GPU WU server down?

[MikeB12]

OK, here's one...

the fah server has gone down more than Paris Hilton...

[Kondik]

Britney Spears ?

[angra]

My office is blissfully quiet....

but no folding being done

[MikeB12]

I just had 2 that got thru 6 and 8 min ago, so maybe things are picking back up..

[Sparky]

My office is blissfully quiet....

but no folding being done

What? Why? I see you've done near 13K today

[angra]

Sparky - I meant from the WU server being down. There is a noticeable noise difference between folding and not folding.

I even can tell a difference from when I sent one of my GPU down the hall to a student to use for some benchmarking.

[Naja002]

Just a couple of thoughts:

I've been having this random inability to upload WUs also. 1 thing that works sometimes on GPU2 is closing the client
and restarting it.

Another thought for those who also run Boinc: GPUGRID.net now has a
GPU client for Boinc. Its not completely....what?....Stable?...No, that's not the word. I don't know....it
can be somewhat of a pita, I guess. But there are work-arounds and they are seriously working on the client.
It runs pretty well overall...with a little bit of understanding. Anyway, anybody that interested in filling in some of
the F@H down-time with another GPU client should check it out. There is an XS team already. I'm #1.....uh....out
of 2 active and 4 total. Anyway, I've no doubt that they can use the help with getting WUs completed and
feedback on the forums would be a big plus...no doubt. If anybody decides to check it out and needs
some help--just let me know.

[Shadowtester]

The servers at F@H have been horrible the last couple of months now I do not think they have made it though a weekend without going down in serveral months. I am not sure even Britney or Paris could claim that one.

[Rise]

it seems to be back up and running now...

sucky servers

[[XC] Synthetickiller]

The servers at F@H have been horrible the last couple of months now I do not think they have made it though a weekend without going down in serveral months. I am not sure even Britney or Paris could claim that one.

So many things to say, but I decline to comment. I don't want to get banned.

[Sparky]

Sparky - I meant from the WU server being down. There is a noticeable noise difference between folding and not folding.

I even can tell a difference from when I sent one of my GPU down the hall to a student to use for some benchmarking.

Oh... that's saddening

[Bobsama]

Too bad: I noticed my HD3850 being quiet for a while. Oh well: it does dozens of WUs daily, so I'm happy I'm contributing to science.

[WFO]

So many things to say, but I decline to comment. I don't want to get banned.

Last nights little outage appears to have cost me at least 10K today. Coo-Coo doesn't get mowed down till tomorrow.

[MikeB12]

again http://171.64.65.20/
http://fah-web.stanford.edu/serverstat.html
171.64.65.20 GPU vspg1 densign full Reject

[Sparky]

8800GT and 9600GSO are both idle right now

[Loony]

Weaksauce...

[p2501]

One 9600GSO is idling quite some time now, the other one is 10% away from running dry. Cheers Standford, they don't seem to have ever heard of redundancy.

[Sparky]

One 9600GSO is idling quite some time now, the other one is 10% away from running dry. Cheers Standford, they don't seem to have ever heard of redundancy.

Well that's the thing, there is more than one server that does GPU units, but for some reason the clients are almost always directed to this one particular server that happens to be the most unstable one

[[XC] Synthetickiller]

Its down for me right now.

2 of 4 cards folding atm. I'm taking a nap. The servers better be up when I return of there's hell to pay!

[Shadowtester]

Great I have a gpu2 client just about to finish and I read about Stanford's dodgy servers going down again just great!

[p2501]

Well that's the thing, there is more than one server that does GPU units, but for some reason the clients are almost always directed to this one particular server that happens to be the most unstable one

Yeah I don't understand their setup at all. Why not setup a redundand load balancing based on some OpenBSD machines (or what ever) so the clients on the outside talk to just one IP. Behind this load balancer they then setup their little server park. Nice and easy. Why rely on this erroneous round robbing on the client side of things.

[MikeB12]

again this morning since

00:50:20 PDT 2008 171.64.65.20 GPU vspg1 densign full Reject

http://fah-web.stanford.edu/logs/171.64.65.20.log.html

[DAK1640]

Mornin Mike...Yepper, the server is doing the Paris Hilton again. Couple that with the windows update. Huge interference with our production.

[MikeB12]

yeah, I got a winders vista update last night on all 4 machines at 3am... l have them set to auto update... and it was a reboot update. I guess I should change them back to "download updates only", but last month I got sick of manually doing it.. so I set them to full auto... it's a pita to go around and click install, then wait and wait while it updates, then reboot each one. I'd rather wake up and find them at the login screen already updated. so I have them on all on full auto.

[Rise]

yeah, I got a winders vista update last night on all 4 machines at 3am... l have them set to auto update... and it was a reboot update. I guess I should change them back to "download updates only", but last month I got sick of manually doing it.. so I set them to full auto... it's a pita to go around and click install, then wait and wait while it updates, then reboot each one. I'd rather wake up and find them at the login screen already updated. so I have them on all on full auto.

me too

i hate waking up to that