Caches are depleting and nothing is coming in. http://milkyway.cs.rpi.edu/milkyway/server_status.php reports 0 ready to send :( .

Yup. Been dry for a few hours now....

Is anybody knows, how long this outage will be ?
Should I search for "spare" WU's from other projects ?

Looks like a few hours ago the problems started as stated in this thread. Guess we'll find out, won't we...
My comps run several projects, so they'll catch up on the missed MW WUs later.

Not a word from Travis. I'm gone to Collatz Conjecture.

For CPUs one of the best projects is WCG in that they rarely go down and I don't think they ever have run out of work (if you select all projects or select one or more but allow work from any if the selected are out). Something about getting backed by IBM I suppose ...

The only depressing thing is that if you are ATI heavy MW is still the only port for work (unless I missed something elsewhere) ...

On the CUDA side I still think that GPU Grid is the best place (stablest application and rarely work outages; downside is that you have to have a pretty good card) ... but that is just me ...

Paul

Collatz ATI. I'm off to try it again now.

Paul

Collatz ATI. I'm off to try it again now.

My god, thats what I get for taking 4 months off due to illness.

I just managed to get it started and running ... :)

I had to make the ATI change to force the use of the ATI GPU, but, 6% done and only time will tell if it works.

{edit - add}
The good news is that maybe I can keep the ATI card humming even if Collatz is not the best choice ... the point is to keep the processors humming ... and that is the whole point ... :)

Validator crapped again. Wanna take turns kicking it?? ;p
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Something is happening, I just got 6 WUs on one of my boxes.

Still have a bunch of pending tho'.

____________

Calm Chaos Forum...Join Calm Chaos Now

Like Labbie, I have a load of pending stuff. But the server statement has been like this for hours.

Any sign the system will restart anytime soon?

Travis?
____________
Go away, I was asleep

Looks like the validater stopped around 10:00 UTC, possibly a bit later. All the WUs uploaded from my comps after that are still pending. Guess the process creating WUs stopped as a result of the stopped validator or something like tha.t. Labbies 6 WUS look to me like some of the always present few resents which somehow still get resent.
We'll find out sooner or later but ofcourse we prefer to hear from Travis or see WUs flowing again sooner rather than later :).

Yeah, I think they were resends. Got a couple more not too long ago that were aborted by someone.

And the funny thing is, I've had three on one machine go thru and validate. Looks strange (for MW) to have a few completed WUs in the middle of a bunch of pendings.

HostID 23774


____________

Calm Chaos Forum...Join Calm Chaos Now

All green on the server status, got a full load on my i7 a few minutes ago, but not on the other machines.

All my pendings are gone too!!! ;)
____________

Calm Chaos Forum...Join Calm Chaos Now

24 tasks received.

Heat output nominal.

Disaster averted.
____________

Not quite sure what happened but I just restarted things not too long ago. Gonna have to watch it and see what's going on.

Not quite sure what happened but I just restarted things not too long ago. Gonna have to watch it and see what's going on.

Travis, it's because you are trying to have a real life.

Computers are vindictive and mean. ;)

>weekends
Or Tuesdays.
Except that for them every day is Tuesday :P
____________
Cheers,

PeterV

.

Not quite sure what happened but I just restarted things not too long ago. Gonna have to watch it and see what's going on.

Travis, it's because you are trying to have a real life.

Computers are vindictive and mean. ;)

If i go out skiing or to some party, almost invariably the server crashes while i'm away :)

And then there's the old 'CD drive in the server opposite, carefully positioned such that the door opening hits your reset button.' So when your wireless buzzer goes off, CPDN staff have had them for years, you just telnet into the server [t'be shure, t'be shure], and when that doesn't work telnet into the one opposite and run 'cd-open, cd-close'.
____________
Cheers,

PeterV

.

And then there's the old 'CD drive in the server opposite, carefully positioned such that the door opening hits your reset button.' So when your wireless buzzer goes off, CPDN staff have had them for years, you just telnet into the server [t'be shure, t'be shure], and when that doesn't work telnet into the one opposite and run 'cd-open, cd-close'.

Good one Peter, like it!

Cheers,

Chris S

Looks like the problem has resurfaced == no work to send and pendings are growing again. 9:45AM PDT
____________

Yep, Results ready to send 0 and milkyway_av Not running - same as yesterday.

Thanks -- though I suppose if it does crash, we'll let you know ;)

I think i've fixed it. If it doesn't crash in the next day or so it should be good :D

If i go out skiing or to some party, almost invariably the server crashes while i'm away :)

If i go out skiing or to some party, almost invariably the server crashes while i'm away :)

You're lucky, computers tend to crash when I get near them :D

Funny thing was that I was just out to lunch -- come back and voila the server had crashed :)

Right.

From the floor and through the Chair, I propose New Rule 37B

"Admins may not go to lunch. If they have to eat they will bring in sandwiches"

I so move.
____________
Don't drink water, that's the stuff that rusts pipes

second
____________

Motion carried.

Now, what about overnight and weekends?

In short - they're not allowed to leave the building. :)
____________
Cheers,

PeterV

.

In short - they're not allowed to leave the building. :)

Skiing or a party can be out of pager range.
Should work within the server building, though.
____________
Cheers,

PeterV

.

