exited with zero status but no 'finished' file
23 Apr 2005 14:06:24 UTC
Topic 189022
(moderation:
I frequently get messages like the following:
Einstein@Home - 2005-04-23 09:31:18 - Result H1_1313.4__1313.8_0.1_T05_Run2_3 exited with zero status but no 'finished' file
Einstein@Home - 2005-04-23 09:31:18 - If this happens repeatedly you may need to reset the project.
A couple of WU's ago, I went ahead and reset the project, but that only resulted in messing up the WU's as far as I could tell. Since then, I still get this type of message, but am chugging along steadily completing WU's successfully.
Should I ignore it?
exited with zero status but no 'finished' file
)
I just found the same question answered in the FAQ's and other posts. I understand that if everything else is working, I can ignore this.
No need to reply. Thanks.
Actually I might re-open
)
Actually I might re-open this. I've had the same message and I looked at the FAQ's. In my case the WU had produced the same message thirty times, so I reset. The WU got blown away, does that get reported anywhere as possibly causing problems? It doesn't seem to have registered on my computer listing.
> Actually I might re-open
)
> Actually I might re-open this. I've had the same message and I looked at the
> FAQ's. In my case the WU had produced the same message thirty times, so I
> reset. The WU got blown away, does that get reported anywhere as possibly
> causing problems? It doesn't seem to have registered on my computer listing.
It won't register because you aborted it...
That message can be ignored in most cases. As long as the WU restarts after this message, you're fine.
On further investigation, the
)
On further investigation, the reason that BOINC starts spewing error messages is nothing to do with BOINC itself. NView (NVIDIA Desktop Manager) seems to hit some sort of problem with CCAPP (Symantec AntiVirus Engine) and throughs a message, something along the lines of "I can't manage this program, so I'm going to slow down to a crawl". Turning all the NView features off (which you don't need anyway) seems to fix it. Or you can re-boot.
I've got an insight I've not
)
I've got an insight I've not heard from anyone else yet. Also I found a weirdness with CPU time reporting in one related case.
On my machine, a dual processor PowerMac G5, I would frequently get this error message when the two units completed within a minute or so. This happened nearly 25% of the time (5 times in 22 results, over a several day span). I'm running Boinc 4.25
At one point, I paused one of the two Einstein processes (kill -STOP pid) and let it sit for about an hour, then continued it (kill -CONT pid). After this it ran for about 20 units with no problems, until the scheduler snafu meant they were re-syncronized when things cleared up and they restared.
Since that time I've gotten it twice more in about 6 or 7 units.
Is there a data or lock file that is getting clobbered when the two Einstein processes complete and try to signal the Boinc client???
Here's an example (with some pruning)
2005-04-26 11:46:54 [Einstein@Home] Starting result H1_1059.9__1060.4_0.1_T08_Run2_0 using einstein version 4.78
2005-04-26 21:46:12 [Einstein@Home] Computation for result H1_1059.9__1060.3_0.1_T08_Run2 finished
2005-04-26 21:46:12 [Einstein@Home] Starting result H1_1176.4__1176.9_0.1_T08_Run2_2 using einstein version 4.78
2005-04-26 21:46:13 [Einstein@Home] Started upload of H1_1059.9__1060.3_0.1_T08_Run2_0_0
2005-04-26 21:46:16 [Einstein@Home] Finished upload of H1_1059.9__1060.3_0.1_T08_Run2_0_0
2005-04-26 21:46:16 [Einstein@Home] Throughput 89816 bytes/sec
2005-04-26 21:46:43 [Einstein@Home] Result H1_1059.9__1060.4_0.1_T08_Run2_0 exited with zero status but no 'finished' file
2005-04-26 21:46:43 [Einstein@Home] If this happens repeatedly you may need to reset the project.
2005-04-26 21:46:43 [Einstein@Home] Restarting result H1_1059.9__1060.4_0.1_T08_Run2_0 using einstein version 4.78
2005-04-26 21:46:48 [Einstein@Home] Scheduler RPC to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
adding: H1_1059.9__1060.4_0.1_T08_Run2_0_0 (deflated 63%)
2005-04-26 23:09:30 [Einstein@Home] Computation for result H1_1059.9__1060.4_0.1_T08_Run2 finished
2005-04-26 23:09:30 [Einstein@Home] Starting result H1_1176.4__1176.5_0.1_T09_Run2_3 using einstein version 4.78
2005-04-26 23:09:31 [Einstein@Home] Started upload of H1_1059.9__1060.4_0.1_T08_Run2_0_0
2005-04-26 23:09:35 [Einstein@Home] Finished upload of H1_1059.9__1060.4_0.1_T08_Run2_0_0
2005-04-26 23:09:35 [Einstein@Home] Throughput 210229 bytes/sec
When the unit was restarted (apparently from a checkpoint) it completed in 4000 seconds, rather than my usual 28,000. It only claimed the 4000 as credit, rather than the total time. Isn't this incorrect?
See http://einsteinathome.org/workunit/911849