Stanford does a Berkeley

They do appear to be suffering at the moment - however we can all get work so I don't see a real problem

There's some quite upset people on the official forums at the moment, silly really given Stanford's excellent record for tracing down lost credit :o
 
I for one am definately not complaining - it coincides with my server migration that has meant my fahstat sigs are currently not updating - great because it'll miss out on less stats :D

(the lack of updating is not the cause for the error currently showing in my sig, that is due to me messing around with permissions in an attempt to get the updates working :p).

Even if it weren't for that I can't really see that I would be that bothered, even if it did mean lost credit.

null :)
 
Berserker said:
And then Berkeley comes back and shows how it's supposed to be done - The project's up, but they can't send out new work (and it's going to be a while before they can).

Hehe ah well that doesn't bother me too much, just means I can whoop beansprout even more at BBC CP! :p

You foldy types have had it too easy with your perfect servers, you needed some kind of let down just to show how good the times really are. ;) :D
 
Amp34 said:
Hehe ah well that doesn't bother me too much, just means I can whoop beansprout even more at BBC CP! :p

You foldy types have had it too easy with your perfect servers, you needed some kind of let down just to show how good the times really are. ;) :D

We endured weeks of crappy low pointer WUs! Now everyone has twigged that we need to turn off -advmethods and now 200, 300, 400, 500 and 600 pointers are the order of the day! Let the good times roll!

SiriusB
 
Amp34 said:
You foldy types have had it too easy with your perfect servers, you needed some kind of let down just to show how good the times really are. ;) :D
bah the servers have never been perfect - the system just has plenty of redundancy so when a few servers go down there is still work to be got :cool:

Borkeley always have been and always will be much better at making a whole system grind to a halt - it's just part of their DC heritage :p
 
SiriusB said:
We endured weeks of crappy low pointer WUs! Now everyone has twigged that we need to turn off -advmethods and now 200, 300, 400, 500 and 600 pointers are the order of the day! Let the good times roll!

SiriusB
I may have missed this somewhere along the line, but should we not be using -advmethods in our configs ?
.
 
KE1HA said:
I may have missed this somewhere along the line, but should we not be using -advmethods in our configs ?
.
recently the choice of WUs when using -advmethods has actually been worse than without

we were getting things like this:
Project : 774
Core : Gromacs
Frames : 100
Credit : 41


-- X2-Core0 --

Min. Time / Frame : 6mn 28s - 91.30 ppd
Avg. Time / Frame : 6mn 36s - 89.45 ppd


-- X2-Core1 --

Min. Time / Frame : 6mn 13s - 94.97 ppd
Avg. Time / Frame : 6mn 16s - 94.21 ppd


-- Louise --

Min. Time / Frame : 4mn 49s - 122.57 ppd
Avg. Time / Frame : 5mn 55s - 99.79 ppd


-- Capricorn --

Min. Time / Frame : 7mn 23s - 79.96 ppd
Avg. Time / Frame : 8mn 24s - 70.29 ppd


-- NewsI --

Min. Time / Frame : 6mn 48s - 86.82 ppd
Avg. Time / Frame : 6mn 48s - 86.82 ppd

then on removing -advmethods we started getting these:
Project : 1814
Core : Amber
Frames : 100
Credit : 153


-- Capricorn --

Min. Time / Frame : 16mn 47s - 131.27 ppd
Avg. Time / Frame : 20mn 10s - 109.25 ppd


-- NewsI --

Min. Time / Frame : 17mn 14s - 127.85 ppd
Avg. Time / Frame : 17mn 28s - 126.14 ppd

and even some of these:
Project : 2414
Core : Gromacs
Frames : 100
Credit : 600


-- Capricorn --

Min. Time / Frame : 40mn 35s - 212.90 ppd
Avg. Time / Frame : 52mn 03s - 165.99 ppd


though from the feel of the current work it might be time to see what's back on -advmethods now :o
 
I think that's actually a coincidence, guys :) I haven't done a thing to my various PC's, (they all have -advmethods on) and they've all started getting 600 pointers and other wu's too. (Although the 380-390 pointers are actually *worse* than the small 40's and 120's were)
 
kinnison said:
(Although the 380-390 pointers are actually *worse* than the small 40's and 120's were)
not on my machines they aren't

Project : 2124
Core : Gromacs
Frames : 100
Credit : 396


-- X2-Core0 --

Min. Time / Frame : 45mn 17s - 125.93 ppd
Avg. Time / Frame : 47mn 30s - 120.05 ppd


-- X2-Core1 --

Min. Time / Frame : 45mn 16s - 125.97 ppd
Avg. Time / Frame : 47mn 26s - 120.22 ppd


-- Capricorn --

Min. Time / Frame : 50mn 21s - 113.26 ppd
Avg. Time / Frame : 1h 00mn 59s - 93.51 ppd


-- NewsI --

Min. Time / Frame : 56mn 48s - 100.39 ppd
Avg. Time / Frame : 57mn 12s - 99.69 ppd

they're not good WUs but definately better than what we were getting previously

you may have a point with the -advmethods on/off not making much difference though - most likely it's just down to what server you end up stuck on, though doing anything to get off that server will always help
 
I think there's something in it, though I suppose it could be coincidence. I had small wus for weeks, and as soon as it went off I got 3x 600 pointers on one machine and another on the other machine. Unfortunately the one machine has a 153 pointer now, but I'm sure the 600's will come back ;)

Also, without advmethods you dont have a chance of the massive and rare 1000 point wu :eek: Still I've gone for consistent 600 pointers for the moment.
 
Back
Top Bottom