Posts by staze


log in
1) Message boards : Technical News : Carolyn = replica (Nov 18 2010) (Message 1049449)
Posted 1341 days ago by Profile staze
Also, it would be cool to see bonnie++ stats for these machines, especially compared to thumper, or the new bruno, etc.
2) Message boards : Technical News : Carolyn = replica (Nov 18 2010) (Message 1049448)
Posted 1341 days ago by Profile staze
How are the RAIDs/MySQL configured? RAID10? Separate the indexes/journals off to another RAID?
3) Message boards : Number crunching : Nehalem Xserve not using all processors (Message 953740)
Posted 1685 days ago by Profile staze
Yup, that did it.

http://www.spy-hill.net/help/apple/SharedMemory.html

Would be cool if BOINC was smart enough to say "Shared memory settings are insufficient, please consider increasing them....".

Now running on all 16 logical cores. Sweet!

Thanks!
4) Message boards : Number crunching : Nehalem Xserve not using all processors (Message 953737)
Posted 1685 days ago by Profile staze
First 20 lines:

Sun Dec 6 21:55:16 2009 Starting BOINC client version 6.10.21 for x86_64-apple-darwin
Sun Dec 6 21:55:16 2009 log flags: file_xfer, sched_ops, task
Sun Dec 6 21:55:16 2009 Libraries: libcurl/7.19.4 OpenSSL/0.9.7l zlib/1.2.3 c-ares/1.6.0
Sun Dec 6 21:55:16 2009 Data directory: /Library/Application Support/BOINC Data
Sun Dec 6 21:55:16 2009 Processor: 16 GenuineIntel Intel(R) Xeon(R) CPU E5520 @ 2.27GHz [x86 Family 6 Model 26 Stepping 5]
Sun Dec 6 21:55:16 2009 Processor features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH DS ACPI MMX FXSR SSE SSE2 SS HTT TM SSE3 MON DSCPL VMX EST TM2 SSSE3 CX16 TPR PDCM SSE4.1 SSE4.2 POPCNT
Sun Dec 6 21:55:16 2009 OS: Darwin: 9.8.0
Sun Dec 6 21:55:16 2009 Memory: 12.00 GB physical, 96.22 GB virtual
Sun Dec 6 21:55:16 2009 Disk: 148.73 GB total, 95.98 GB free
Sun Dec 6 21:55:16 2009 Local time is UTC -8 hours
Sun Dec 6 21:55:16 2009 No usable GPUs found
Sun Dec 6 21:55:17 2009 Not using a proxy
Sun Dec 6 21:55:17 2009 SETI@home URL http://setiathome.berkeley.edu/; Computer ID 5030462; resource share 100
Sun Dec 6 21:55:17 2009 SETI@home General prefs: from SETI@home (last modified 06-Dec-2009 21:39:47)
Sun Dec 6 21:55:17 2009 SETI@home Computer location: home
Sun Dec 6 21:55:17 2009 SETI@home General prefs: no separate prefs for home; using your defaults
Sun Dec 6 21:55:17 2009 Reading preferences override file
Sun Dec 6 21:55:17 2009 Preferences:
Sun Dec 6 21:55:17 2009 max memory usage when active: 6144.00MB
Sun Dec 6 21:55:17 2009 max memory usage when idle: 11059.20MB
Sun Dec 6 21:55:17 2009 max disk usage: 3.00GB
Sun Dec 6 21:55:17 2009 (to change, visit the web site of an attached project,
Sun Dec 6 21:55:17 2009 or click on Preferences)

In computing prefs, I have:

24 processors, 100%, and 100%.

There are 8 tasks running, the rest say: "waiting for shared memory"...

Now that I google that... I find I am an idiot. I will post back when I have a result.
5) Message boards : Number crunching : Nehalem Xserve not using all processors (Message 953076)
Posted 1688 days ago by Profile staze
Hello,

I've looked and looked at this point, and can't seem to find someone having a similar issue. I'm stress testing a brand new 2009 Xserve, that's got 8-cores, with HT, and no matter what, while BOINC says "16 processors" and startup, and I have all my prefs set to use 100% and 24 processors, etc... I cannot get BOINC/SETI to use more than 8 processors.

It sounds like this should be doable... yet I can't figure out how. The biggest issue with this is that, it doesn't seem to favor the real cores vs. the "logical" cores, as the OS doesn't seem to really know the difference.

Anyone have any ideas?
6) Message boards : Technical News : Returning (Jul 14 2008) (Message 782791)
Posted 2199 days ago by Profile staze
So, I guess the question becomes either... could you place that index on another drive/partition, and therefore give it room to grow, without fragmenting everything else... or, have you tried raw device support within mysql, and forgo the filesystem altogether?

how big did the index get? we're obviously talking millions (if not, hundreds of millions) of entries...
7) Message boards : Number crunching : "No work available (daily quota exceeded)" - what gives (Message 1385)
Posted 3680 days ago by Profile staze
I seriously doubt either. I have no network trouble from anything else... I'm on cable, so I doubt there's a timeout unless the data being sent back is huge and it wants it a lot faster than my 256kbit/sec connection can handle upstream.

Just for s's and g's, I'll reinstall boinc... but somehow I doubt that'll fix it.

the error is exactly that, an access violation... but of what?

will post back when I know more... at this point though, I am starting to want to go back to classic until the next rev of boinc. or just switch back to fah...

-Staze
8) Message boards : Number crunching : "No work available (daily quota exceeded)" - what gives (Message 1364)
Posted 3680 days ago by Profile staze
> > I have done no where near 50 units in a day (I only have 1 machine
> working on
> > it right now), and I'm getting a quota exceeded message. This seems
> entirely
> > bogus. I have tried several settings for the amount of work to be
> > downloaded, and it makes no difference.
>
> If you go into your account and bring up your computer and look at the
> results, you will see that you have been assigned 50 work units. You
> successfully completed 1 of them, and you had client errors on the other 49.
> Something is wrong on your system.
>
k... then where is the verbose error? I see nothing in stderr.txt besides the quota issue... and in stdout, all I see is:

2004-06-24 13:08:31 [---] Fewer active results than CPUs; requesting more work
2004-06-24 13:08:31 [SETI@home] Requesting 86399 seconds of work
2004-06-24 13:08:31 [SETI@home] Sending request to scheduler: http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
2004-06-24 13:08:35 [SETI@home] Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
2004-06-24 13:08:35 [SETI@home] Message from server: Not sending work - last RPC too recent: 64 sec

That doesn't seem like something that is an issue with me. If the best I can get is "client error" it becomes a little hard to diagnose a problem. And the fact that I've done thousands of units with classic, and it seems that really all boinc is doing is running classic, then it seems even more illogical that the problem is anything else but a bug.

If you could point me in the direction of some place to get some real locations to look for errors from the client, then cool... if not, then again, lack of documentation, and the project probably shouldn't be public.
9) Message boards : Number crunching : "No work available (daily quota exceeded)" - what gives (Message 1341)
Posted 3680 days ago by Profile staze
I have done no where near 50 units in a day (I only have 1 machine working on it right now), and I'm getting a quota exceeded message. This seems entirely bogus. I have tried several settings for the amount of work to be downloaded, and it makes no difference.

if you ask me, there's a bug in the system, or it's including the other machines I have running classic in my daily quota.

Incase you need it, my host ID is 14616. I am running 3.19 for windows.

Please fix this, either document the project better, or start posting errors you're seeing because at this point, I don't see how the project went public.

-Staze


Copyright © 2014 University of California