BOINC V5.8.11 Computer Info/Listings Bug
19 Feb 2007 21:59:57 UTC
Topic 192458
(moderation:
I recently switched all my Systems to V5.8.11 and noted the following :
- Computer Listings now include lots of additional CPU Information (while useful for BOINC itself to have it, does it really belong here ?)
Differences across Operating Systems :
Windows2000 SP4
AuthenticAMD
AMD Athlon(tm) XP 3000+ [x86 Family 6 Model 10 Stepping 0] [fpu tsc sse 3dnow mmx]
Linux (2.4.x upto 2.6.15 Kernel, various Distributions)
AuthenticAMD
AMD Athlon(tm) XP 3000+ [fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow]
Linux (2.6.18 Kernel - Fedora Core 6)
AuthenticAMD
----------------------
Windows Systems list Family/Model/Stepping, but only a few CPU Extensions.
Linux Systems upto 2.6.15 Kernel list only the CPU Extensions, but there it lists just about... everything.
Linux Systems with 2.6.18 Kernel however, show only CPU Manufacturer and that's it (?)
(have this on two separate Systems running FC6)
Note :
BOINCview V1.4.2b which oversees the entire Network sees the CPU Identifiction correctly on the 2.6.18 Systems, so BOINC 5.8.11 is reading/using it there - it just fails to report it into the Computer listing.
-----------------
Apart from the additional Info completely clogging up the Computer Listing pages, I believe this is a Bug worth looking into.
And while doing, BOINC Devs could further improve Linux BOINC Benchmark results by about 14% to match their Win32 counterpart (it's beyond me why they still fail to achieve within 5% results across Operating Systems after more than 2 years).
BOINC V5.8.11 Computer Info/Listings Bug
)
Reasons different OSs show different information is it's what the OS can use internally, and understands.
These are here to help the developers ability to create platform specific software. I believe on the next few versions it will be separated on it's own section, and not included as part of the processor. BOINC is continuous in development, so it will go through more iterations.
On a SuSE Linux 10.1 BOINC
)
On a SuSE Linux 10.1 BOINC 5.8.8 worked beautifully. Since switching to 5.8.11, as suggested, I get an error message such as:
Task xxx exited with zero status but not 'finished' file
This happens with both Einstein and SETI. Crunching goes on. Einstein graphics windows works when called, SETI only erratically.
Tullio
RE: On a SuSE Linux 10.1
)
Known issue, error is harmless. DO NOT RESET.
What is happening is something is interrupting BOINC for more than a certain period of time. I think they changed this period of time to a shorter amount. I can tell you that NNTP calls, and a few other things like that are sometimes the cause. In Windows OS it's been seen happening when it does NNTP, AV scans, etc.
BOINC is a continuous development.
RE: I recently switched all
)
@FalconFly,
I also have installed Boinc 5.8.11, installed on a new AMD Opteron 285 with Fedora Core 6 but during installation it upgraded from 2.6.18 to 2.6.19. My computer now displays everything, taking up heaps of space on the computer listings.
I agree with you with regard to benchmarks, I still don't know why it has taken so many Boinc versions (and years) to give decent benchmarks for Linux machines, it could of saved lots of attempts to cheat if done earlier.
RE: Known issue, error is
)
Thanks Pooh Bear 27 for the info. Another funny thing: any BOINC update increases the firepower of my Pentium II Deschutes. Now It boasts 334 MIPS floating and 457 integer. Wow!
Tullio
RE: @FalconFly, I also have
)
Okidok, so I think either the next Kernel or BOINC Version will likely and automatically fix it :)
(not sure about the Benchmark though, it'll likely take them another 2 years *lol*)
RE: RE: @FalconFly, I
)
You guys do know that you never had to put up with it. You could have used other means to determine more appropriate benchmark values, "baked" the client_state file to reflect that, and told the CC not to run the BM's anymore with a command line switch when starting up.
I will grant you though, it is nice that the BM's do seem to be getting closer to reality on other platforms (or at least more comparable) and definitely helps to even the playing field for projects using BM based scoring without providing motivation for things like CWCP'ing and even less desirable behaviour. ;-)
Alinator
Since starting BOINC 5.8.11
)
Since starting BOINC 5.8.11 my stderr.txt has grown to a larger file with the following lines repeated many times
2007-02-22 07:49:21.4342 [detail]: Checkpointed state (10093/74036771/1483484).
2007-02-22 07:50:32.8713 [detail]: Checkpointed state (10099/74054860/1483843).
2007-02-22 07:51:44.2133 [detail]: Checkpointed state (10105/74065209/1484050).
2007-02-22 07:52:55.6030 [detail]: Checkpointed state (10111/74065209/1484050).
It seems that a check is done every 1'11". Is this correct?
Tullio
@Alinator I want to keep
)
@Alinator
I want to keep everything at stock settings, also the dynamic Benchmark results are an early indication of Problems with a machine for me (the numbers I receive over the Network is the only thing I know about the machines' state)
--------
Overall, my hopes that a new BOINC Version would solve the small annoyance weren't fulfilled.
BOINC V5.8.15 still only reports CPU Vendor on the two (soon three) affected machines but nothing else.
...guess I'll have to live with a few "Incognito" machines for a while ;)