Now that v7 has rolled out..........

Message boards : Number crunching : Now that v7 has rolled out..........
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 14 · 15 · 16 · 17 · 18 · Next

AuthorMessage
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1375109 - Posted: 2 Jun 2013, 4:33:08 UTC

It appears that the maximum run time setting of V7 units may be set a bit tight.

For install and testing purposes. I started V7 crunching running only one unit per GPU. When I was happy all was working properly I switched back to 2 per GPU.

As a result of this I trashed 60 odd units with -177 errors, all timing out after ~250 seconds. There has been no errors with units downloaded since that batch.

I thought the timeout was 10x the expected run time but these failures were only 2x ?

T.A.
ID: 1375109 · Report as offensive
Profile Vipin Palazhi
Avatar

Send message
Joined: 29 Feb 08
Posts: 286
Credit: 167,386,578
RAC: 0
India
Message 1375113 - Posted: 2 Jun 2013, 4:54:09 UTC - in response to Message 1374842.  
Last modified: 2 Jun 2013, 4:55:07 UTC


You might want to double check what that AV is doing.

Since some of the tasks do manage to start and then die to that error another possibility is that you have permissions problems in the project directory and/or in slots directory. Check those too.

edit: That same computer is having problems with Einstein (host there) tasks too. The error there is:

The system cannot find the file specified.
(0x2) - exit code 2 (0x2)

This is really strange as this is my main rig and has had no changes in it. The issue began after the v7 release. I have now reset the rig and tuned down the AV scans. File and folder permissions are all ok. The system has downloaded a fresh batch of WUs along with the executables, seems to be crunching along. Will report back later.
ID: 1375113 · Report as offensive
Profile Francesco Forti
Avatar

Send message
Joined: 24 May 00
Posts: 334
Credit: 204,421,005
RAC: 15
Switzerland
Message 1375277 - Posted: 2 Jun 2013, 10:22:06 UTC

I repeat here part of a discussion started in the news forum

1) All the Stock v7 GPU apps are Lunatics apps now!
2) You did go and enable 'SETI@home v7: yes' in your project preferences didn't you?
3) No, you copy and paste it from one of the posts in Number Crunching, and edit it to your requirements.



1: great! But Only GPU? Why not CPU Too? Sooner or Later?
2: NO, I didn't. Now I've done it on project page and I'v done "update" on the host 25 min ago. Nothing, untill now. Have I to reset again or have I only to wait?
3: What do you mean with (mine) requirements?

I have found this:

Make a new text file named app_config.xml and have this in the contents

<app_config>
<app>
<name>setiathome_v7</name>
<gpu_versions>
<gpu_usage>0.5</gpu_usage>
<cpu_usage>.04</cpu_usage>
</gpu_versions>
</app>
</app_config>


and this

<app_config>
<app> <name>setiathome_enhanced</name>
<max_concurrent>6</max_concurrent>
<gpu_versions>
<gpu_usage>0.5</gpu_usage>
<cpu_usage>0.04</cpu_usage>
</gpu_versions>
</app>
</app_config>

... ad many others ....
Where can I find instructions to build different files?

As I understand I will have V6 and V7, both GPU and CPU.
On some Hosts I run only one GPU task, on three ones I run two.

Thanks in advance for any clear answer
ID: 1375277 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1375285 - Posted: 2 Jun 2013, 11:09:04 UTC - in response to Message 1375277.  
Last modified: 2 Jun 2013, 11:13:02 UTC

I repeat here part of a discussion started in the news forum

1) All the Stock v7 GPU apps are Lunatics apps now!
2) You did go and enable 'SETI@home v7: yes' in your project preferences didn't you?
3) No, you copy and paste it from one of the posts in Number Crunching, and edit it to your requirements.



1: great! But Only GPU? Why not CPU Too? Sooner or Later?
2: NO, I didn't. Now I've done it on project page and I'v done "update" on the host 25 min ago. Nothing, untill now. Have I to reset again or have I only to wait?
3: What do you mean with (mine) requirements?

