All Over (Dec 10 2008)

Message boards : Technical News : All Over (Dec 10 2008)
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile Matt Lebofsky
Volunteer moderator
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 1 Mar 99
Posts: 1444
Credit: 957,058
RAC: 0
United States
Message 838672 - Posted: 11 Dec 2008, 0:21:20 UTC

During the wee hours this morning our upload server (bruno) froze. We are still unsure why, but recovery was a comedy of errors. Jeff was already about to power cycle it (having little other choice given the unresponsive console) when I got in around 8am. After rebooting bruno failed to mount its result storage drives due to some kind of mdadm mismanagement. This forced us into a read-only please-fsck-your-drives mode. The drives, outside of pointless resyncing due to hard power cycle, were fine - they didn't need to be fsck'ed. Still, being root (/) was read-only we couldn't edit /etc/fstab to prevent this from happening again upon every reboot.

So I tried to get it into a real single user mode to make such an edit - all I wanted to do was comment out that one mount line. However, thus started a series of about 8 consecutive reboots, each taking about five minutes, and all wastes of time due to a typo or an unresponsive kvm. I ultimately gave up and booted from DVD in "rescue mode" where I could finally make the fstab edit. Finally all was well with the mount (which I did on the command line), but then I had all kinds of network errors with the system. More tweaks, more reboots... Long story short this server is being held together with figurative duct tape at the moment. We'll get it all sorted out later.

Jeff and I also worked together to get the remaining pieces of the "donation drive" in place, such as it is. I'm sending out test e-mails out now, and will probably start sending in earnest on Friday. Please send all questions/comments about our fundraising efforts to the principal investigators (Dan, Dave, Eric). I am simply implementing the technical aspects of this endeavor, though I would like to point out we finally updated the text on the plans page.

By the way.. did anybody notice this?

- Matt
-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude
ID: 838672 · Report as offensive
Profile Dr. C.E.T.I.
Avatar

Send message
Joined: 29 Feb 00
Posts: 16019
Credit: 794,685
RAC: 0
United States
Message 838676 - Posted: 11 Dec 2008, 0:37:42 UTC


. . . Kudo's to Each of You @ Berkeley - 'a job Well done' - Thanks

@ Matt - Thanks for the Updates and the Link to 'CUDA-capable GPU'


BOINC Wiki . . .

Science Status Page . . .
ID: 838676 · Report as offensive
DJStarfox

Send message
Joined: 23 May 01
Posts: 1066
Credit: 1,226,053
RAC: 2
United States
Message 838720 - Posted: 11 Dec 2008, 3:57:56 UTC - in response to Message 838672.  

Did you try this earlier?
# mount -o remount,rw /

It's a good thing you updated the plans page before the Xmas season "donation" email newsletter goes out. :)
ID: 838720 · Report as offensive
Profile popandbob
Volunteer tester

Send message
Joined: 19 Mar 05
Posts: 551
Credit: 4,673,015
RAC: 0
Canada
Message 838751 - Posted: 11 Dec 2008, 6:21:05 UTC

About the gpu app...
5-10x faster than stock or opti app?
Also is it only in beta right now?
~BoB


Do you Good Search for Seti@Home? http://www.goodsearch.com/?charityid=888957
Or Good Shop? http://www.goodshop.com/?charityid=888957
ID: 838751 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 838983 - Posted: 11 Dec 2008, 19:44:44 UTC - in response to Message 838751.  
Last modified: 11 Dec 2008, 19:44:52 UTC

Rudimentary, very preliminary, standalone tests here say, on my rather ordinary 8600GT, about 6x faster than AKv8SSE4.1 running on my dual core E8400@3.6GHz... naturally though I'd like to be running all three apps at the same time, but will have to poke around with that another day ;D
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 838983 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 839034 - Posted: 11 Dec 2008, 22:09:25 UTC - in response to Message 838983.  

Rudimentary, very preliminary, standalone tests here say, on my rather ordinary 8600GT, about 6x faster than AKv8SSE4.1 running on my dual core E8400@3.6GHz... naturally though I'd like to be running all three apps at the same time, but will have to poke around with that another day ;D


Withdrawn... Was accurately measuring CPU time, which of course the GPU does not use much of :D .. wall clock time is somewhat longer here.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 839034 · Report as offensive
Profile Jack Zhang
Volunteer tester
Avatar

Send message
Joined: 2 Jul 06
Posts: 206
Credit: 6,142,449
RAC: 0
Canada
Message 839036 - Posted: 11 Dec 2008, 22:16:47 UTC - in response to Message 839034.  
Last modified: 11 Dec 2008, 22:59:23 UTC

Retracted previous statement.

Right now I'm not seeing much of a difference between this and AK V8 SSSE3 in the first half of processing, but it does go 5-10x faster in the end half. But remember, this is a different project from the official S@H. If you're encouraged to do it, your RAC will count for the Beta project, not the official one.

