Oops... Long runs = 1
Really ?


____________
[CSF] Thomas H.V. Dupont
Founder of the team CRUNCHERS SANS FRONTIERES 2.0
www.crunchersansfrontieres

Sometimes experiments finish. That's a good thing.
A lot of what's been sent out still has a few steps left.

Plus us 918 users, on the whole, already have a spare work unit.
That should give a bit of leeway so that most people are crunching when there are steps left.

Perhaps though some notice or discussion of this on the News board would be appreciated.

I appreciate that experiments conclude, but also, as I am out of work on some systems already, would like to have an idea of the length of the 'no work to send'

I think it will have to wait until early next week.
The team is in weekend like us !
Happy weekend to all elsewhere :)
____________
[CSF] Thomas H.V. Dupont
Founder of the team CRUNCHERS SANS FRONTIERES 2.0
www.crunchersansfrontieres

Thomas, I expect that will be the case.

But just for the sake of discussion, do you think the team was caught by surprise as to the timing of this?

I suspect they saw this coming at least during the past week, and perhaps before this (as they've launched the AMD GPU work).

So my lament as it were, is that there was no prior alert.

I'm able to shift (or add) other projects to use the GPU's and will do so as the GPUGrid work clears out.

I'm not part of the GPUG crew so I don't know and I don't want to speculate on this subject, Barry.
I guess we will have news very quickly.
Just be a little patient.
BTW, happy weekend to you, Barry :)
____________
[CSF] Thomas H.V. Dupont
Founder of the team CRUNCHERS SANS FRONTIERES 2.0
www.crunchersansfrontieres

Well - in the last 30 minutes I've done a BOINC "update" on my two hosts and got a new WU each time. Perhaps the server status is telling porkies... :)

Well - in the last 30 minutes I've done a BOINC "update" on my two hosts and got a new WU each time. Perhaps the server status is telling porkies... :)

Thanks so much tomba and happy weekend to you :)

Thanks so much tomba and happy weekend to you :)

Et un week-end heureux de vous aussi, Thomas. Salutations de les collines au-dessus de St Tropez.

Dans la langue de Molière ! Merci tomba !

Yes -- I see that new work was released shortly after my whining. The power I have <rueful self-deprecating smile>

Well - in the last 30 minutes I've done a BOINC "update" on my two hosts and got a new WU each time. Perhaps the server status is telling porkies... :)

4 systems sitting idle and I was trying to make the push for 100,000,000 by tomorrow midnight. That ain't happening now.
____________
1 Corinthians 9:16 "For though I preach the gospel, I have nothing to glory of: for necessity is laid upon me; yea, woe is unto me, if I preach not the gospel!"
Ephesians 6:18-20, please ;-)
http://tbc-pa.org

905-NOELIA_PNP-9-10-RND6302_1
Downloaded this morning and my teammates do not fail to work either.
The SSP tells nonsense ?
____________
[CSF] Thomas H.V. Dupont
Founder of the team CRUNCHERS SANS FRONTIERES 2.0
www.crunchersansfrontieres

Out of 7 NVidia GPUs, I have 3 idle. I was able to bog down the other 4 in order to make them last longer (until Monday) so they won't go completely idle. They are checking for tasks every once in a while each and I am getting only tasks that time out on other people's computers or error out on their computers. Oddly enough, because of the timing, I have the one computer with 2 cards that has 4 tasks: 2 working and 2 waiting. Depending on new tasks coming out, I may make the 100,000 on Monday or Tuesday now.
____________
1 Corinthians 9:16 "For though I preach the gospel, I have nothing to glory of: for necessity is laid upon me; yea, woe is unto me, if I preach not the gospel!"
Ephesians 6:18-20, please ;-)
http://tbc-pa.org

I've posted something similar on another thread, but just to reiterate: Einstein at home has GPU work available, despite not making that clear in the BOINC project selection page. So that's my backup project, set at 0 priority. Much better for the GPU work to go somewhere then to just let them idle while we wait for more GPUgrid resends.

hahaha, the others still idle of GPUGrid work, the one with the WUs to spare just got a 3rd spare.
Is there a way to transfer the WUs from one PC to another over the network or are they assigned and stuck?

