| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Incorrect tally of WU's |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 7
|
|
| Author |
|
|
mike047
Senior Cruncher Joined: Aug 22, 2006 Post Count: 262 Status: Offline Project Badges: 
|
Two of my boxes have incorrect tallies of wu's.
----------------------------------------ID 60481 ID 168998 The wu list in "my grid" indicates many MORE units than the boxes actually list. I have taken into cosideration the pending, in work, etc units and directly compared each box to the my grid list. Anyone have any idea where these wu's are going??? These boxes only keep about 12 hrs in queue and I have it set for 48 hrs. I have 41 other boxes set for the same queue and they are all correct. Also, yesterday another box faulted out with a " Oswaldo C collaboration" error. Can anyone shed any light on this?? It shut down the Boinc client and I had to reinstall to correct the problem. 5.4.11 all boxes Thanks in advance.
mike
----------------------------------------Crunch Hard, Crunch Often  [Edit 1 times, last edit by mike047 at May 8, 2007 9:52:05 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
If you are talking about queued work, then knreed has temporarily disabled the "resend lost results" feature. This may mean that sometimes the servers will send a work unit, but the agent won't receive it. Is that what you are seeing?
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Hi Mike047,
----------------------------------------Can you post the specific error message. Can't locate any previous reports on the fora with this, what i presume to relate to Genome Comparison jobs. What OS is this on? (blasphemous maybe to ask if you are on Linux :O) Also, yesterday another box faulted out with a " Oswaldo C collaboration" error. Can anyone shed any light on this?? It shut down the Boinc client and I had to reinstall to correct the problem. 5.4.11 all boxes Thanks in advance. Like Didactylos, presume you are talking about the Result Status page showing more work than you actually received and is tallied in the Task Tab of the BOINC agent.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
mike047
Senior Cruncher Joined: Aug 22, 2006 Post Count: 262 Status: Offline Project Badges:
|
If you are talking about queued work, then knreed has temporarily disabled the "resend lost results" feature. This may mean that sometimes the servers will send a work unit, but the agent won't receive it. Is that what you are seeing? Yes, that is what I see, I think  I'll wait it out, the box is crunching fine.
mike
Crunch Hard, Crunch Often |
||
|
|
mike047
Senior Cruncher Joined: Aug 22, 2006 Post Count: 262 Status: Offline Project Badges:
|
Hi Mike047, Can you post the specific error message. Can't locate any previous reports on the fora with this, what i presume to relate to Genome Comparison jobs. What OS is this on? (blasphemous maybe to ask if you are on Linux :O) Also, yesterday another box faulted out with a " Oswaldo C collaboration" error. Can anyone shed any light on this?? It shut down the Boinc client and I had to reinstall to correct the problem. 5.4.11 all boxes Thanks in advance. Like Didactylos, presume you are talking about the Result Status page showing more work than you actually received and is tallied in the Task Tab of the BOINC agent. OS is XP  The error was multiple times, the windows variety....do you want to send a report one that flashes after an error. I did not write down the entire error, just what I posted. It was definitely linked to boinc as after reinstall no further errors. I think the box was working on a FCG unit, it errored out many[8??] FCG wu's as it was having its problem. These reported as error work. This ID is 60461 It also has some VERY large credit awards for zero cpu time and the wu's are marked valid. Someone might want to check on this as it is TOO much credit if it was actually granted. These are HPF2 units. These occurred on 5-6/7-2007.
mike
----------------------------------------Crunch Hard, Crunch Often [Edit 4 times, last edit by mike047 at May 8, 2007 11:13:32 PM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Unless there is a bug in the validation process, assume a valid WU is really a valid WU. In that case, where zero time for instance is reported, the credit is granted as per algorithm. If it's higher than the other WU's in the FCG quorum, something is awry. For HPF2 it is actually possible as was confirmed. In that case more attempts were computed, which is not seen very often.
----------------------------------------Those 'Not Received' WU's will eventually be cleaned out in a periodic scheduler process. edit: grammar and spelling corrections 
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at May 9, 2007 8:08:44 AM] |
||
|
|
mike047
Senior Cruncher Joined: Aug 22, 2006 Post Count: 262 Status: Offline Project Badges:
|
Unless there is a bug in the validation process, assume a valid WU is really a valid WU. In that case, where zero time for instance is reported, the credit is granted as per algorithm. If it's higher than the other WU's in the FCG quorum, something is awry. For HPF2 it actually possible as was confirmed. In that case more attempts were computed, which is not see very often. Those Not Received WU's will eventually be cleaned out in a periodic scheduler process. THANKS, I just wanted to be sure everything was OK I'll wait it out.
mike
Crunch Hard, Crunch Often |
||
|
|
|