And honestly, why is it asking for so much work?

Folding optimized their engine to use 100% of the GPU power, nearly using up every processing thread related to the GPU. I'm not seeing that right now for the first half of WU processing taking 24 minutes but then sharply speeds up at the end to finish the WU in 30 minutes. There should be optimizations for the entire length of the WU processing.
What if Fiction was Fact and Fact was Fiction and vice versa?
ID: 839036 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19012
Credit: 40,757,560
RAC: 67
United Kingdom
Message 839077 - Posted: 12 Dec 2008, 0:28:05 UTC - in response to Message 839036.  

Retracted previous statement.

Right now I'm not seeing much of a difference between this and AK V8 SSSE3 in the first half of processing, but it does go 5-10x faster in the end half. But remember, this is a different project from the official S@H. If you're encouraged to do it, your RAC will count for the Beta project, not the official one.

And honestly, why is it asking for so much work?

Folding optimized their engine to use 100% of the GPU power, nearly using up every processing thread related to the GPU. I'm not seeing that right now for the first half of WU processing taking 24 minutes but then sharply speeds up at the end to finish the WU in 30 minutes. There should be optimizations for the entire length of the WU processing.

Shouldn't your concerns and question be on the SetiBeta pages?
ID: 839077 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 839084 - Posted: 12 Dec 2008, 0:52:17 UTC - in response to Message 839036.  
Last modified: 12 Dec 2008, 0:52:48 UTC

There should be optimizations for the entire length of the WU processing.

There should be, but coding is work, and what they're running over there is a beta version -- it isn't necessarily done.
ID: 839084 · Report as offensive
Haos.PL
Volunteer tester

Send message
Joined: 18 Mar 04
Posts: 63
Credit: 3,268,546
RAC: 0
Poland
Message 839089 - Posted: 12 Dec 2008, 1:12:17 UTC

I`m not sure if i`m doing it wrong. Was using AKv8SSE4.1 with APv5.0SSE3 optimized apps. I installed new BOINC and my GPU was detected propely. Is it enough or do i have to revert to stock apps that come with the 6.4.5 BOINC?
ID: 839089 · Report as offensive
Profile Jack Zhang
Volunteer tester
Avatar

Send message
Joined: 2 Jul 06
Posts: 206
Credit: 6,142,449
RAC: 0
Canada
Message 839091 - Posted: 12 Dec 2008, 1:15:34 UTC - in response to Message 839089.  
Last modified: 12 Dec 2008, 1:21:32 UTC

You have to join the Beta program, attach your BOINC Manager to the URL "http://setiweb.ssl.berkeley.edu/beta/" to join the beta. Then you'll have access to the CUDA program.

It's also worth noting that your login doesn't cross-over between the official and the beta. You'll have to create a new account.

----

Retracted previous statement.

Right now I'm not seeing much of a difference between this and AK V8 SSSE3 in the first half of processing, but it does go 5-10x faster in the end half. But remember, this is a different project from the official S@H. If you're encouraged to do it, your RAC will count for the Beta project, not the official one.

And honestly, why is it asking for so much work?

Folding optimized their engine to use 100% of the GPU power, nearly using up every processing thread related to the GPU. I'm not seeing that right now for the first half of WU processing taking 24 minutes but then sharply speeds up at the end to finish the WU in 30 minutes. There should be optimizations for the entire length of the WU processing.

Shouldn't your concerns and question be on the SetiBeta pages?


Can't post there, "You must reach this certain credit limit" protection for that message board is on.
What if Fiction was Fact and Fact was Fiction and vice versa?
ID: 839091 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 839284 - Posted: 12 Dec 2008, 18:27:39 UTC

Matt, is the GPU app in Beta truly being tested with Adaptive Replication turned on or is it just that the <target_nresults> in the APP record for setiathome_enhanced has been left at 2 so testers can't see workunit details? If that entry, and the similar one for astropulse, could be reduced to 1 or 0 then participants would have a better view of what's happening.
                                                                Joe
ID: 839284 · Report as offensive
Profile Matt Lebofsky
Volunteer moderator
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 1 Mar 99
Posts: 1444
Credit: 957,058
RAC: 0
United States
Message 839285 - Posted: 12 Dec 2008, 18:37:02 UTC - in response to Message 839284.  

For the record, I have *zero* idea what's going on with the CUDA/SETI@home beta project, except that it now exists and I was just merely pointing out its existence. Please forward all comments/questions to the beta forum, which I assume is being read by the beta developers. Sorry for any confusion.

- Matt
-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude
ID: 839285 · Report as offensive
Filipe

Send message
Joined: 12 Aug 00
Posts: 218
Credit: 21,281,677
RAC: 20
Portugal
Message 839311 - Posted: 12 Dec 2008, 19:48:47 UTC

Matt, is it possible to split the "science status page" between multibeam and astropulse in the same way it has been done for the "server status page"?
ID: 839311 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 839529 - Posted: 13 Dec 2008, 12:06:33 UTC - in response to Message 839285.  

