Distributed Bit-Tech Folding@Home Team (CLOSED: now superseded by joint CPC/bit-tech team)

i've put mine back up, they don't seem to be any cooler today.
(yesterday was the hottest day in May for over 50 years!)

 

Wowza those are some nice graphs. I am happy to see that i am the 4th highest (average) producer at the moment! How did you find that site? I like being able to look at the more advanced stats.

Hint for keeping temps down.

Move to Alaska. The hottest its been this year is 75 F. I love being ale to keep my OC's up and full load.

Anyone else notice that we've jumped 5 team positions in the last 2 weeks? Good job everyone! Keep it up!

-Brad

 

Its a good stats site definatly!

Cant remember how I found it though...

 

My systems are all OC'd and all 24x7 stable... I just don't like getting $200 electricity bills, once it cools off in a couple of months I'll be back to full power again, till than I guess i may lose my 5th place spot...

 

Power went out tonight, so i lost a it of work, and some folding time oh well, i'll just have to make it up.

-Brad

 

Jeez spacehopper your churning out WUs at quite a rate!

 

hehehe

i still need to get most of the rooms transferred over to the new proxy settings too!

 

I'm at max capacity now. Darn you and you more computers

 

Hey can anyone explain to me the cause of 'Timered Checkpoint Triggered' and why it started happening on all 4 of my machines? It seems as if it is really bogging my folders down.

Thanks,
-Brad

 

The timered checkpoint is just where the client periodically saves intermediate results. This means that if there is a sudden crash or powercut or something, only a small amount of calculations are lost.

The default time between checkpoints is 15 minutes and this can be changed by using the -config switch.

Timered checkpoints are a feature of the client and do not affect performance, they are not a problem and won't be the cause of your slow folding. If something else is slowing your folding, then the times between frames will increase and hence more timered checkpoints will be performed between frames since they are taking longer.

I believe it might only do them (or at least only tell you about them) for Gromacs WUs. If you have never noticed them before, it might be because you have changed the -verbosity switch to a higher number, whereby the client tells you in greater detail what it is doing.

 

Well we've now hit 1.5 million points. Keep up the good work guys

 

without extreme cooling? Im using my 30cm fan blowing into my computer to keep it cool!

 

Gonna have my Lab PCs on 24/7 now (rather than 9-11)

production boost

 

starting up again... expecting 100-150pts a day

 

Good good Anything helps!

Was quite a boost eh! Topping 3,500 points a day now

 

Joined and added a few machines, been looking for another dist. comp project since Seti went all wierd looking, hopefully this is the one.

 

wow, its been a while since ive been in here (dont ask, r/l and work has been hell). good to see we are still vhurning out points hopefully, i should be getting most of my machines back on and folding

 

Looks like I may have to retire almost all of my machines from F@H, we have a new program at work that goes skitso when F@H is installed too.

I'm trying to find a way around it but I've only got two weeks...

 

out of curiosity, are your machines on 24/7?

ive had success with machines that are heavily used during the day, and have difficulties running F@H during "working hours" too. i found a rather nice solution, it might be helpfull for your network too. i created a "background local user account", namely "caretaker", for each machine. then logged into it, set up a few items on task scheduler (defrag, backup etc) and removed it from the XP login screen. on the task scheduler i also set up f@h to run during lunch hour, and from 5pm to 9am. the task scheduler is able to kill processes if they have not quit after X hours, and f@h never quits. with the checkpoint saving on the latest clients, the maximum you will lose for a forced kill is whatever has been computed after the most recent checkpoint.

i.e. i set f@h to launch at 1200 and kill if the task has not finished in 1 hour (thus it runs from 1200-1300). then set it to run at 1700 and kill if not finished in 16 hours (thus it runs from 1700-0900).

you can also have the task launched on idle, say, schedule the task to launch every day a 9am, but only if the machine has been idle for 30 minutes, and "retry if the machine is not idle" to 99999 times. you can also stop the task if the machine ceases being idle. so the f@h client will only ever run if the machine is sat idle for 30 minutes, and will be killed off when someone resumes work on the machine.

im not sure how applicable this would be for your network, but its something to think about

 

yer 48 to 52 machines on 24/7.

I'll have a play with the task sched see what I can do...