Smoked em'. Yeah so Matroxusers will be next easy, then the "Dutch Power Cows". hehe bring em on.

I think before long we will be well inside the top 10

;)

I was at work for ~ 3hours before I noticed that it was silent in my office. Must have lost power in a thunderstorm last night. Firing the armada back up now :)

since they killed our internet connection at work, i'm loosing about 10+ 2+GHz p4's at work..., bummer

could you take the work that your 10 pc's at work did and take that home and upload it?
do that once a week or so?

we made some rela good progress this weekend... its going to be a struggle getting to the power cows... as their main dude seems to have fired up half the country (they are a dutch team) folding.

added 15 more machines onto the run for tongiht. Those ****ing cows arent going to get away from us that fast! :)

good idea, would need a usb stick to be workable

It's a nice little set of jobs there Joe - but my environment is a little bit more complicated. It sounds like you're using mostly single-processor workstations to fold at night, while I'm using multi-processor servers. I could cut'n'paste the batch files to double-up everything and run two instances of the service on each machine - but that doesn't help if i want to push it to a quad-processor or two, and I don't think that the service-mode supports more than two instances. Also - using scheduled jobs and services are just two more things that have a possibility of leaving entries in the log files if something wierd does happen.

My current method has shortcuts on the desktop (in my profile only) one for each processor/instance and I remote-desktop to the server (these are all Win2K Server so this gives me a virtual terminal, not the local console), click my shortcuts, and then close the terminal window without logging off. The batch-files I've got of the desktop are set to kill the folding client after 12 hours - so I start them between 5:00 and 6:00 AM (I work later than everyone else - most leave at 4:30), and they shut themselves off early next morning before people arrive (which is good because I start later than everyone else - flex-time is good) I'm spending my free time here investigating a way I can run a single file from my workstation to start all of the clients and the 12-hour timer than kills them, but still have them running as a standard process instead of a service on each server. And for weekends I just have a second copy of the batch file with a 60-hour timer instead. I didn't get to run it last weekend though because I took Friday off to go camping early.

Yes services mode supports 4 instances easy. I run 4 instances right now.

you could easily modify the scripts to support 2 or 4 instances.

Just an idea ..
but what is to stop someone from doing a weeks worth of work without uploading results...
then backing up the entire directory... then uploading the results...
restoring from backup... and uploading the results again.
and again and again.....
I could put us in 1'st place in a day....
it would be wrong... unethical...
am i the only one that thinks that devianetly?
or is that the reason there are a few users putting out large numbers...

NO Joe i don't think your doing that!!!
but otheres that are in first place?

just something to wonder about....

I would assume there was a mechanism in place to stop that from happening.

I'm sure if it were that easy there would be MANY more users putting out millions and millions of folds per day.

And I went through and built some batch files in some free time today. I've simplified my scripts to the point of clicking a single icon to start a 12-hour folding session (I still use timers to stop instead of a scheduled stop time), and that icon will handle pushing the client to boxes that need it, multiple instances per client, etc. It does NOT use AT or any other scheduler service, and it does NOT run them as a service. It uses windows RPC calls to start the process(es) remotely and a handy freeware utility that lets me do these things from the command-line that I've been using for various remote-admin type tasks for quite a while now.

I just started a 12-hour batch to go tonight, and if it worked well in the morning when i get back I'll post up my batch files and a link to said utility.

This is one of the cheats in SETI. After 10 days or so the validation information is reset and you can re-upload the same results over again.

DF uses a results varification method that prevents this from happening. How it works I am not sure but it does work. I would think if someone was good enough at decifering code they could find a work around for this. I suspected CodeBlaster the The Rebels Haven Computer Forum team but then went to there forum and did some digging and it seems he just has a bunch of comps that crunch off line pretty much throughout a whole protein and then he uploads the results before the new protein takes over. Something like that anyway.

funny
speaking of the Rebel geeks...

they just bumped us down to 13th after they seem to have recruited half of the teamless users back to their team... not to worry we will walk over them in no time on our way to walk over the cows.

BTW thats 78 machines now... 78 machines and only 260k/hour

the 1.8 P4's dont seem too powerful the last 30 machines have been all 1.8 P4's Hell I may need to just add 70 more machines on (which I have at my disposal with no problem, and kick this bitch up a notch.. how does 400 - 500k/hour sound?

Do that!

I did some benchmarks at P4 1.5 and they are only a litle more speedy that p3 1000, but hey, its one more pc...

Joe,

Just throw 350 or so computers on our team for a weekend. See if everything works out ok. If they do, then don't run those 350, run the other 350 for a weekend. See if it works ok. If both halves work fine, then run all of them for 1 weekend. I'll wub you forever if you do it! :dome: :dome: :dome:

Joe, you realize that there are only 5 users with higher daily production than you, right? do you have any idea how ridiculous "only" sounds from you? :)

There is only about 10-12 TEAMS putting out more than Joe is! :)

UltraMassiv :D :D :D

just added 20 more ;)

so that should make 98 here, and with the extras I have running its over 100 now.

This should be the cap for a while, I dont think I will be adding any more than that.

Also.. for an upcoming 3 weeks I will be out of the office, durring that time there will be no high end folding love from me. so hopefully Cova can step up by then :) last week in sept is my first time out of the office. Also Thursday I may not kick off the folding cause I may not be in work ( possability I am getting a tooth pulled tomorrow night.

We got 53 people with points on the board! Rock on! Welcome new guys! :)

I am doing 325K now and the team overall is at 540 - 600k an hour
;)

we are going to be up 4 or 5 million by morning at this rate :)
we will pass that Rebel computing again in a day at this rate. and in a couple days we should be past the cows... then Anandtech falls fast. and we are into the top 10 ;)

200K from taking 12th place!
 

Then we have a haul up to 11th... the cows got a lil power going, but I think by the end of the weekend we will get em.