I have noticed that there are some users which have worse or same hardware as I, but they have like 1000 tasks in progress where I rarely get 2 at same time...

Is there a way to fix this? I need more tasks.

I have noticed that there are some users which have worse or same hardware as I, but they have like 1000 tasks in progress where I rarely get 2 at same time...

Is there a way to fix this? I need more tasks.

Actually according to the website you have 350 workunits "in progress" right now. But your problem appears to be lots of projects and only one pc, meaning you are spreading your crunching across lots of projects and each gets some of your processing power but none get alot of it.

Hi,

I can see this symptom on my own system runnig BOINC Client v7.0.28 and it makes it switch to a backup GPU project all the time. :( I'm not trying deliberately limit the number of active tasks to ridiculously low number like 1 or 2. Must be another v7 bug but I'll see if a BOINC Server code upgrade helps here. Other thing to check are the BOINC Client preferences.

-w

i was quite amazed; after making this topic (30min+-) i got a lot of tasks.

*(my settings are 300 resource share) and stock work for 10days and additional 10days. I remember in past I did cancel a whole bunch of tasks when i was given over 2k and I wanted to do some other projects. After that i didn't get more than 3 tasks

Thanks for whoever what did.
---------------------------------------------------------------------

@mikey
That is incorrect; my currently gpu active project is only moo.

i was quite amazed; after making this topic (30min+-) i got a lot of tasks.

*(my settings are 300 resource share) and stock work for 10days and additional 10days. I remember in past I did cancel a whole bunch of tasks when i was given over 2k and I wanted to do some other projects. After that i didn't get more than 3 tasks

Thanks for whoever what did.

That could be the answer then...when you cancel units, or return them late, the Boinc software marks your pc as 'unreliable' and starts limiting the units you can get until your pc proves it is reliable again. It does the proving by returning unit on time and valid and for every 1 you return that is good you can get 2 in return. So the return to normality is fairly quick. You could see this thru the daily quota of X units per day as your limit.

---------------------------------------------------------------------

@mikey
That is incorrect; my currently gpu active project is only moo.

Sorry on your list of top projects these all take gpu's:
PrimeGrid, MilkyWay@home and SETI@home. I just figured you were crunching optimally for them too.

Hi,

I think I fixed this just now and according to logs the scheduler is now pushing more work out for you all. So much that we just run out of huge work units. :) But don't despair, our work generator is hard at work creating moar.

-w