I just wondering why this project is deleting finished tasks so fast. As a volunteer cruncher I want to see at least my work for some time to check the "quality" of my work. Thus is denied to me. I even do not see if credits are granted or not for certain tasks (e.g. when set on inconclusive) Why is that so?
Storage Memory can't be the reason, since the cost for it is really low.

It was an easy fix to keep server load down at times in the past couple years and was left this way. I don't think it is needed as much now but it is still in place.

---

Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

It was an easy fix to keep server load down at times in the past couple years and was left this way. I don't think it is needed as much now but it is still in place.

It would REALLY be nice if they stuck around long enough to see! At least a day or two.

---

This was from when I was debugging the server awhile ago. I've increased the purge to be 12 hours (from 0). Kind of need to move in small steps because we don't want to bring down the database due to the high volume of workunits being processed.

If this works over the next couple weeks just make another post reminding me (or send an email) and I'll bump it up to a day and see if that works out.

--Travis

---

Good thing! :-)

But I think 12 hours are enough to detect faults etc.

With a fast GPU you get tons of WU's in your lists.
A much better improvement will be if we try to increase the WU limit per GPU in steps from 12 to 24 and finaly to 48.
That would bridge small server outages.

regards

Franz

Yes 12 Hours is waaay enough.

I agree with the wanted raise from 12 to 24 GPU WU Limit. The lowest card that can run MW is my ATI 3850 that works this 12 Tasks out in round 1,5-2 hours (depends on the WU Unit Type from 7-15 Mins/WU). Thats a bit hard when server fail and the Computer is running a longer time without a control and eats Energy without computing anything ^^ Normaly i turn it of cos there are no other interesting ATI projects for me ^^

---

DSKAG Austria Research Team: http://www.research.dskag.at

Good thing! :-)

But I think 12 hours are enough to detect faults etc.

With a fast GPU you get tons of WU's in your lists.
A much better improvement will be if we try to increase the WU limit per GPU in steps from 12 to 24 and finaly to 48.
That would bridge small server outages. regards Franz

+2

I would like to see an increase in the limit as well.

For me is one or two hours more then enough.

---

Greetings from,
TJ

I support the increase of GPU WUs limit too. My dual HD5870 machine runs out of WUs pretty fast.

---