Getting rejected by assignment server (GPU)

[mlpickerel]

Looks like it's back up... I just got an 8018 WU.

Mark

[netblazer]

TURN OFF AUTOMATIC UPDATES!!!

[ThePretzelGuy]

got up this morning and came to my homeoffice. Room was at least 4-5 degrees colder than normal. I thought -I know what happened -checked HFM and sure enough I have 3 slots waiting for assignments . Maybe we should startup a fund we could all paypal to that buys PG a backup server. I'll bet what we have saved in electric bills out east here would easily buy one or two new rigs!

[ThePretzelGuy]

Oh Snap!
Good news-

[ThePretzelGuy]

Oh Snap!
Good news-
just got three new WUs

Bad news-
they are all 8018s

Will these never end? Even the 8900 was better for me, and they were slow with no QRB points

What is the link to see what projects are set up waiting for assignment?

[Macaholic]

Yes. It appears to be fixed, for now. Last attempt for my six GPUs was 45 minutes ago. Back to space heating.

[bruce]

What is the link to see what projects are set up waiting for assignment?

The link at the top of this page says "Project Summary " but updates to it are far from frequent.

[ThePretzelGuy]

Oops! Google Chrome could not connect to fah-web.stanford.edu
Try reloading: fah-­web.­stanford.­edu/­psummary.­html


looks like that server is affected as well

thanks for the tip

[johnblommers]

"I felt a great disturbance in the Force, as if thousands of GPUs suddenly cried out in terror and are without work units. I fear something terrible has happened." REF

"The haggis is in the fire for sure." REF

[ChristianVirtual]

Thanks for fixing so quickly.


That's how we like it, nice and warm

[Clyde46]

Got one WU on my 580 rig thats on clean up? Its done 100% but it hasn't uploaded anything.

[johnblommers]

So my FAH client was going to wait 45 minutes before requesting another GPU work unit. Being rather impatient I opened a terminal and entered:

sudo /etc/init.d/FAHClient restart

and sure enough it requested and got a GPU work unit with sugar on top. The room temperature is climbing again to normal.

[PantherX]

FYI, if the GPU Slot is waiting for a WU to be assigned and the timeout is several minutes, pause the GPU Slot and then Fold it. This will automatically bump the timeout from several minutes to a minute or so which means that you can avoid restarting the client. This works in V7.2.9 so in theory, V7.3.6 should also support it.

[ChristianVirtual]

Tried that, too. But seems 7.3.6 keep the waiting time over the pause/start cycle. For me also a restart was the only way to skip the hours of waiting time.

[DocJonz]

Tried that, too. But seems 7.3.6 keep the waiting time over the pause/start cycle. For me also a restart was the only way to skip the hours of waiting time.

Works for me on 7.3.6.