terryry

Terry E Ryterski
Terry E Ryterski
Joined: 3 Apr 05
Posts: 3
Credit: 55641
RAC: 0
Topic 190530

my computer works on 2 units at a time einstine is the only one that only sends one work unit at a time why

Stick
Stick
Joined: 24 Feb 05
Posts: 790
Credit: 33738776
RAC: 20838

terryry

Quote:
my computer works on 2 units at a time einstine is the only one that only sends one work unit at a time why

Your description is a little too vague and your computers are hidden - so it is difficult to answer your question. I would suggest that you read through the Work Scheduler of the BOINC Wiki. It is a very good explanation of how BOINC requests and schedules work.

Tern
Tern
Joined: 27 Jul 05
Posts: 309
Credit: 99440614
RAC: 0

As Stick said, unless you

As Stick said, unless you "show" your computers, we really can't give you much help. (Look at mine to see what would be visible, it's nothing that would put you "at risk".)

Just taking a wild guess - you have a Hyperthreading P4 CPU and are running SETI and Einstein, your Work tab normally contains either one or two SETIs and one Einstein. Therefore your CPU is "running" either one SETI and one Einstein, or two SETIs, at any given moment. This is because you probably have your cache size set to the "default" 0.1 day (2.4 hours) and an Einstein result takes more than 2.4 hours on your machine, while a SETI result takes less than 2.4 hours.

IF you are giving 50% to SETI and 50% to Einstein, the "best case" for performance on your machine is to always run one SETI and one Einstein in parallel. To do that, you have to _lower_ your cache size so that only one SETI will "fit". You can set it to 0.01 and it will do this. The problem then becomes, if SETI isn't sending work, for even an hour or so, you will be running Einstein-only, and _possibly_ even have an idle CPU for a bit. Because SETI is very, um, unreliable on work availability, most people _raise_ their cache size to at least 0.25 days, which will give them enough work to "ride out" at least the every-Wednesday outage. With a fast P4 and only two projects, _I_ would set the cache to 0.5 days. It will sometimes be running two SETIs or two Einsteins, but the performance hit from doing this instead of "one of each" is minimal.

Let me know if I'm anywhere close on my guess - I'm trying to work on my mind-reading skills. :-)

Terry E Ryterski
Terry E Ryterski
Joined: 3 Apr 05
Posts: 3
Credit: 55641
RAC: 0

I have alowed computers to be

I have alowed computers to be shown now.work on 3 projects Einstein Predictor & Setti all set at 100%. Changed cache to.5 Predictor downloaded 12 work units Einstein still only one

Terry E Ryterski
Terry E Ryterski
Joined: 3 Apr 05
Posts: 3
Credit: 55641
RAC: 0

Scratch last message that did

Scratch last message that did solve the problem all have multiple work units

Thanks

Terry

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.