1) No, Lunatics CPU apps won't become Stock apps (not at present anyway), it'll be too complicated project side making sure each app only goes to the right host, better to have a single app as Stock that covers all bases,
Lunatics has supplied improvements to the Stock v7 CPU app, it now does more function tests than 6.03 and choices the functions that are fastest, so it is faster than the old 6.03 app, especially on AVX hosts,
it also uses a newer fftw dll giving a further ~10% speedup.
2) Most of your hosts are running anonymous platform with Setienhanced v6 apps in them, you can't get v7 work until until either you update your app_info's with v7 apps, or you revert to Stock.
3) Depends on how many total tasks you want to run at once, and how many total GPU tasks you want to run at once, so your requirements, Edit the app_config.xml as you see fit,
But it's pointless setting a v7 app_config entry if you can't do v7 work, see answer 2.

Where can I find instructions to build different files?

What do you mean by that? are you after the source code and instructions for building applications or Boinc?, or how to do a app_config.xml?

Claggy
ID: 1375285 · Report as offensive
Profile Francesco Forti
Avatar

Send message
Joined: 24 May 00
Posts: 334
Credit: 204,421,005
RAC: 15
Switzerland
Message 1375323 - Posted: 2 Jun 2013, 12:44:03 UTC - in response to Message 1375285.  
Last modified: 2 Jun 2013, 12:44:51 UTC

answer to claggy 1375285

1) clear
2) I have renamed app_inf in two hosts (one yesterday [6334665] and the other hours ago [5381180]) reverting to stock but no V7 at all. Only V6 of different types. Of course I have restard Boinc and reset project. Preferences are "SETI@home v7: yes". I usually walk step by step. If something works I extend to all hosts. I' still waitng to know if I have only to wait or if I have some mistake.
3)for CPU max can be all CPUS, for GPUS it depends from GPU type. I known which GPU is able to run 1 or 2 task, with the previous lunatics apps. I think it should be the same with new one, correct? Does exist some page showing us how to write a good app_config.xml?
ID: 1375323 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1375341 - Posted: 2 Jun 2013, 13:12:37 UTC - in response to Message 1375323.  

answer to claggy 1375285

1) clear
2) I have renamed app_inf in two hosts (one yesterday [6334665] and the other hours ago [5381180]) reverting to stock but no V7 at all. Only V6 of different types. Of course I have restard Boinc and reset project. Preferences are "SETI@home v7: yes". I usually walk step by step. If something works I extend to all hosts. I' still waitng to know if I have only to wait or if I have some mistake.
3)for CPU max can be all CPUS, for GPUS it depends from GPU type. I known which GPU is able to run 1 or 2 task, with the previous lunatics apps. I think it should be the same with new one, correct? Does exist some page showing us how to write a good app_config.xml?

2) Host 6334665 has had plenty of v7 work. Host 5381180 has already got a whole shitload of work allocated to it, including 75 Astropulse v6 tasks, you'll have to get those resent, and crunched before you'll start getting Seti v7 tasks.

3) It is not CPU max, it is max_concurrent, that is GPU & CPU, if you put that as 4, and put gpu_usage as 0.5 you'll run two on the GPU and two on the CPU (a total of four), As with all Boinc configuration, it is in the Wiki: Client configuration

Claggy
ID: 1375341 · 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 1375373 - Posted: 2 Jun 2013, 14:13:51 UTC - in response to Message 1375109.  

It appears that the maximum run time setting of V7 units may be set a bit tight.

For install and testing purposes. I started V7 crunching running only one unit per GPU. When I was happy all was working properly I switched back to 2 per GPU.

As a result of this I trashed 60 odd units with -177 errors, all timing out after ~250 seconds. There has been no errors with units downloaded since that batch.

I thought the timeout was 10x the expected run time but these failures were only 2x ?

T.A.

