint. The work processed since that previous checkpoint will need to be reprocessed. The frequency of checkpoints for that WU will matter. Without a more complete segment of the messages associated with that WU, it's difficult to tell if that explanation makes sense or whether you're reporting a bug.

Assuming there is an appropriate checkpoint, can work be unpaused? What was reported when you tried to resume that WU?

If you're running FAHControl on another machine, you can configure the "Clients" tab to do the Pause/Unpause manually.