Splitters

Message boards : Number crunching : Splitters
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 864
Credit: 138,079,224
RAC: 87,270
United States
Message 1929935 - Posted: 14 Apr 2018, 16:51:07 UTC

Why are the splitters not running?

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1929935 · Report as offensive
rob smith Special Project $250 donor
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 16448
Credit: 338,601,400
RAC: 213,026
United Kingdom
Message 1929937 - Posted: 14 Apr 2018, 17:17:29 UTC

As I type only the AP splitters are off line, which is their normal state when there is nothing for them to split.
Also, they run in short bursts, not continuously, so if the stats sample is taken in one of their "rests" they will appear to be off-line.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1929937 · Report as offensive
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 864
Credit: 138,079,224
RAC: 87,270
United States
Message 1929943 - Posted: 14 Apr 2018, 18:05:32 UTC - in response to Message 1929937.  

They were all Disabled. They were that was for at least 30 minutes until I moved on to something else. Well they are back now and running just fine it would seem.

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1929943 · Report as offensive
Unixchick

Send message
Joined: 5 Mar 12
Posts: 122
Credit: 1,628,493
RAC: 866
United States
Message 1929978 - Posted: 14 Apr 2018, 21:02:24 UTC

I think that the splitters stop when there is "enough" ready to send work units.
I don't know what the magic number is where it stops the splitters or when it restarts them (guessing when ready to send falls below some other magic number).
ID: 1929978 · Report as offensive
juan BFP Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 7118
Credit: 410,001,319
RAC: 272,198
Panama
Message 1929992 - Posted: 14 Apr 2018, 22:09:15 UTC - in response to Message 1929978.  
Last modified: 14 Apr 2018, 22:16:26 UTC

I think that the splitters stop when there is "enough" ready to send work units.
I don't know what the magic number is where it stops the splitters or when it restarts them (guessing when ready to send falls below some other magic number).

Yes the magic number is in the rangue of 620K. give or take few 10K WU. Maybe somebody could post for us.
ID: 1929992 · Report as offensive
Profile Bill G Special Project $75 donor
Avatar

Send message
Joined: 1 Jun 01
Posts: 864
Credit: 138,079,224
RAC: 87,270
United States
Message 1930003 - Posted: 14 Apr 2018, 22:43:31 UTC - in response to Message 1929992.  

I just have never seen it before, and I have looked a lot.
I think that explanation certainly makes sense.

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours
ID: 1930003 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 10313
Credit: 138,361,826
RAC: 80,978
Australia
Message 1930005 - Posted: 14 Apr 2018, 22:48:02 UTC - in response to Message 1929992.  

Yes the magic number is in the rangue of 620K. give or take few 10K WU. Maybe somebody could post for us.

Pretty sure the Ready-to-send target is around 600k; they generally shut down by 650k & restart at around 550k.
It's only now with the reduced server load of NVidia GPUs doing Arecibo VLARs that the Ready-to-send buffer has actually been able to refill for long enough for the splitters to shut down for a while. For quite a while now the load had been such that if the splitters even slowed down, the Ready-to-send buffer would quickly start draining and get below the refill trigger point before the splitters had a chance to actually stop.
Grant
Darwin NT
ID: 1930005 · Report as offensive
Unixchick

Send message
Joined: 5 Mar 12
Posts: 122
Credit: 1,628,493
RAC: 866
United States
Message 1930062 - Posted: 15 Apr 2018, 4:42:47 UTC - in response to Message 1930005.  

looks like we are getting near the 650K magic number in the ready to send queue. It will be interesting to watch the stop and start of the spliters.
thanks for the "magic" numbers as it makes less worrisome to know that the program has a way to handle this event.
ID: 1930062 · Report as offensive

Message boards : Number crunching : Splitters


 
©2018 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.