Message from server: This computer has reached a limit on tasks in progress

[PatrickV2]

Hi there,

I recently noticed the cache on my main crunching rig (Q6600 + 8800GTX) was getting smaller. I have no exact number, but it's only about 4 screenfulls in BOINC Manager.

EDIT: When browsing the link below, there seem to be 139 tasks in progress.

I also see the following message every time the 'puter tries to request more work:

24-Sep-11 11:57:40   SETI@home   Message from server: This computer has reached a limit on tasks in progress

Link to host:

http://setiathome.berkeley.edu/show_host_detail.php?hostid=5188158

Yes, I am aware I have set the cache probably a bit (too?) high (5 days + 5 days), but I did this after I ran out of WUs several times during the heavy outages a while ago. Is there anything I can do? Should I lower these values and request less work? What would be a more safe value?

Regards, Patrick.

[Mike]

Noticed that. too.

So far i know its limited to 50 WUs per core a day for now.

[Richard Haselgrove]

This is a server issue at Berkeley, nothing to do with local or user account settings.

We've been told that the current (temporary) limits are 50 CPU tasks, and 400 GPU tasks - in total, not per core/GPU this time. It's only a quick'n'dirty safety net while other problems are sorted out, and it may not be working quite as expected - but since the complete end-to-end server system can't process any more WUs than it is at present (because of the VHAR recordings from Arecibo this summer), there's not much point in tweaking it.

Sit tight, and hold on for a rough ride. Better to look for more details on the main SETI message boards.

[PatrickV2]

Thanks for the info.

Just got home again, and that same machine has just downloaded a bunch of Einstein@Home work (I set it's resource share to 0.0%).

It seems that works as intended: get work from Einstein@Home if it looks like Seti work dries up.

I sure hope they get their act together soon; until then it looks like I'll be donating my cycles to Einstein@Home.

Regards, Patrick.

[PatrickV2]

Looking just at the cache (and not having examined the forums at Seti in detail) it looks like the above-mentioned limit has been lifted/extended?

My main cruncher is still working away some Einstein@Home tasks, but after that I see a 'sea' of Seti tasks being queued up...

What are the 'new' rules?

Regards, Patrick.

[Richard Haselgrove]

What are the 'new' rules?

I'm not aware of any change from 50 CPU, 400 GPU - but I'm nowhere near those levels myself, so it will take time to verify.

[arkayn]

I am still hovering around 400 GPU units on my GTX560.

[Josef W. Segur]

As of 2011-10-01  03:27:29 UTC the CPU limit for my quad core Q8200 was still just 50. I'm only running a half day cache in the hope that SETI Beta will be reactivated, and got some non-shorty tasks since then so don't know if the limit has been revised.
                                                        Joe

[Mike]

Got this message a couple times as well.
But didnīt get the amount of units to reach the limits.
But i confirmed a couple times that server code isnīt to specific. LOL

[PatrickV2]

Sorry for the false 'alarm' then. The main cruncher does seem to have a mix of Seti CPU and Einstein@Home CPU work in the queue.

However, no GPU Einstein@Home work is queued, probably because I have a boatload of seti-GPU units lined up.

We'll see how it goes...

Regards, Patrick.