For the record, I have *zero* idea what's going on with the CUDA/SETI@home beta project, except that it now exists and I was just merely pointing out its existence. Please forward all comments/questions to the beta forum, which I assume is being read by the beta developers. Sorry for any confusion.

- Matt

Matt,

New testers being drawn into the Beta project for the first time by the CUDA application are finding that they can't participate in the Beta SETI@home Enhanced Message Board because it has a minimum credit level set. They are able to post in the Astropulse Message Board because the credit restriction has been eliminated there because of similar problems when Astropulse first went into testing. Could you pass a message along to the relevent committee, please?

At the same time, could you also pass on a message to say that the ongoing "single redundancy mode" test at Beta is making CUDA testing unecessarily complicated. Could you suggest they review that one too, please?
ID: 839529 · Report as offensive
Profile Dr. C.E.T.I.
Avatar

Send message
Joined: 29 Feb 00
Posts: 16019
Credit: 794,685
RAC: 0
United States
Message 839533 - Posted: 13 Dec 2008, 12:22:11 UTC - in response to Message 839529.  
Last modified: 13 Dec 2008, 12:27:47 UTC

For the record, I have *zero* idea what's going on with the CUDA/SETI@home beta project, except that it now exists and I was just merely pointing out its existence. Please forward all comments/questions to the beta forum, which I assume is being read by the beta developers. Sorry for any confusion.

- Matt

Matt,

New testers being drawn into the Beta project for the first time by the CUDA application are finding that they can't participate in the Beta SETI@home Enhanced Message Board because it has a minimum credit level set. They are able to post in the Astropulse Message Board because the credit restriction has been eliminated there because of similar problems when Astropulse first went into testing. Could you pass a message along to the relevent committee, please?

At the same time, could you also pass on a message to say that the ongoing "single redundancy mode" test at Beta is making CUDA testing unecessarily complicated. Could you suggest they review that one too, please?


. . . Richard - i believe there are Forums etc @ CUDA that one can Post to there

> hope your day is goin' well Sir . . .
BOINC Wiki . . .

Science Status Page . . .
ID: 839533 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 839535 - Posted: 13 Dec 2008, 12:30:39 UTC - in response to Message 839533.  

. . . Richard - i believe there are Forums etc @ CUDA that one can Post to there.

I doubt the Berkeley SETI Beta team will want to use that forum to retrieve bug reports relating to their own application.
ID: 839535 · Report as offensive
Profile Dr. C.E.T.I.
Avatar

Send message
Joined: 29 Feb 00
Posts: 16019
Credit: 794,685
RAC: 0
United States
Message 839541 - Posted: 13 Dec 2008, 12:51:31 UTC - in response to Message 839535.  

. . . Richard - i believe there are Forums etc @ CUDA that one can Post to there.

I doubt the Berkeley SETI Beta team will want to use that forum to retrieve bug reports relating to their own application.


. . . ok - and there's a lot of Postings by Jord here that may well enable one to get to the 'right individuals' ?

hope this can help a bit ;)

< Contact Jord [Ageless] ?


BOINC Wiki . . .

Science Status Page . . .
ID: 839541 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 839545 - Posted: 13 Dec 2008, 13:03:12 UTC - in response to Message 839541.  

[. . . ok - and there's a lot of Postings by Jord here that may well enable one to get to the 'right individuals' ?

The 'right individuals' are those who work on the SETI applications, not the BOINC developers. And it's not me that wants to talk to them - unlike you, I'm already a 'Volunteer tester' with an account and full posting rights. It's the new participants attracted by Matt's post that are having the difficulty.

< Contact Jord [Ageless] ?

I'm already talking to him.
ID: 839545 · Report as offensive
Profile Dr. C.E.T.I.
Avatar

Send message
Joined: 29 Feb 00
Posts: 16019
Credit: 794,685
RAC: 0
United States
Message 839549 - Posted: 13 Dec 2008, 13:20:32 UTC - in response to Message 839545.  
Last modified: 13 Dec 2008, 13:21:06 UTC

[. . . ok - and there's a lot of Postings by Jord here that may well enable one to get to the 'right individuals' ?

The 'right individuals' are those who work on the SETI applications, not the BOINC developers. And it's not me that wants to talk to them - unlike you, I'm already a 'Volunteer tester' with an account and full posting rights. It's the new participants attracted by Matt's post that are having the difficulty.

< Contact Jord [Ageless] ?

I'm already talking to him.


saw that too as well as here: s@h on GPU CUDA now?!

well - hopefully all gets answered for You Sir and have a great day / weekend

ps - anybody know how to spell wiki? [unofficially - of course] ;))
BOINC Wiki . . .

Science Status Page . . .
ID: 839549 · Report as offensive

Message boards : Technical News : All Over (Dec 10 2008)


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.