I'll knock off the app_config.xml temporarily and do a work fetch this evening, so I can check what the server response looks like after that change.
That still got me tasks set to use 0.9 CPUs - which isn't really much use (see GitHub issue 2949). I think the runtime estimate has increased, but I wasn't specifically watching for that. Take 2, tomorrow morning.
My host with a GTX1660s is out of all GPU work and is unable to to get any, meanwhile the other host with 2 3GB GTX1060s has a fair number of GWs and has snagged the odd pulsar,
To add to the fun neither of my back up project, Asteroids and GPU seem to have work. The bottom line is the GTX1660s done nothing since I got up today.
I've only been getting the new O3ASE tasks on Linux. The Windows machine that should be running O2MDF seems to have reached the same state as yours, except that it caught one of the rare GPUGrid tasks this morning.
WCG has a lot of GPU tasks at the moment, so I'll let them run overnight and see how things look in the morning.
My thread. I mentioned WCG for Betreger, as a possible alternative GPU project while we're transitioning from O2 to O3 here. Tullio's brief response with running times (which I hadn't mentioned) is welcome here.
I solved my work fetch problem on the GTX1660s host. For some reason I had not enabled beta tasks and when I did O3 work gushed forth and it is happily crunching away.
Ian&Steve C. wrote: this
)
I think you've answered both your own and Bernd's question.
Richard Haselgrove
)
That still got me tasks set to use 0.9 CPUs - which isn't really much use (see GitHub issue 2949). I think the runtime estimate has increased, but I wasn't specifically watching for that. Take 2, tomorrow morning.
I agree that 0.9 isn't much
)
I agree that 0.9 isn't much use.
for nvidia cards, best to just ship them out as 1.0 CPU for 1.0 GPU, as that's what's actually required.
_________________________________________________________________________
My host with a GTX1660s is
)
My host with a GTX1660s is out of all GPU work and is unable to to get any, meanwhile the other host with 2 3GB GTX1060s has a fair number of GWs and has snagged the odd pulsar,
To add to the fun neither of my back up project, Asteroids and GPU seem to have work. The bottom line is the GTX1660s done nothing since I got up today.
I've only been getting the
)
I've only been getting the new O3ASE tasks on Linux. The Windows machine that should be running O2MDF seems to have reached the same state as yours, except that it caught one of the rare GPUGrid tasks this morning.
WCG has a lot of GPU tasks at the moment, so I'll let them run overnight and see how things look in the morning.
I am running
)
I am running OpenPandemics-COVID-19 on my GTX 1060 and they take about 20 minutes.
Tullio
tullio wrote: I am running
)
What does this have to do with GW GPU tasks on Einstein?
_________________________________________________________________________
off topic
)
off topic
My thread. I mentioned WCG
)
My thread. I mentioned WCG for Betreger, as a possible alternative GPU project while we're transitioning from O2 to O3 here. Tullio's brief response with running times (which I hadn't mentioned) is welcome here.
I solved my work fetch
)
I solved my work fetch problem on the GTX1660s host. For some reason I had not enabled beta tasks and when I did O3 work gushed forth and it is happily crunching away.