I've had several of my (WinXP C2D) results validate with a Linux quorum partner. (at least three on at least two different hosts, one Athlon and one another AMD) I've seen no failures to validate for my boxes that involved a Linux partner.
I've seen no statistical care in assessing OS mismatch vs. CPU type mismatch, vs. possibly excessive CPU frequency vs. ... for contribution to invalidated successful results.
My impression was the validate issue was with early 4.13 and maybe 4.14 versions. My Linux wingman was running 4.15 on the successful one.
Another possibility was part of the problem was with the valdiator itself. They might have fixed that, but I have no way of knowing that from what I can see.
Although the bug Bernd is mention could be related to validation, I think it's more to do with execution failures. I'm still seeing a fairly high occurance of those for Linux hosts when I go rooting through WU Summary pages.
In any event, it looks like progress is being made and am still hopeful that by the end of this month they will have the execution failure solved and be well on the way to getting the app optimized again to bring the runtimes back down to something more manageable. This should help flatten out the rate curve again, even though it will be lower than it was before.
My impression was the validate issue was with early 4.13 and maybe 4.14 versions. My Linux wingman was running 4.15 on the successful one.
Another possibility was part of the problem was with the valdiator itself. They might have fixed that, but I have no way of knowing that from what I can see.
Although the bug Bernd is mention could be related to validation, I think it's more to do with execution failures. I'm still seeing a fairly high occurance of those for Linux hosts when I go rooting through WU Summary pages.
In any event, it looks like progress is being made and am still hopeful that by the end of this month they will have the execution failure solved and be well on the way to getting the app optimized again to bring the runtimes back down to something more manageable. This should help flatten out the rate curve again, even though it will be lower than it was before.
1st copy => Windows XP => Compute error : Exit status :-1073741502 (0xc0000142)..whatever
2nd copy => Linux => Success !!!!! :-)
3rd Copy => Win XP => Client Error
4th copy => Win XP => Client error (unable to create process)
5th copy => Linux => Client error (maximum disk space exceeded)
6th copy => currently running on Darwin , we'll see.
There must be a bad spell on this one ...
BRM
Wow... when WU's fight back!!
There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman
Agreed, I didn't mean to imply that problems were exclusive to Linux hosts. There are a number of seriously messed up WinBoxes out there, but that's true even when the app is 'perfect'. :-)
So it looks like there's still enough problems to around, but you Penguins seem to have been taking the brunt of them lately even though the care and feeding of Linux hosts is higher than you would see on Windows generally. Let's face it, most Windows users don't have a clue what's going on with their box (present company excepted), where you have to be fairly good with the machine just to get it setup correctly for Linux (and that's not a bad thing overall).
Although yours seem to be doing pretty well for themselves comparatively. I only saw a few failures in your listings, all with 4.14. You're clean on 4.18 so far (knock on wood).
I guess the bottom line is if your host is having trouble (doesn't matter what platform), it really has problems! ;-)
Although yours seem to be doing pretty well for themselves comparatively. I only saw a few failures in your listings, all with 4.14. You're clean on 4.18 so far (knock on wood).
I guess the bottom line is if your host is having trouble (doesn't matter what platform), it really has problems! ;-)
Alinator
I had the "signal 11" problem as well and then prevented einstein from loading the libGL.so.* library, since then (coincidence or not) everything works just fine. Not a single client error for the last week or so on six Linux CPUs.
Pleasant Surprise
)
I've had several of my (WinXP C2D) results validate with a Linux quorum partner. (at least three on at least two different hosts, one Athlon and one another AMD) I've seen no failures to validate for my boxes that involved a Linux partner.
I've seen no statistical care in assessing OS mismatch vs. CPU type mismatch, vs. possibly excessive CPU frequency vs. ... for contribution to invalidated successful results.
'Twould be good to know, however.
The problem is a bug in the
)
The problem is a bug in the linux app. See the post from Bernd here:
http://einsteinathome.org/node/191393&nowrap=true#67194
My impression was the
)
My impression was the validate issue was with early 4.13 and maybe 4.14 versions. My Linux wingman was running 4.15 on the successful one.
Another possibility was part of the problem was with the valdiator itself. They might have fixed that, but I have no way of knowing that from what I can see.
Although the bug Bernd is mention could be related to validation, I think it's more to do with execution failures. I'm still seeing a fairly high occurance of those for Linux hosts when I go rooting through WU Summary pages.
In any event, it looks like progress is being made and am still hopeful that by the end of this month they will have the execution failure solved and be well on the way to getting the app optimized again to bring the runtimes back down to something more manageable. This should help flatten out the rate curve again, even though it will be lower than it was before.
Alinator
RE: My impression was the
)
Well, Windows wingmen can be a pest as well :-)
See this one:
http://einsteinathome.org/workunit/33436306
1st copy => Windows XP => Compute error : Exit status :-1073741502 (0xc0000142)..whatever
2nd copy => Linux => Success !!!!! :-)
3rd Copy => Win XP => Client Error
4th copy => Win XP => Client error (unable to create process)
5th copy => Linux => Client error (maximum disk space exceeded)
6th copy => currently running on Darwin , we'll see.
There must be a bad spell on this one ...
BRM
RE: Well, Windows wingmen
)
Wow... when WU's fight back!!
There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman
RE: There must be a bad
)
Agreed, I didn't mean to imply that problems were exclusive to Linux hosts. There are a number of seriously messed up WinBoxes out there, but that's true even when the app is 'perfect'. :-)
So it looks like there's still enough problems to around, but you Penguins seem to have been taking the brunt of them lately even though the care and feeding of Linux hosts is higher than you would see on Windows generally. Let's face it, most Windows users don't have a clue what's going on with their box (present company excepted), where you have to be fairly good with the machine just to get it setup correctly for Linux (and that's not a bad thing overall).
Although yours seem to be doing pretty well for themselves comparatively. I only saw a few failures in your listings, all with 4.14. You're clean on 4.18 so far (knock on wood).
I guess the bottom line is if your host is having trouble (doesn't matter what platform), it really has problems! ;-)
Alinator
RE: Although yours seem to
)
I had the "signal 11" problem as well and then prevented einstein from loading the libGL.so.* library, since then (coincidence or not) everything works just fine. Not a single client error for the last week or so on six Linux CPUs.
CU
BRM