The timeout is 10x the server estimate of runtime, the splitters still set that 10x factor. Maybe the reason it looks like 2x on your systems is because you're running BOINC 6.10.x so duration_correction_factor (DCF) affects your local run time estimates. For those running BOINC 7.x the local estimate matches the server estimate because DCF is disabled.
                                                                  Joe
ID: 1375373 · Report as offensive
JohnDK Crowdfunding Project Donor*Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 28 May 00
Posts: 1222
Credit: 451,243,443
RAC: 1,127
Denmark
Message 1375613 - Posted: 2 Jun 2013, 21:40:16 UTC - in response to Message 1374379.  

One WU finished waiting for validation.

http://setiathome.berkeley.edu/result.php?resultid=3023153356

Will keep an eye one things.


Much appreciated here :) Dr Ivan mentioned, buried amidst the chaos, that he's going to have a look for x41zc update needs, and Urs provided some helpful starting suggextions.

Seems 41g works OK, so far no errors and only a few inconclusive.

http://setiathome.berkeley.edu/results.php?hostid=6636133&offset=0&show_names=0&state=0&appid=11

Well hope you and Ivan get the x41zc update working :)
ID: 1375613 · Report as offensive
Profile Francesco Forti
Avatar

Send message
Joined: 24 May 00
Posts: 334
Credit: 204,421,005
RAC: 15
Switzerland
Message 1375789 - Posted: 3 Jun 2013, 6:46:14 UTC - in response to Message 1375341.  


2) Host 6334665 has had plenty of v7 work. Host 5381180 has already got a whole shitload of work allocated to it, including 75 Astropulse v6 tasks, you'll have to get those resent, and crunched before you'll start getting Seti v7 tasks.

3) It is not CPU max, it is max_concurrent, that is GPU & CPU, if you put that as 4, and put gpu_usage as 0.5 you'll run two on the GPU and two on the CPU (a total of four), As with all Boinc configuration, it is in the Wiki: Client configuration

Claggy

2) Ok, Now (this morning, at my time) I see them too.
3) Ok, everything is clear, now. I will read that page.
Thanks
ID: 1375789 · Report as offensive
Darth Beaver Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 20 Aug 99
Posts: 6728
Credit: 21,443,075
RAC: 3
Australia
Message 1375796 - Posted: 3 Jun 2013, 7:03:31 UTC - in response to Message 1375341.  
Last modified: 3 Jun 2013, 7:05:26 UTC

answer to claggy 1375285

1) clear
2) I have renamed app_inf in two hosts (one yesterday [6334665] and the other hours ago [5381180]) reverting to stock but no V7 at all. Only V6 of different types. Of course I have restard Boinc and reset project. Preferences are "SETI@home v7: yes". I usually walk step by step. If something works I extend to all hosts. I' still waitng to know if I have only to wait or if I have some mistake.
3)for CPU max can be all CPUS, for GPUS it depends from GPU type. I known which GPU is able to run 1 or 2 task, with the previous lunatics apps. I think it should be the same with new one, correct? Does exist some page showing us how to write a good app_config.xml?

2) Host 6334665 has had plenty of v7 work. Host 5381180 has already got a whole shitload of work allocated to it, including 75 Astropulse v6 tasks, you'll have to get those resent, and crunched before you'll start getting Seti v7 tasks.

3) It is not CPU max, it is max_concurrent, that is GPU & CPU, if you put that as 4, and put gpu_usage as 0.5 you'll run two on the GPU and two on the CPU (a total of four), As with all Boinc configuration, it is in the Wiki: Client configuration

Claggy



Claggy are you saying I can get even the cpu to double up on how many it does with Lunatics ?
Or you talking stock v7

Currently my Quaddy does max of 4 ..so I can get it to do 8 ?
if so now I get why my Rac doesn't show upwards of 9000 I've been expecting 20000 or close to that
ID: 1375796 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1375808 - Posted: 3 Jun 2013, 7:40:04 UTC - in response to Message 1375796.  

