This may or may not be related, but today I just installed Windows 2003 Enterprise Server(MS Promotional copy) on a Dual AMD MP2200. I then installed Service Pack 1. However, since installing SP1, I find that the only Boinc version that actually runs is 4.19. Every other version I've tried(4.45, 5.2.2, and 5.2.13) all fail with same lockups described by the OP. I've not installed any other MS patches yet.
Edit: I do have Acrobat Reader 7.0.0 installed and all is working fine.
I'm glad I'm not the only one seeing this problem. I guess I'll try installing 4.19.
This may or may not be related, but today I just installed Windows 2003 Enterprise Server(MS Promotional copy) on a Dual AMD MP2200. I then installed Service Pack 1. However, since installing SP1, I find that the only Boinc version that actually runs is 4.19. Every other version I've tried(4.45, 5.2.2, and 5.2.13) all fail with same lockups described by the OP. I've not installed any other MS patches yet. ...
I see that 4.45 is the oldest version available for Windows on the download page. Is there somewhere else I can get 4.19, to try it?
I see that 4.45 is the oldest version available for Windows on the download page. Is there somewhere else I can get 4.19, to try it?
idhcria,
Unfortunately, 4.19 is incompatible with the new Einstein apps needed to process the new albert WUs which are about all of those being sent out these days. :-(
Sorry.
Michael
microcraft
"The arc of history is long, but it bends toward justice" - MLK
I see that 4.45 is the oldest version available for Windows on the download page. Is there somewhere else I can get 4.19, to try it?
Unfortunately, 4.19 is incompatible with the new Einstein apps needed to process the new albert WUs which are about all of those being sent out these days. :-(
Michael
Michael, I'm not sure if this is true, but I don't think so. The main disadvantage of 4.19 is that it does not fix the 'ghost workunit' problem, and contains other known bugs that have been subsequently fixed. But I think it still works reasonably well (though I personally prefer and use the 5.2.X clients).
Michael, I'm not sure if this is true, but I don't think so. The main disadvantage of 4.19 is that it does not fix the 'ghost workunit' problem, and contains other known bugs that have been subsequently fixed. But I think it still works reasonably well (though I personally prefer and use the 5.2.X clients).
Dr. Allen,
I'm not sure what gave me that impression (about 4.19), and pain has kept me from searching for whatever it was that did. Stick questioned me on the subject almost immediately, via email, and I promised to try to run it down. In the absence of info to the contrary, I'll bow to your superior knowledge. Chalk it up to a rogue neuron firing off and, after the impulse bounced around the skull, received the echo and interpreted it as affirmation. :-)
Respects,
Michael
microcraft
"The arc of history is long, but it bends toward justice" - MLK
So far I've not encountered any issue(that I'm aware of) using the 4.19 client on my 2003 Server rig. Just yesterday, I finished patching the server current will all of the available MS updates, and I can report that I'm still unable to upgrade the Boinc client to a newer version. :(
I don't know what the problem may be.
"Chance is irrelevant. We will succeed."
- Seven of Nine
So far I've not encountered any issue(that I'm aware of) using the 4.19 client on my 2003 Server rig. Just yesterday, I finished patching the server current will all of the available MS updates, and I can report that I'm still unable to upgrade the Boinc client to a newer version. :(
I don't know what the problem may be.
GRRO,
In looking over the machine in question, I just noticed something quite odd - the benchmarks. v4.19 was very generous on benchmarks, especially the integer side, if I recall correctly, and AMDs' integer benchmarks are always]/i] well above their floating point benchmarks. The Server Edition MP 2200, however, shows a very low int bench, [i]lower than the float bench!, and way below the float bench of your other MP 2200. Not only is BOINC not correctly identifying your processor (which most likely means that Windows also does not correctly identify it), but the extremely low int bench severely impacts your credit claim - it should be claiming much higher, by at least a factor of 2x.
I have no idea whether that anomoly affects upgrade installation, though my guess is that it does not.
Michael
microcraft
"The arc of history is long, but it bends toward justice" - MLK
So far I've not encountered any issue(that I'm aware of) using the 4.19 client on my 2003 Server rig. Just yesterday, I finished patching the server current will all of the available MS updates, and I can report that I'm still unable to upgrade the Boinc client to a newer version. :(
I don't know what the problem may be.
