Folding @ Home F.A.Q.

[pscout]

Originally Posted by sswilson

Just a quick note that fahmon now works right out of the box with the SMP beta.

Seems to me I looked at it a while back and it required a few hoops to jump through with SMP so I set it off to the side to look at later.

I installed it last month and was pleased to see that it worked without any problems.

Yes .. the last upgrade i did to v 2.2.1a fixed a lot of problems. It works very well now.

U_F has done a great job with it adding a few enhancements. Before i switched to smp, i was monitoring up to 35 standard client instances and fahmon was more usable for simple monitoring than emiii, although i ran both.

[pscout]

Originally Posted by 3.0charlie

Thanks Pete, that's exactly what I was looking for.

John

You are welcome.

Always happy to help

[3.0charlie]

Odd message after 44%:

"Gromacs cannot continue further.
Going to send back what have done"

???

[sswilson]

Has anything changed on the folding rig? Did you change the OC on it?

It means that the core program became corrupted, or that it thinks it's corrupted because an OC isn't stable.

Normally it'll dump whatever data you had, and dl a new copy of the core.

[sswilson]

I just had a wierd one myself.......

The laptop was slowing down and finally displayed a "your hard drive is too full" error message.......

At first I thought one of the young lad's had filled it up with some form of crap, but it turned out to be the FAH log file....It had grown to 36 Gig!!!!!!!!!!

That's a hell of a lot on an 80 gig drive. :)

[3.0charlie]

Originally Posted by sswilson

Has anything changed on the folding rig? Did you change the OC on it?

It means that the core program became corrupted, or that it thinks it's corrupted because an OC isn't stable.

Normally it'll dump whatever data you had, and dl a new copy of the core.

It's a customer rig that is back for maintenance - and the owner specified that the OC (E6400 @ 3.4) was gone. He didn't notice when.

So it's unstable - oddly enough, Orthos ran for 25 hours with no errors.

I installed F@H for the fun of it... knowing that it is very sensible to instability.

Thanks.

John

[3.0charlie]

Originally Posted by sswilson

I just had a wierd one myself.......

The laptop was slowing down and finally displayed a "your hard drive is too full" error message.......

At first I thought one of the young lad's had filled it up with some form of crap, but it turned out to be the FAH log file....It had grown to 36 Gig!!!!!!!!!!

That's a hell of a lot on an 80 gig drive. :)

36Gig of text, right? Wow. How long has it been running?

[sswilson]

Long enough apparently.... it's up to 22 WUs since I first installed it.

I didnt' bother to to try to open it.... probably would have crashed on me as I only had 196 meg free space....... :)

[pscout]

Originally Posted by 3.0charlie

Odd message after 44%:

"Gromacs cannot continue further.
Going to send back what have done"

???

This is on standard client?

The parameters used by the researchers in the simulation can result in some wu's not being able to complete. These are not the same as EUE's caused by an unstable rig.

Assuming the results were returned to the server then the wu will be reassigned to 3? other rigs to see if they also error.

If you only rarely see these it should not be an unstable oc causing it.

Some projects are more prone to these kind of errors than others but stanford tries not to release projects past beta or advmethods if the failure rate is too high (~ > a few %) since it can cause a lot of concern with the average folder.

But at that high of an oc i suspect it was unstable.

[pscout]

Originally Posted by sswilson

I just had a wierd one myself.......

The laptop was slowing down and finally displayed a "your hard drive is too full" error message.......

At first I thought one of the young lad's had filled it up with some form of crap, but it turned out to be the FAH log file....It had grown to 36 Gig!!!!!!!!!!

That's a hell of a lot on an 80 gig drive. :)

If the client doesn't get restarted once in a while the fahlog.txt will just keep growing. the verbosity 9 flag makes it worse since more is written to the log.

The fahlog gets pruned once it get past ~ 50 K and the most recent is moved to fahlog-prev.txt .... but this only happens once when the client is started.

But I dunno how you could have filled that much of a log with only 22 wu's tho. The beta does have problems deleting files from the work directory. The biggest fahlog.txt I ever recall having seen was one of my gpu clients that produced ~1000 eue's in less than a day and it was only 1-2 MB.