SETI@Home CPU cores

Message boards : Number crunching : SETI@Home CPU cores
Message board moderation

To post messages, you must log in.

AuthorMessage
Demented-Idiot

Send message
Joined: 30 Aug 16
Posts: 6
Credit: 3,544,155
RAC: 0
United Kingdom
Message 1834455 - Posted: 6 Dec 2016, 11:55:46 UTC

Hi Guys

I run a Seti VM on my hypervisors. Boinc is set to use all available CPU and I limit seti by virtue of the VM only having a certain number of cores.

I've recently upgraded one of my servers and I've now allocated 6 CPU cores to the Seti VM but it seems to consistently only run at about 80% of my CPU. I can see there're 4 setiathome clients running under boinc so I presume its to do with threading and the fact that more threads are available than processes running?

Is there a way i can make boinc/seti spawn more processes to make use of all available CPU? Will it see the change automatically and start ramping up work thus i just need to be patient?

I'm running Linux setiathome_8.00

Thanks
ID: 1834455 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13736
Credit: 208,696,464
RAC: 304
Australia
Message 1834543 - Posted: 7 Dec 2016, 5:00:42 UTC - in response to Message 1834455.  

I'd suggest exiting & restarting BOINC and see if Seti picks up the extra available cores. If not, check in the event log to see if the manager is recognising all of the available cores.
Grant
Darwin NT
ID: 1834543 · Report as offensive
Demented-Idiot

Send message
Joined: 30 Aug 16
Posts: 6
Credit: 3,544,155
RAC: 0
United Kingdom
Message 1835028 - Posted: 9 Dec 2016, 19:07:23 UTC

Hi

It sees the correct number of cores, but it only uses about 80% CPU - I've made sure the preferences are set for 100% so i assume its just the individual processes that spawn under boinc arent maxing out the cores
ID: 1835028 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1835041 - Posted: 9 Dec 2016, 20:45:40 UTC - in response to Message 1835028.  

Hi

It sees the correct number of cores, but it only uses about 80% CPU - I've made sure the preferences are set for 100% so i assume its just the individual processes that spawn under boinc arent maxing out the cores

If you are using local preferences you might want to give clearing them a try. You can also delete global_prefs_override.xml before starting BOINC.
Sometimes BOINC gets a little off kilter and that helps it out.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1835041 · Report as offensive
Demented-Idiot

Send message
Joined: 30 Aug 16
Posts: 6
Credit: 3,544,155
RAC: 0
United Kingdom
Message 1835107 - Posted: 10 Dec 2016, 1:56:03 UTC - in response to Message 1835041.  

Hi

It sees the correct number of cores, but it only uses about 80% CPU - I've made sure the preferences are set for 100% so i assume its just the individual processes that spawn under boinc arent maxing out the cores

If you are using local preferences you might want to give clearing them a try. You can also delete global_prefs_override.xml before starting BOINC.
Sometimes BOINC gets a little off kilter and that helps it out.



# cat global_prefs_override.xml 
<!--
This configuration file global_prefs_override.xml for the BOINC core client
can be used to override global preferences locally. For a complete list of
all preferences which can be overridden see:
http://boinc.berkeley.edu/trac/wiki/PrefsOverride
-->
<global_preferences>
</global_preferences>


:)
ID: 1835107 · Report as offensive
Demented-Idiot

Send message
Joined: 30 Aug 16
Posts: 6
Credit: 3,544,155
RAC: 0
United Kingdom
Message 1835923 - Posted: 14 Dec 2016, 18:40:58 UTC - in response to Message 1835107.  

Probably should have been clearer with my last reply

That is my prefs file, so no local prefs - but it still only uses between 60 and 80% CPU
ID: 1835923 · Report as offensive
Juha
Volunteer tester

Send message
Joined: 7 Mar 04
Posts: 388
Credit: 1,857,738
RAC: 0
Finland
Message 1835930 - Posted: 14 Dec 2016, 20:11:18 UTC - in response to Message 1835923.  

Are you giving BOINC enough RAM (in both VM and BOINC prefs)?

If you have multiple venues make sure you edit the right one.
ID: 1835930 · Report as offensive
Demented-Idiot

Send message
Joined: 30 Aug 16
Posts: 6
Credit: 3,544,155
RAC: 0
United Kingdom
Message 1835972 - Posted: 14 Dec 2016, 23:21:00 UTC - in response to Message 1835930.  

Hmm, it wasnt struggling for ram as far as I could tell, but never the less i've increased RAM by a few GB for the VM and it seems to be spawning more processes (thus using more CPU now)

cheers :)
ID: 1835972 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1835978 - Posted: 15 Dec 2016, 0:07:52 UTC - in response to Message 1835972.  

Hmm, it wasnt struggling for ram as far as I could tell, but never the less i've increased RAM by a few GB for the VM and it seems to be spawning more processes (thus using more CPU now)

cheers :)

They were actually talking about how the limits in your BOINC Computing preferences
You can change the BOINC limits instead of increasing the RAM for the VM.

My settings may be overkill for most people.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1835978 · Report as offensive

Message boards : Number crunching : SETI@Home CPU cores


 
©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.