So, after several postings, it sounds like the 4.19 client does still work with 2003 server (and none of the newer versions do). If that is true, can I download the 4.19 client from somewhere and use it until a newer version that works is available?
I found the page to download version 4.19 on the einstein web site via a web search and installed in on my 2003 server. That runs OK. But it started back on the work that it downloaded 1/5/2006 (when it was last connected), that has a report deadline on 2006-01-19 (in other words, it is past its deadline). I don't see a way to abort that work in that version. Am I not seeing how to do that or do I need to just wait until it completes (about 7.5 hours from now). Telling the project to update didn't change anything.
Never mind. I did a complete uninstall and deleted the contents of the BOINC directory, then reinstalled v4.19 and attached to the project again (then merged the old computer id with the new one). Things seem to be working as expected now.
I'll be looking for a newer version that works on 2003 server PCs.
RE: This may or may not be
)
I'm glad I'm not the only one seeing this problem. I guess I'll try installing 4.19.
RE: This may or may not be
)
I see that 4.45 is the oldest version available for Windows on the download page. Is there somewhere else I can get 4.19, to try it?
RE: I see that 4.45 is the
)
idhcria,
Unfortunately, 4.19 is incompatible with the new Einstein apps needed to process the new albert WUs which are about all of those being sent out these days. :-(
Sorry.
Michael
microcraft
"The arc of history is long, but it bends toward justice" - MLK
RE: RE: I see that 4.45
)
Michael, I'm not sure if this is true, but I don't think so. The main disadvantage of 4.19 is that it does not fix the 'ghost workunit' problem, and contains other known bugs that have been subsequently fixed. But I think it still works reasonably well (though I personally prefer and use the 5.2.X clients).
Director, Einstein@Home
RE: Michael, I'm not sure
)
Dr. Allen,
I'm not sure what gave me that impression (about 4.19), and pain has kept me from searching for whatever it was that did. Stick questioned me on the subject almost immediately, via email, and I promised to try to run it down. In the absence of info to the contrary, I'll bow to your superior knowledge. Chalk it up to a rogue neuron firing off and, after the impulse bounced around the skull, received the echo and interpreted it as affirmation. :-)
Respects,
Michael
microcraft
"The arc of history is long, but it bends toward justice" - MLK
So far I've not encountered
)
So far I've not encountered any issue(that I'm aware of) using the 4.19 client on my 2003 Server rig. Just yesterday, I finished patching the server current will all of the available MS updates, and I can report that I'm still unable to upgrade the Boinc client to a newer version. :(
I don't know what the problem may be.
"Chance is irrelevant. We will succeed."
- Seven of Nine
RE: So far I've not
)
GRRO,
In looking over the machine in question, I just noticed something quite odd - the benchmarks. v4.19 was very generous on benchmarks, especially the integer side, if I recall correctly, and AMDs' integer benchmarks are always]/i] well above their floating point benchmarks. The Server Edition MP 2200, however, shows a very low int bench, [i]lower than the float bench!, and way below the float bench of your other MP 2200. Not only is BOINC not correctly identifying your processor (which most likely means that Windows also does not correctly identify it), but the extremely low int bench severely impacts your credit claim - it should be claiming much higher, by at least a factor of 2x.
I have no idea whether that anomoly affects upgrade installation, though my guess is that it does not.
Michael
microcraft
"The arc of history is long, but it bends toward justice" - MLK
RE: So far I've not
)
So, after several postings, it sounds like the 4.19 client does still work with 2003 server (and none of the newer versions do). If that is true, can I download the 4.19 client from somewhere and use it until a newer version that works is available?
Leland.
I found the page to download
)
I found the page to download version 4.19 on the einstein web site via a web search and installed in on my 2003 server. That runs OK. But it started back on the work that it downloaded 1/5/2006 (when it was last connected), that has a report deadline on 2006-01-19 (in other words, it is past its deadline). I don't see a way to abort that work in that version. Am I not seeing how to do that or do I need to just wait until it completes (about 7.5 hours from now). Telling the project to update didn't change anything.
Thanks,
Leland.
Never mind. I did a complete
)
Never mind. I did a complete uninstall and deleted the contents of the BOINC directory, then reinstalled v4.19 and attached to the project again (then merged the old computer id with the new one). Things seem to be working as expected now.
I'll be looking for a newer version that works on 2003 server PCs.
Thanks,
Leland.