.....Is there a way to transfer the WUs from one PC to another over the network or are they assigned and stuck?

thx

It looks like they might have just got new WUs together maybe. I just filled up all my GPUs for short and long runs all at once.
____________
1 Corinthians 9:16 "For though I preach the gospel, I have nothing to glory of: for necessity is laid upon me; yea, woe is unto me, if I preach not the gospel!"
Ephesians 6:18-20, please ;-)
http://tbc-pa.org

I think you must have gotten a few resends at once. Certainly the servers are still running low.



What work units did you get? Just tried getting more work, and I got
2/8/2015 4:11:31 PM | GPUGRID | Scheduler request completed: got 0 new tasks
2/8/2015 4:11:31 PM | GPUGRID | No tasks sent
2/8/2015 4:11:31 PM | GPUGRID | No tasks are available for Short runs (2-3 hours on fastest card)
2/8/2015 4:11:31 PM | GPUGRID | No tasks are available for ACEMD beta version
2/8/2015 4:11:31 PM | GPUGRID | No tasks are available for Long runs (8-12 hours on fastest card)
2/8/2015 4:11:31 PM | GPUGRID | No tasks are available for OpenCL for AMD GPUs

Both queues needs to be refilled.
Until then my new-old spare project kicked in.
This is SETI@Home at the moment, but they produce less heat than GPUGrid does, so if I catch cold, I'll put the blame on the staff. :)

This is SETI@Home at the moment, but they produce less heat than GPUGrid does, so if I catch cold, I'll put the blame on the staff. :)

Always happy to feed the beast with some fresh WU :D

Thanks for the new WU's. While my boxes were idle, BarryAZ passed me on RAC. Time to fix that.

Happy Crunching.

Always happy to feed the beast with some fresh WU :D

Thanks for the new WU's.

Short runs are back!
THX!

Thanks for the new WU's. While my boxes were idle, BarryAZ passed me on RAC. Time to fix that.

Happy Crunching.

Attached to BOINC client Einstein@home, with resource sharing of 0. Removed CPU tasks from project preferences, let's see how it works.

Always happy to feed the beast with some fresh WU :D

Sniff, sniff...
2/11/2015 2:25:39 PM | GPUGRID | No tasks are available for Short runs (2-3 hours on fastest card)
2/11/2015 2:25:39 PM | GPUGRID | No tasks are available for Long runs (8-12 hours on fastest card)
____________
[CSF] Thomas H.V. Dupont
Founder of the team CRUNCHERS SANS FRONTIERES 2.0
www.crunchersansfrontieres

I'm preparing a new set of WU. In a couple of hours the equilibration should have finished and they will be launched. Stay tuned!

I got one long Gerard ten minutes ago.

-timon

I got one long Gerard ten minutes ago.

-timon

I'm preparing a new set of WU. In a couple of hours the equilibration should have finished and they will be launched. Stay tuned!

Why exactly is this happening again?

Weekend : Sniff, sniff.

Short runs (2-3 hours on fastest card) 1
Long runs (8-12 hours on fastest card) 59
____________

Guys:

No need to be sad. Be happy!

If we run out of work here, it happens. Just make sure your clients are attached to other GPU projects, to keep the GPUs busy. You can even attach with Resource Share 0, such that they will be backup projects, and will only get work when your main projects are out.

I just sent some more simulations for the weekend. Remember that boinc and our project in particular, as it is set up right now, will respawn approximately 1 WU for each correctly completed WU; therefore you shouldn't look at the unsent WU but at the "in progress" queue to assess the load of GPUGRID. In other words, if you see some WU in the unsent queue it just means that GPUGRID is running at its maximum speed and as soon as you finish the current workunit a new one will be generated in the immidiate future. Happy crunching!