2 WU\'s Computation Error

Author	Message
old_user72842 Send message Joined: 2 May 05 Posts: 21 Credit: 1,290 RAC: 0	Message 12405 - Posted: 8 May 2005, 15:35:39 UTC I just had 2 WU's in a row have a computation error under BOINC 4.37...what is up with that?! <a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a> ID: 12405 · Reply Quote

Andrew Hingston Volunteer moderator Send message Joined: 17 Aug 04 Posts: 753 Credit: 9,804,700 RAC: 0	Message 12416 - Posted: 8 May 2005, 19:37:42 UTC They both seem to have returned the same -1 error code. I don't recall reports of this very often, nor what it signifies, but somebody else has recently encountered the same error. I suspect the problem may well be BOINC related - was there anything significant at the time? For example, were you online? Anything that could have interrupted a process, eg disk access? ID: 12416 · Reply Quote

Ananas Volunteer moderator Send message Joined: 31 Oct 04 Posts: 336 Credit: 3,316,482 RAC: 0	Message 12417 - Posted: 8 May 2005, 20:33:43 UTC I had a -1 when the CPDN client exited without having been stopped by BOINC and without having left a message for BOINC about the reason. This can happen when the process gets killed by shutdown or through the task manager before BOINC has a chance to end the client. The behaviour of BOINC will hopefully be fixed some day though, at least I heard it has been planned (somewhere in the Einstein forum). BOINC should ignore if the project client process just exits and give the WU /model another try. When the next start of the model issues an error, it will still be time to report a failure status to the scheduler. ID: 12417 · Reply Quote

old_user72842 Send message Joined: 2 May 05 Posts: 21 Credit: 1,290 RAC: 0	Message 12538 - Posted: 12 May 2005, 20:52:38 UTC - in response to Message 12417. > I had a -1 when the CPDN client exited without having been stopped by BOINC > and without having left a message for BOINC about the reason. > > This can happen when the process gets killed by shutdown or through the task > manager before BOINC has a chance to end the client. > > > The behaviour of BOINC will hopefully be fixed some day though, at least I > heard it has been planned (somewhere in the Einstein forum). BOINC should > ignore if the project client process just exits and give the WU /model another > try. When the next start of the model issues an error, it will still be time > to report a failure status to the scheduler. > it seems to be fixed w/ 4.38, maybe or maybe not... <a href="http://tinyurl.com/9hemz"><img src="http://www.boincsynergy.com/images/stats/comb-1441.jpg"></img></a> ID: 12538 · Reply Quote