Be a spendthrift! Get a pager! :)
____________
Cheers,

PeterV

.

Got a Windows machine sitting somewhere? Then it *might* be possible have one of these alert you via email or sms:

http://www.ks-soft.net/hostmon.eng/
http://www.whatsupgold.com/
____________

Hmmmm

9/1/2009 9:13:14 AM|Milkyway@home|Message from server: Project is temporarily shut down for maintenance

I think i've fixed it. If it doesn't crash in the next day or so it should be good :D

NO WU !!!

Here we go again for our new weekend ritual.
____________

4870 GPU
4870 GPU

All I can get is requesting tasks for GPU. My GPU's are running GPU-grid. That is on the boxes with GPU's. The non-gpu boxes don't seem to be getting anything at all.

I want some CPU WU'S, but cannot get the darned server to give me any! Well, guess the other projects will kick in higher up the chain.

Dave
____________

There has been an apparent server crash -- at the moment one can't report work or download new work. We can hope that Dave or his DS (Designated Stickee) is aware of this and is in the process of bringing things back to life.

That being said, it IS Friday afternoon and there has been no comment from the RPI side of the world regarding this outage since it first surfaced a couple of hours ago.

So we might be in for a long wait.....

All I can get is requesting tasks for GPU. My GPU's are running GPU-grid. That is on the boxes with GPU's. The non-gpu boxes don't seem to be getting anything at all.

I want some CPU WU'S, but cannot get the darned server to give me any! Well, guess the other projects will kick in higher up the chain.

Dave

I just got some work but out of the 24 units I got 23 finished in 0 seconds. I have no way to see if they are errors yet but I would guess that they will be. I've not had any problem with errors in the past. May be a new run?
<edit>
units finished in less than 1 second and got 0.08 credits. WTF?
____________

4870 GPU
4870 GPU

There are 0 WU ready to send, whats up??
;-p
____________

Again, no work.

Saw that for 5 minutes or so. Back to normal now, must have been new searches being implemented.

So far all seems to be working OK.

But, there must be some server hold ups due to demand. This is stopping my RAC building back to it's normal level, when trying to recover from a 4 day dry due to a failed PSU.

Total RAC was climbing slowly back, needing another 32K for normality, but it is drifting down very slowly.
____________
Go away, I was asleep

Results ready to send 19 ((

Results ready to send 0 :(

We are pulling them faster than the server is making them... :(

But I just got a batch.

I also just tired and got a selection of Collatz tasks (2.0) to see how that works and as a safety project for my ATI card...

WTF, is the problem today?
____________

WTF, is the problem today?

>MW
As does cpdn, except that it waits until the pubs are closed and everyone's laying in an Oxford gutter somewhere.
>SaH
But every day's tuesday there. :)
____________
Cheers,

PeterV

.

And now is validator down
____________

And now is validator down

And now is validator down

And just when I was going to try doing CUDA ... rats! :(

My 4850 is definitely outta work, but the CPU is crunching Einstein.

The 3850 is just uploaded it's last result and crunching Malaria.

The rest (all CPU based) are crunching OK for the next little while.

I see the number of results "in progress" has dropped nearly 10K since I last posted 15 - 20 minutes ago.
____________
Go away, I was asleep

Let's think positive,the site is much faster now!

max

The 3850 is just uploaded it's last result and crunching Malaria.

The 3850 is just uploaded it's last result and crunching Malaria.

Malaria has a GPU application?

First I heard of it ...

Let's think positive,the site is much faster now!

max

Malaria has no GPU application, especially ATI based, it was the way I worded it badly.

The CPUs are crunching Malaria, usually in parallel with MW, and FreeHAL across the top of the CPUs when it is running. I do Einstein in parallel with MW on the other rig.

I reported a completed Collatz work unit about an hour ago with no problem and then when I tried to visit the site it was unresponsive.

Most likely MYSQL died again.
____________

Collatz is working now...

Somethings going on as I got a MW unit a few minutes ago...only 1...and its long gone and done now.

http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=49956
____________

Collatz is working now...

OK, Time for a little tough love.
It's time for:....
a. Project split GPU/CPU
b. Server repair (probably not the problem)
c. Server replacement or upgrade
IMO a or c or a and c is the best solution
You've had donations ( however no statement has been issued as to how well the donations went) and offers have been made for hardware donations (i've not seen any requests for these).
What can we do to help you help us?
We want to help and I believe we've shown that. How are we supposed to do that without requests for specific items or goals we need to reach in order to accomplish specific upgrades.
Once again, help us to help you.
JMO
Randy
____________

4870 GPU
4870 GPU

OK, Time for a little tough love.
It's time for:....
a. Project split GPU/CPU
b. Server repair (probably not the problem)
c. Server replacement or upgrade
IMO a or c or a and c is the best solution
You've had donations ( however no statement has been issued as to how well the donations went) and offers have been made for hardware donations (i've not seen any requests for these).
What can we do to help you help us?
We want to help and I believe we've shown that. How are we supposed to do that without requests for specific items or goals we need to reach in order to accomplish specific upgrades.
Once again, help us to help you.
JMO
Randy

Results ready to send 0

eports 0 ready to send(

Got a few 'ubnable to contact the database', but now it seems that things are working and the server is slowly catching up.
____________

Here is another no work thread - Just bumping it to the top!
____________
.