LHC@home

yesterday and today, I had two Theory tasks where I noticed only after long time that the CPU was no longer doing any work, so I abortet them:

https://lhcathome.cern.ch/lhcathome/result.php?resultid=285041017
https://lhcathome.cern.ch/lhcathome/result.php?resultid=284980288

stderr in both cases showed:
cranky: [ERROR] 'cvmfs_config probe sft.cern.ch' failed

anyone any idea what's going wrong?

The internet connection was too slow as the tasks were starting.


This happens in the first 3 or 4 minutes after it starts but they just keep on running anyway.
You can check the VB Console on any running task to see if they are doing this.

thanks for the explanation.
In fact, I wouldn't even need to check the VB console; a short view at the Windows task manager would show me if there is no CPU activity under "VBoxHeadless.exe". Too bad though that, in both cases, so much time went by until I took a look, and then terminated the task manually.

It's surprising that there was a problem with the internet connection. Since no other of my machines (on some of which also Theory tasks are running) showed an internet connection problem, I guess it might have been a problem of the WLAN adapter of that notebook).
Meanwhile, two other Theory tasks were processed successfully, and another one is running okay just now.

At any rate, I need to keep a closer eye on this notebook.

this morning, I detected another such case.
Unfortunately, due to lack of time, I havn't taken a close look at this task for some time, and so I found out, after a runtime of 4 days, that this task was running but not doing any computation. So I aborted it manually. Too bad for the time waste.

I am wondering why these tasks don't stop themselves once this problem comes up within the first few minutes, making the tasks invalid right from the beginning on.
Kind of an automatic stop procedure would be very helpful and would make sense.

The same has just happened to me. 22 hours of doing nothing.