{macOS} Client stopped receiving work (SOLVED)

[asm]

I've been running 7.5.1 for a few days now and all of a sudden it stopped receiving work today.

The logs show:

Code: Select all

...
20:08:50:Trying to access database...
20:08:50:Successfully acquired database lock
20:08:50:Enabled folding slot 00: READY cpu:15
20:08:50:WU00:FS00:Connecting to 65.254.110.245:8080
20:08:51:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
20:08:51:WU00:FS00:Connecting to 18.218.241.186:80
20:08:51:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
20:08:51:ERROR:WU00:FS00:Exception: Could not get an assignment
20:08:51:WU00:FS00:Connecting to 65.254.110.245:8080
20:08:52:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
20:08:52:WU00:FS00:Connecting to 18.218.241.186:80
20:08:52:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
20:08:52:ERROR:WU00:FS00:Exception: Could not get an assignment
20:09:52:WU00:FS00:Connecting to 65.254.110.245:8080
20:09:52:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
20:09:52:WU00:FS00:Connecting to 18.218.241.186:80
20:09:52:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
20:09:52:ERROR:WU00:FS00:Exception: Could not get an assignment
20:11:29:WU00:FS00:Connecting to 65.254.110.245:8080
20:11:29:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
20:11:29:WU00:FS00:Connecting to 18.218.241.186:80
20:11:29:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
20:11:29:ERROR:WU00:FS00:Exception: Could not get an assignment
20:14:06:WU00:FS00:Connecting to 65.254.110.245:8080
20:14:06:WARNING:WU00:FS00:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
20:14:06:WU00:FS00:Connecting to 18.218.241.186:80
20:14:06:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
20:14:06:ERROR:WU00:FS00:Exception: Could not get an assignment

I have tried restarting the client with:

Code: Select all

FAHClient --send-command shutdown

This did not help.

Is this a problem at my end or F@H?

[Joe_H]

Will have to check further, but I have heard that one of the servers was about to be taken offline for a day or two. Elecetrical work needs to be done in the server room. Assignments may have been turned off to prepare for that. That server appears to have most of the larger CPU projects that would scale to 15 cores.

Will update here if I learn more.

[asm]

Ah - thanks!

If I dial it back to 7 cores will that make a difference?

Aside: Is there a way to have finer-grained control over # of cores in FAHControl? It seems like I'm limited to 7, 15, or 16 (Light, Medium, Full).

[asm]

If I dial it back to 7 cores will that make a difference?

To answer my own question: I set it to 7 cores (Light), restarted the client, and it downloaded an assignment properly. Either this is coincidental timing, or it worked

I would have expected that if the assignments are based on number of cores, then there should be some kind of fallback if you have "too many" cores for the available assignments. Otherwise a lot of potential processing is wasted.

[Joe_H]

There is a fallback process, but it sometimes does not fall back far enough and fast enough for higher thread counts.

One note, the Gromacs code used in the CPU folding cores has trouble with what is considered "large" primes and their multiples. Currently that is enforced for primes 7 or greater. In your case the 7 setting will have been changed to 6 if you examine the log.

Some smaller projects will have issues with 5 and its multiples.

Using the Configure function of FAHControl you can set a specific number of threads to use, That gives a bit more control than just using the slider.

[toTOW]

Some researchers were backing up data out of work servers and had disabled work distribution to speed up the process.

Since they are now aware of the lack of work, they have re-enabled some projects. Things should be better now.

[asm]

OK - thanks to both of you for the info!

@Joe: Using the slider would be a lot more convenient. I find I'm switching it up depending on what I'm doing at the time. Open FAHControl -> click the slider -> close FAHControl. (Actually, on macOS a system menu would make most sense... start/stop/slider for # of cores & display the status and progress.)