Claggy are you saying I can get even the cpu to double up on how many it does with Lunatics ?

In the past running more than 1 WU per CPU core has always resulted in a signifcant reduction in processing done.
With a GPU running 2 or 3 at a time just takes the GPU utilisation up to 100%, which is what a CPU core runs at when processing a WU.
Grant
Darwin NT
ID: 1375808 · Report as offensive
Darth Beaver Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Avatar

Send message
Joined: 20 Aug 99
Posts: 6728
Credit: 21,443,075
RAC: 3
Australia
Message 1375820 - Posted: 3 Jun 2013, 7:50:31 UTC - in response to Message 1375808.  

Claggy are you saying I can get even the cpu to double up on how many it does with Lunatics ?

In the past running more than 1 WU per CPU core has always resulted in a signifcant reduction in processing done.
With a GPU running 2 or 3 at a time just takes the GPU utilisation up to 100%, which is what a CPU core runs at when processing a WU.


Umm so no then ? I already have the GPU doing 3

I also checked my app_info file and it does not have

<app_config>
<app>
<name>uppercase</name>
<max_concurrent>1</max_concurrent>
<gpu_versions>
<gpu_usage>.5</gpu_usage>
<cpu_usage>.4</cpu_usage>
</gpu_versions>
</app>
</app_config>

the example from Bionic page you put up Claggy

I could not figure out thou if this is stock
seeing as it's not in my xml file I just leave it till I read up some more thanks thou Grant
ID: 1375820 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1375842 - Posted: 3 Jun 2013, 8:52:54 UTC - in response to Message 1375820.  

Claggy are you saying I can get even the cpu to double up on how many it does with Lunatics ?

In the past running more than 1 WU per CPU core has always resulted in a signifcant reduction in processing done.
With a GPU running 2 or 3 at a time just takes the GPU utilisation up to 100%, which is what a CPU core runs at when processing a WU.


Umm so no then ? I already have the GPU doing 3

I also checked my app_info file and it does not have

the example from Bionic page you put up Claggy

I could not figure out thou if this is stock
seeing as it's not in my xml file I just leave it till I read up some more thanks thou Grant

Not app_info.xml, But app_config.xml, different files doing different things, app_info.xml specifies what apps Boinc should run via anonymous platform, app_config.xml puts limits on how many from a app name that can run at once,
enables other projects to run an instance or two, again you can specify how many of their apps names you want to run at once, So rather than Boinc 7 flipping between projects as the priorities change, it runs work from each constantly.

Claggy
ID: 1375842 · Report as offensive
Profile Francesco Forti
Avatar

Send message
Joined: 24 May 00
Posts: 334
Credit: 204,421,005
RAC: 15
Switzerland
Message 1376402 - Posted: 4 Jun 2013, 6:38:55 UTC

Hi, as suggested I have prepared an app_config.hml file for hosts in which GPU I want to run 2 intances.

Here is my file
<app_config>
    <app>
    <name>setiathome_v7</name>
    <gpu_versions>
       <gpu_usage>0.5</gpu_usage>
       <cpu_usage>.4</cpu_usage>
    </gpu_versions>
</app>
<app>
    <name>astropulse_v6</name>
    <gpu_versions>
       <gpu_usage>0.5</gpu_usage>
       <cpu_usage>0.4</cpu_usage>
    </gpu_versions>
    </app>
</app_config>

As I can see, it' working.
In one hosts now I see two
SETI@home 6.04 AstroPulse v6 (opencl_nvidia_100) running
and in the other I see two
SETI@home 7.00 SETI@home v7 (cuda50) Running

Same file in both host.
But in both host I'v got the message : app setiathome_v7 not found in app_config.xml
ID: 1376402 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6325
Credit: 106,370,077
RAC: 121
Russia
Message 1376434 - Posted: 4 Jun 2013, 8:21:14 UTC - in response to Message 1374156.  

