Resource sharing between SAH and PAH

Message boards : Number crunching : Resource sharing between SAH and PAH
Message board moderation

To post messages, you must log in.

AuthorMessage
eberndl
Avatar

Send message
Joined: 12 Oct 01
Posts: 539
Credit: 619,111
RAC: 3
Canada
Message 5825 - Posted: 9 Jul 2004, 22:18:00 UTC

So, I'm running both the above projects. I currently have SAH at 100 and PAH at 10, so under normal circumstances, it should be ~91%SAH and ~9% PAH. But sometimes, after PAH reports in to its server, they go back to 50:50.

Does anyone know why this might be?? I don't know if it'll effect the actual number of WU's d/led by the various projects, but it.... just bothers me.

I know, I know, not a big problem, but when SAH thinks it takes 53h to run a WU (I haven't been able to get SAH WUs in more than a week)... I don't want more going to PAH than expected.
ID: 5825 · Report as offensive
Profile D. Drake

Send message
Joined: 21 Aug 01
Posts: 23
Credit: 1,866
RAC: 0
United States
Message 5857 - Posted: 9 Jul 2004, 23:38:16 UTC - in response to Message 5825.  

> So, I'm running both the above projects. I currently have SAH at 100 and PAH
> at 10, so under normal circumstances, it should be ~91%SAH and ~9% PAH. But
> sometimes, after PAH reports in to its server, they go back to 50:50.
>
> Does anyone know why this might be?? I don't know if it'll effect the actual
> number of WU's d/led by the various projects, but it.... just bothers me.
>
> I know, I know, not a big problem, but when SAH thinks it takes 53h to run a
> WU (I haven't been able to get SAH WUs in more than a week)... I don't want
> more going to PAH than expected.
>

The settings from one site should propogate to the other site, but there is a delay in this happening. What I did to fix this was to quickly change my general settings on both projects so that they were the same and when they next contacted the schedulers, they were set at 50/50 like I wanted.

SAH thinks it takes 22 hours to run a WU for me, too. It actually takes just over 4 hours. PAH also gives a high expected run time for me. It thinks about 3 hours and takes right around an hour. I do know that the Angle Rate (slew I guess?) of the WU for SAH makes a big difference in how long it takes to complete. I suspect, but am not sure, that is why sometimes SAH Classic would abort a WU after about 10% processing (too noisy from the speed the telescope was tracking across the sky).

You might try changing your cache to like 0.5 - 1.2 days. But since BOINC thinks it takes you 2.5 days to complete a WU, I dunno if that will help. I had mine set at 0.1 - 0.5 and wasn't getting SETI WUs. Changed to the above cache and now have 4 SETI WUs and no PAH. (shrug)
ID: 5857 · Report as offensive
Profile Keck_Komputers
Volunteer tester
Avatar

Send message
Joined: 4 Jul 99
Posts: 1575
Credit: 4,152,111
RAC: 1
United States
Message 5878 - Posted: 10 Jul 2004, 0:53:05 UTC - in response to Message 5825.  

> So, I'm running both the above projects. I currently have SAH at 100 and PAH
> at 10, so under normal circumstances, it should be ~91%SAH and ~9% PAH. But
> sometimes, after PAH reports in to its server, they go back to 50:50.
>
> Does anyone know why this might be?? I don't know if it'll effect the actual
> number of WU's d/led by the various projects, but it.... just bothers me.
>
> I know, I know, not a big problem, but when SAH thinks it takes 53h to run a
> WU (I haven't been able to get SAH WUs in more than a week)... I don't want
> more going to PAH than expected.
>
Another possible explanation is that your emails are different on the two projects. They should both be in all lower case, since some of the database software will automatically convert to lower case. This could cause the client to use the preferences of whichever project was contacted last instead of propagating them.

John Keck
BOINCing since 2002/12/08
ID: 5878 · Report as offensive
eberndl
Avatar

Send message
Joined: 12 Oct 01
Posts: 539
Credit: 619,111
RAC: 3
Canada
Message 5891 - Posted: 10 Jul 2004, 1:21:57 UTC - in response to Message 5878.  

> Another possible explanation is that your emails are different on the two
> projects. They should both be in all lower case, since some of the database
> software will automatically convert to lower case. This could cause the client
> to use the preferences of whichever project was contacted last instead of
> propagating them.
>
John Keck
> BOINCing since 2002/12/08

no, my email is the same for both, written the same way and everything... Perhaps it was just that while SAH was down PAH was assuming that they had equal weight... I dunno, everything seems to be back to normal, but I won't know until I get this "massive" WU out of the way.
ID: 5891 · Report as offensive
Profile Keck_Komputers
Volunteer tester
Avatar

Send message
Joined: 4 Jul 99
Posts: 1575
Credit: 4,152,111
RAC: 1
United States
Message 5929 - Posted: 10 Jul 2004, 3:19:09 UTC - in response to Message 5891.  

> > Another possible explanation is that your emails are different on the
> two
> > projects. They should both be in all lower case, since some of the
> database
> > software will automatically convert to lower case. This could cause the
> client
> > to use the preferences of whichever project was contacted last instead
> of
> > propagating them.
> >
John Keck
> > BOINCing since 2002/12/08
>
> no, my email is the same for both, written the same way and everything...
> Perhaps it was just that while SAH was down PAH was assuming that they had
> equal weight... I dunno, everything seems to be back to normal, but I won't
> know until I get this "massive" WU out of the way.
>
Could be. BOINC forgets alot of project specific info when a master fetch fails, resource share could be part of it.

John Keck
BOINCing since 2002/12/08
ID: 5929 · Report as offensive

Message boards : Number crunching : Resource sharing between SAH and PAH


 
©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.