Message boards : Number crunching : Server refuses to send more work
Author | Message |
---|---|
Plasmon_attack Send message Joined: 2 May 10 Posts: 13 Credit: 15,451,384 RAC: 0 |
I have a 64 bit win7 machine that I updated with the correct software this morning. I only got 8 tasks (I have 16 cores) and all completed just fine and have been reported. However, the server refuses to send any more work (reached daily quota of 8 results). I've detached and reattached to the project, but it still won't send anything and is deferring communication for 9 hours. Why is this computer limited to 8 tasks/day when my other ones aren't? This machine COULD do >100 tasks/day but work needs to be sent. I know sometimes if a computer sends in too many bad work units (I tried adding this computer yesterday and 8 work units ended with computation errors) it gets stalled but since it's returned good ones now why the withholding? Thoughts? |
Link Send message Joined: 4 May 07 Posts: 356 Credit: 382,349 RAC: 0 |
With your computers hidden it's hard to tell, but your machine probably trashed over 100 WUs due to the 3.43 issue and it is down to 1WU/day/core. But actually you should get than 16 per day. Each of those 8 successfully crunched WUs should however double the quota as soon as they validate, which should not take too long here. So just wait and eventually crunch some other projects while you wait. If the problem won't go away by itself, you'd have to unhide your computers. . |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
You are correct that returning good work units should permit more work to be sent. Perhaps the servers are still behind and there was a delay in allowing additional work to be sent to your machine. It is also possible that the successfully completed tasks later received validation failures. You can see that if you click in to your profile and machines on the project webpages. Rosetta Moderator: Mod.Sense |
Message boards :
Number crunching :
Server refuses to send more work
©2024 University of Washington
https://www.bakerlab.org