exited with zero status but no 'finished' file
![[B^S] Elmerfudd](https://einsteinathome.org/sites/default/files/imagefield_default_images/default-avatar.png?1347855344 "[B^S] Elmerfudd"){kind=avatar}
12 Mar 2007 19:03:46 UTC
Topic 192522
(moderation:
3/12/2007 12:01:59 PM|Einstein@Home|Task h1_0496.5_S5R1__1703_S5RIa_1 exited with zero status but no 'finished' file
Anyone getting this error? I am not sure what to do. Any help would be appreciated.
Elmerfudd
exited with zero status but no 'finished' file
)
It's usually just a warning that something has not allowed the unit to update / work for a bit. There are a few things that can cause this. 5.8.15 fixed most of them, but with the Daylight Saving Time that happened over the weekend, I see more people with this error, today. The shift of the clock probably confused how long it was truly working on certain files, and may cause this error.
If you are not on 5.8.15, update, otherwise ignore and all will continue normally.
Was using 5.8.11. Will try
)
Was using 5.8.11. Will try the new. Thanks
Yep daylight savings are a
)
Yep daylight savings are a very common cause for this error, just like adjusting your clock too much at a time via a timeserver. I have had this errors loads of times but never lost a WU because of them, so don't worry too much.
Btw, adjusting your local time can also cause a "no heartbeat from core client" error, which seems even more dramatic but never harmed my WUs either.
RE: Yep daylight savings
)
Acutally the two go hand in hand. First thing that happens is the science app loses the heartbeat, reports that in stderr, and then exits. Usually when whatever was forcing BOINC to 'oversleep' clears a short time later, the CC sees that the science app went bye-bye, assumes it was finished, but has the wrong status, and thus throws the misleading message (for this kind of event) in the message tab, and restarts it.
In any event, it seems reasonable that anything which would mess with the clock, timers, etc. would be a candidate to trigger this kind of event.
The worst thing which would normally happen is if you set the clock/calendar enough into the future as to cause a deadline issue and force BOINC into EDF mode.
Alinator
Ok, upgraded to 5.8.15
)
Ok, upgraded to 5.8.15 version and all is running great again. It even appears that I got full credit for the crunched work that had the error.
Thanks all!!!
OK, that's good news. One
)
OK, that's good news.
One thing to keep in mind is anything, including Windows itself, which all of a sudden demands virtually full time attention from the CPU for an extended period at high priority can trigger this event. Most likely you'll still see it once a while with the latest version of BOINC. It would have probably been better to word the warning message something like: "Unusual Application exit event, attempting to resume...". Also as has been said before, it's completely harmless for the vast majority of times it happens.
Alinator