The second WU finished okay, too, which I consider a near miracle given that my external hard disk is dying on me and I'm busy trying to back up the relevant data, which means a lot of disk operations and extra CPU usage for my poor little laptop making gzipped archives and copying them around... an app which survives this can hardly be anything less than rock solid.
Jos, that seems to be a system-related problem rather than sth to do with the new science app. Maybe some permissions got changed by accident?
Funny how that works, the second result on Linux+CC4.19 has 0 CPU time also, plus both have some message in it that I would have assumed to be critical : "Can't set up shared mem: -1 Will run in standalone mode." but still it gets the WUs done :-)
It's probably an API vs. CC version or an API vs. kernel version issue, sure not an Einstein specific problem.
The change in behaviour happened between 4.02 and 4.20 btw.
CC4.19 seems to handle EAH4.20 nearly as it was a non-BOINC application but it can still run it without a wrapper. I will check the CC logfile sometimes.
Just restarted crunching after a 10 days stop due to a hospital stay, and my WU is hopelessly late. I switched to 4.20 from 4.16 and things seem to run smoothly.
Tullio
You might want to have a look at this; apparently I'm the first person to get a signal 11 with the new app.
This system is a Core Duo laptop, Kubuntu 7.10, kernel version 2.6.22-14. If there is anything I can do to help just let me know.
You seem to be the natural-born-beta-tester :-). Well, you could try to catch one this signal 11 exceptions with the ddd debugger, apparently this is still enabled on your BOINC installation but the debugger doesn't show up because BOINC runs as a daemon without X11 .
"Application caught signal 8" it says (afaik. that's an FPU error) and your previous result with 4.02 has been invalid, there might be something wrong with your box (dust bunnies in the cooler, PC placed in a corner with insufficient air flow ...)
"Application caught signal 8" it says (afaik. that's an FPU error) and your previous result with 4.02 has been invalid, there might be something wrong with your box (dust bunnies in the cooler, PC placed in a corner with insufficient air flow ...)
I'm running this box as my main desktop system. It runs with about 37 to 42°C CPU temperature. Cleaned of dust recently and supported with fresh thermal grease.
Alternativly with sudoku@home there are no errors (only one for a power fail).
Might there be a difference in FPU usage?
Sorry of reporting my failure so late, but 50% of my Einstein@home WU previously have succeded.
Things were going quite well here, but now I've just received my first "signal 11's". In fact, I received three consecutive ones. They're on machine number 1060000 in case anyone wants to look.
(Of course, I was also having some network problems at the time. Could that have caused this?)
Donald, that's interesting... I was at Uni at the time and the WiFi access point had died again, so, no internet... and BOINC might have wanted to contact the server to report the two previous WUs, which had just finished... I already wondered if there was a connection but thought it was unlikely. But after reading your post, I'm not so sure any more.
The second WU finished okay,
)
The second WU finished okay, too, which I consider a near miracle given that my external hard disk is dying on me and I'm busy trying to back up the relevant data, which means a lot of disk operations and extra CPU usage for my poor little laptop making gzipped archives and copying them around... an app which survives this can hardly be anything less than rock solid.
Jos, that seems to be a system-related problem rather than sth to do with the new science app. Maybe some permissions got changed by accident?
Funny how that works, the
)
Funny how that works, the second result on Linux+CC4.19 has 0 CPU time also, plus both have some message in it that I would have assumed to be critical : "Can't set up shared mem: -1 Will run in standalone mode." but still it gets the WUs done :-)
It's probably an API vs. CC version or an API vs. kernel version issue, sure not an Einstein specific problem.
The change in behaviour happened between 4.02 and 4.20 btw.
CC4.19 seems to handle EAH4.20 nearly as it was a non-BOINC application but it can still run it without a wrapper. I will check the CC logfile sometimes.
Just restarted crunching
)
Just restarted crunching after a 10 days stop due to a hospital stay, and my WU is hopelessly late. I switched to 4.20 from 4.16 and things seem to run smoothly.
Tullio
Bernd? I had sth very strange
)
Bernd? I had sth very strange here. Apparently, I got two signal 11 errors with the 4.20 app.
I'm referring to these WUs:
http://einsteinathome.org/task/89595020
http://einsteinathome.org/task/89596329
You might want to have a look at this; apparently I'm the first person to get a signal 11 with the new app.
This system is a Core Duo laptop, Kubuntu 7.10, kernel version 2.6.22-14. If there is anything I can do to help just let me know.
Hi Annika! You seem to be
)
Hi Annika!
You seem to be the natural-born-beta-tester :-). Well, you could try to catch one this signal 11 exceptions with the ddd debugger, apparently this is still enabled on your BOINC installation but the debugger doesn't show up because BOINC runs as a daemon without X11 .
Bikeman
Hi, i received a "process
)
Hi,
i received a "process exited with code 38 (0x26, -218)"
with 4.20.
http://einsteinathome.org/task/89326361
Hartmut
RE: Hi, i received a
)
"Application caught signal 8" it says (afaik. that's an FPU error) and your previous result with 4.02 has been invalid, there might be something wrong with your box (dust bunnies in the cooler, PC placed in a corner with insufficient air flow ...)
RE: RE: Hi, i received a
)
I'm running this box as my main desktop system. It runs with about 37 to 42°C CPU temperature. Cleaned of dust recently and supported with fresh thermal grease.
Alternativly with sudoku@home there are no errors (only one for a power fail).
Might there be a difference in FPU usage?
Sorry of reporting my failure so late, but 50% of my Einstein@home WU previously have succeded.
Hartmut
Well folks, Things were
)
Well folks,
Things were going quite well here, but now I've just received my first "signal 11's". In fact, I received three consecutive ones. They're on machine number 1060000 in case anyone wants to look.
(Of course, I was also having some network problems at the time. Could that have caused this?)
Donald, that's interesting...
)
Donald, that's interesting... I was at Uni at the time and the WiFi access point had died again, so, no internet... and BOINC might have wanted to contact the server to report the two previous WUs, which had just finished... I already wondered if there was a connection but thought it was unlikely. But after reading your post, I'm not so sure any more.