| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: Suggestions / Feedback
Thread: Work Units returned as 'error' |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 6
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have noticed that there are a lot of work units that have been returned and marked as 'error'. Perhaps there could be a change made to automatically turn off a project if the number of work units retunred with a result of 'error' exceeds a pre-determined threshold. This would then limit the problems of resending out work units that have been returned in a problem status.
|
|||
|
||||
|
Vester
Senior Cruncher USA Joined: Nov 18, 2004 Post Count: 325 Status: Offline Project Badges: 
|
I have noticed that there are a lot of work units that have been returned and marked as 'error'. One can cause an "error" status by aborting a job, having corrupted results, or, with Windows Vista, by not extending the shutdown time thereby not allowing files to be written to the hard drive before shutdown. Are you running a desktop or a laptop? What operating system? Overclocked? I recommend that you run Memtest86 from a floppy or USB drive depending on your computer's capabilities. You need to ensure that the drive you choose is before the C drive in the boot sequence. If that is too hard to do, try Prime95. Someone else can address the remainder of your post: Perhaps there could be a change made to automatically turn off a project if the number of work units retunred with a result of 'error' exceeds a pre-determined threshold. This would then limit the problems of resending out work units that have been returned in a problem status.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Vester,
I am not having a problem. I do get many work units that have been resent with 'errror' as having been returned by other machines due to the fact that I turn them back arround within about 8 hrs of clock time cleaning up the mess created by others.  |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
There's nothing wrong with any project of material nature. If you got a lot of the back up jobs it means that you have a high reliability and fast return rate, thus you would see an above normal number of error returns in the quorum detail view. The system tends to send 'repair' jobs to these outstanding clients.
----------------------------------------I barely see any, but I'm a slower returner.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
There's nothing wrong with any project of material nature. If you got a lot of the back up jobs it means that you have a high reliability and fast return rate, thus you would see an above normal number of error returns in the quorum detail view. The system tends to send 'repair' jobs to these outstanding clients. I barely see any, but I'm a slower returner. I guess that makes me a fast returner.. adverage for me on one rig a dual core is doing 2 in little over 4 hours each.. on most WU.. been running BOINC about a day and I don't see a hitch yet.. does that mean when my rig becomes a favorite for the server am I going to see a larger WU queue on this end?? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello choj01,
The size of your work queue is determined by the 'connect every x days' parameter (BOINC 5.8 and earlier). In BOINC 5.10 it is controlled by the 'additional work buffer' parameter. If you are always connected, there is no need for a large work queue, which only slows down the validation of returned work units. Lawrence |
||
|
|
|