Page 1 of 1

Project 13444 WU corrupted, but accepted on upload

Posted: Fri Feb 05, 2021 1:08 pm
by paulch2
No idea if this one would be caugt automatically, but looked like the server accepted it.
If possible, can somebody delete that from the finished queue and re-submit it.

Odd behaviour after a Radeon software update + reboot.
First WU after the reboot zipped through in a few minutes, rather than a few hours, and seemed to be accepted OK.
Others seem to be fine since then:

10:58:19:WU00:FS01:0x22:Project: 13444 (Run 6934, Clone 54, Gen 0)

10:58:19:WU00:FS01:0x22:Unit: 0x00000000000000000000000000000000

10:58:19:WU00:FS01:0x22:Digital signatures verified

11:01:26:WU00:FS01:Sending unit results: id:00 state:SEND error:NO_ERROR project:13444 run:6934 clone:54 gen:0 core:0x22 unit:0x00000036000000000000348400001b16

Re: Project 13444 WU corrupted, but accepted on upload

Posted: Fri Feb 05, 2021 1:33 pm
by Neil-B
Any chance of a fuller log post (verbosity 3) from the reboot to the completion and confirmation of the upload of the wu>

Re: Project 13444 WU corrupted, but accepted on upload

Posted: Fri Feb 05, 2021 4:07 pm
by bruce
Based on https://apps.foldingathome.org/wu#proje ... e=54&gen=0, it looks like two corrupt WUs were accepted and one was successfully completed. The user names don't look like they belong to you, paulch2, so it looks like the valid simulation was accepted and the bad ones rejected, which means the FAH servers successfully dealt with whatever the corruptions were.

Re: Project 13444 WU corrupted, but accepted on upload

Posted: Fri Feb 05, 2021 4:29 pm
by paulch2
That one was using the vfpc username, marked as OK in that list.
Sorry, should have included the log, but that's since gone - got a bit paranoid and wiped everything to start again.
Seems to have been a one off glitch, but I'll keep an eye on it for a while.

Thanks for checking.

Re: Project 13444 WU corrupted, but accepted on upload

Posted: Fri Feb 05, 2021 4:50 pm
by bruce
Did your drivers change recently? Maybe a bad revision got distributed. and then it was corrected.

The main point is the bad results didn't make it into the scientific results.