Now that my GTX460 cruncher is on stock app for the first time, it downloaded its first Nvidea AP to crunch using opencl. However, GPU utilization is very low. For MB WUs, GPU utilization is typically 80-90% (for 1 v7 WU), but the AP was only showing ca 4% GPU utilization. I freed up one CPU core and this increased to 18% with occaisional bursts up to 100%. Freeing up a second CPU core made no difference. Is this normal?

The AP WU is currently 24% complete and has been running for 1.5 hours, so it is heading for about 6 hours to complete. Does this sound right for a GTX460 running at 857MHz?

Check blanking % in results file. Worth to discuss only 0% blanking tasks. Blanking on CPU and it's known limitation.
Also, for FERMI and higher you could try to run 2 or even more AP tasks at once. It could be good even for pre-FERMI GPUs if AP tasks from noisy (high blanking) tape.

SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1376434 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6325
Credit: 106,370,077
RAC: 121
Russia
Message 1376436 - Posted: 4 Jun 2013, 8:26:17 UTC - in response to Message 1374669.  

Maybe this is the correct thread to ask this question. The auto correlation increases the sensitivity for weak signals by how many db?

None. Autocorrelation is a search for a different type of signal, where the transmission is the combination of an original signal combined with a delayed copy of the same signal. Such a combination would be a good indication of intelligent intent given a significant delay.
                                                                   Joe

Or gravitational lensing.

I had the same thought. There's also another phenomenon I've seem mention of in visible light, where secondary light is scattered off surrounding debris (from a supernova, e.g) -- but that delay is measured in years*10^N so we're unlikely to catch it... :-O!

Anyway even being non-intellegent such kind of signals interesting enough to be investigated.

SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1376436 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6325
Credit: 106,370,077
RAC: 121
Russia
Message 1376438 - Posted: 4 Jun 2013, 8:35:06 UTC - in response to Message 1374620.  

I just don´t understand why not put a instruction to allow us to choose the best app to our host manualy, that could save a lot of time and resources.

Because those that know will use the Lunatics installer.
The stock application is for those that don't know.

Actually being on stock apps has own advantages so if something can be implemented as stock feature better to implement it as such. And leave anonymous platform for "hard cases" :)

SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1376438 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1376540 - Posted: 4 Jun 2013, 16:49:02 UTC - in response to Message 1374620.  

I just don´t understand why not put a instruction to allow us to choose the best app to our host manualy, that could save a lot of time and resources.

Because those that know will use the Lunatics installer.
The stock application is for those that don't know.

Maybe it's just me, but that comment comes across as arrogant and condescending. There are quite a few of us here that have the level of knowledge and skill to run Optimized Apps, but choose NOT to. I could run the Lunatics package on my P4HT/WinXP boxes, but I don't have a need or desire for the extra production that the Opt Apps provide - my RAC is sufficient, and my results are as valid as any others.

And remember, too, that the Lunatics package only runs on Windows boxes - there is no equivalent for my Macs or other non-Windows OSes.
Donald
Infernal Optimist / Submariner, retired
ID: 1376540 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1380593 - Posted: 13 Jun 2013, 8:24:50 UTC - in response to Message 1376540.  


Probably as good a place as any-

Is there any one around that had been running the older stock or optimised CPU applications and is now running the new optimised application & making use of the AVX support?

Since upgrading from the old optimised application to the current v7 optimised with AVX supprt, my i7 2600 CPU is now running about 5°c hotter.
Grant
Darwin NT
ID: 1380593 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34253
Credit: 79,922,639
RAC: 80
Germany
Message 1380603 - Posted: 13 Jun 2013, 9:33:20 UTC

I noticed that on my FX also.
Only 2 - 3°C but its there.




With each crime and every kindness we birth our future.
ID: 1380603 · Report as offensive
Previous · 1 . . . 14 · 15 · 16 · 17 · 18 · Next

Message boards : Number crunching : Now that v7 has rolled out..........


 
©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.