| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Unrecoverable Error |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 10
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I posted this in the FightAIDs@Home forum on 5 Feb but was told by L.C. that he thought it was an application error, and that I should have posted it here...
 I was wondering if anyone else has seen the following? I've had several of these lately...  That popped up right after this happened: 2/4/2006 10:17:20 PM|World Community Grid|Unrecoverable error for result faah0174_diversity0687_x6084_01_3 ( - exit code 1282 (0x502)) |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Morning Nite Owl,
Haven't come across the error myself, so this is a bit of a stab in the dark, but did come across a post on another BOINC site on exit code 1282. It appears to be a problem with checkpointing of the file if it is removed from memory before preempting (Whatever that may mean  )Are your preferences set to leave the applications in memory while preempted? (The post I'm refering to is http://africa-home4.cern.ch/malariaControl/forum_thread.php?id=19.) Hope that may help. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks Slug,
Every BOINC message forum that I have looked through has posts about the 'preempt' problem. It is a known bug in the BOINC system. BOINC allows time slices for multiple projects, so with X projects, there is the potential to have X work units being worked on with none of them complete. This could really hog Virtual Memory. So BOINC has an option to delete a project from memory when it is preempted by another project. This should work the same way as a reboot. When the project is selected again, it should be reloaded from the last check point and proceed from there. Unfortunately, this often fails for reasons unknown. Until this bug is corrected, always choose the option to leave projects in memory when preempted and make sure that you have enough Virtual Memory to handle your BOINC projects. mycrofth |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Morning mycrofth
Thanks for a full explaination. Maybe this is one for the FAQ section? Regards Slug |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Owlie used to do Rosetta and that has a serious bug requiring to leave in memory so I think he could be aware of this one!
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have always had "Leave in memory" selected in my preferences... on all projects. However, the machines that are experiencing this problem are running WCG only...
---------------------------------------- [Edit 1 times, last edit by Former Member at Feb 10, 2006 8:36:34 AM] |
||
|
|
RT
Master Cruncher USA - Texas - DFW Joined: Dec 22, 2004 Post Count: 2636 Status: Offline Project Badges: 
|
Nite Owl
----------------------------------------I have the same options chosen and I only run WCG. I have seen the same problem but not chased it assuming that the dump sent to Microsoft will find its way to Berkley for debugging. I run multiple problems and some other resourse hogging programs at the same time. (Machine is a dual core P4 with HT). I suspect mycrofth is correct (usually is  ).Perhaps the debug info is not getting through to Berkley, anyone know for sure |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The debug info is sent to Microsoft and isn't going to help the problem in the least. I suspect the WCG Agent has a driver that periodically screws up, causing Windows to reject the work unit. I have WCG running the AIDs project on all 30 of my computers and I'm seeing more and more of these rejects each day.
The question is.... What's up?  |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Good point, Nite Owl.
It might be an application problem rather than a BOINC problem. But if it is not replicable, then . . . Sometimes, erratic bugs like that take a long time to track down. If it is not keeping us from completing work units, just slowing us down, then we can live with it until somebody in TSRI has a bright idea. The question is - - should we be asking for some special report on errors of this kind? mycrofth |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Good point, Nite Owl. It might be an application problem rather than a BOINC problem. But if it is not replicable, then . . . Sometimes, erratic bugs like that take a long time to track down. If it is not keeping us from completing work units, just slowing us down, then we can live with it until somebody in TSRI has a bright idea. The question is - - should we be asking for some special report on errors of this kind? mycrofth ask away I'd be more than happy to post or send anything I can track down!  |
||
|
|
|