Folding @ Home F.A.Q.

[pscout]

Originally Posted by MacJunky

Would it actually say I was having EUE issues? Because I was getting no errors at all. If it does keep track of errors, would there be a way to check past WUs?

If the machine was just freezing than it would not necessarily have any errors unless a wu got corrupted by freeezing while writing to the work directory, which is unlikely unless it was freezing very often.

You can look in the fahlog.txt file as well as the archived log called fahlog-prev.txt to see if you have had early-unit-end messages, or other error messages. I have no experience with the mac client so am not sure if there are any mac specific things to look for beyond what the client logs.

Here is a link to the mac osx client forum at fc which may have some more light to shed on it. The most expert mac folder i know is a mod there - Macaholic - although i haven't seen him around the fora i visit much lately.

Folding-community.org :: View Forum - Mac OS X client [PPC or Rosetta]

[MacJunky]

[offtopic]Well, at that time it was not freezing often (unless I cranked my graphics card too high. But even then it was still kinda barely partially functional most of the time) so much as it was losing power and booting up again. It never did this with it's old PSU. Though then again with it's old PSU it needed around 5 boot attempts to actually cold boot last I checked and if I continued to use that PSU I bet it would have continued increasing (worked fine during warm reset). However with this PSU it cold boots in one go.
Heh, I need a _good_ PSU that is capable of running my system without issues. In the meantime I setup a second PSU to offload around 80w-100w ish at most from my primary PSU. Hopefully it will help a bit. No, I have not wired the PSUs directly together. I am not that stupid.
Now that I think of it, I should probably actually update my sig with the two new PSUs so as to not confuse people who might actually read sigs.[/offtopic]

Anyway, I just checked ~/Library/Folding@home/FAHlog.txt and I can see no mention of EUE issues at all in it.
This is at the end of the section with the WU in question:
[15:22:03] Finished Work Unit:
[15:22:03] - Reading up to 232536 from "work/wudata_01.arc": Read 232536
[15:22:03] - Reading up to 452908 from "work/wudata_01.xtc": Read 452908
[15:22:03] goefile size: 0
[15:22:03] logfile size: 169004
[15:22:03] Leaving Run
[15:22:07] - Writing 1052172 bytes of core data to disk...
[15:22:07] ... Done.
[15:22:07] - Shutting down core
[15:22:07]
[15:22:07] Folding@home Core Shutdown: FINISHED_UNIT
[15:22:07] CoreStatus = 64 (100)
[15:22:07] Sending work to server


[15:22:07] + Attempting to send results
[15:28:37] + Results successfully sent
[15:28:37] Thank you for your contribution to Folding@Home.

>_< I have spent at least an hour continually tweaking and adding to this post. I think it is close enough to done as-is. I hate it when I do this.

[pscout]

The log snippets you posted look fine.

15:22:07] Folding@home Core Shutdown: FINISHED_UNIT

The finished unit message means sucessfully completed ... an EUE would not log this message.


[15:22:07] CoreStatus = 64 (100)
[15:22:07] Sending work to server

This is also an important message since it means the EUE or successful wu results will be reported. There have been problems, but mostly in beta testing where a failure is not reported to the server which is bad since the server cannot reassign the work to another rig until the preferred deadline for the wu passes.


[15:22:07] + Attempting to send results
[15:28:37] + Results successfully sent
[15:28:37] Thank you for your contribution to Folding@Home.

These messages tell you that you wu is uploaded ok and your points for it will get credited.

Since the single core wu's can take 3-4 days on a fast machine to finish, if the rig keeps crashing on you due to the constant 100% load folding creates, it is not suprising that you only have 1 wu credited.

If you want to keep trying to run it, i suggest changing the checkpoint value in the config (under advanced options) to the minimum time - 3 minutes i think. The default is 15 minutes. The client will restart from the last checkpoint which is either 15 minutes, or every 1% complete, whichever comes first. So with a 15 minute default, when your rig reboots you lose up to 15 minutes worth of work. Shortening the checkpoint time to the mininmum 3 minutes will reduce the amount of lost work. But if it is rebooting that constantly, I doubt you would want to keep restarting and it is also not good for your rig. And it is likely the 100% folding load is contributing to these constant reboots. So I would not bother folding until it can run stable for long periods.

[MacJunky]

I know about the checkpoints and it was one of the first things I changed when I started. I _always_ poke around an application's preferences before actually using it. As to my G4's previous restart rate with this PSU, I would say something like once per day or two under constant heavy load I suppose. Never really recorded it.

But, as I said, I added a second PSU for the first time since it has been out of it's case and some extra fans. Want to ditch this damn 50mm evercool EC5010M12CA though, it is total crap and the bearings started failing less than a year after being purchased new. I will run some system stressing apps and stuff for extended periods of time and see if I have any issues.
About me only having done one wu, I stopped after that first one.

[3.0charlie]

The little rig that could just may become the Big Rig That Can...






I smell a Q6600. I'll know more in a week's time...

[pscout]

Originally Posted by 3.0charlie

The little rig that could just may become the Big Rig That Can...






I smell a Q6600. I'll know more in a week's time...

/crosses fingers .... /

[qwerty]

Yeah I just installed F@H on a p3 733 and it is ridiculously slow, I reme,ber you guys warning but would I listen? Of course not...Qwerty must find everything out the hard way, naturally.

I think that I'm going to wait until all my new parts for my new rig arrive then go again. Should be by the next week I am hoping. Damn mail service...

It is trying to tell me that The WU will not be done until Feb 8, 2008...LOL.

Madness!

[Babrbarossa]

Trying to start folding- Anyone know what's wrong?

Couldn't open fahcore_78.exe1
Error: could not extract core
Core download error #7, waiting before retry

I downloaded the text only console for vista

EDIT: never mind- got 'er going now.

[Mr. Cipher]

I've already been tagged thankfully for my 2 computers, Eir and Eir_DualCore. The former isn't running atm but I plan on getting it going soon again.

My issue is that I'm running an e2180 but stanford's stats keep on saying that the # of active cores on my computer is only 1, rather than 2.

I read the entire FAQ, here and on other sites. Am I missing something? I want to get as many points as possible for the team!

[3.0charlie]

Are you running the SMP client? If so, it will detect 2 cores no problem. As for Standford, one active core equals one CPU. A Quad does not equal 4 active cores....

[Mr. Cipher]

I'm running two command-line clients now. I went through and installed them in two different directories, assigned affinity through the setup and changed the registry settings. It's all good now! Thanks :)

I wish I understood why some people have low WUs and wickedly high points, and then people like myself have low WUs and very little points. Either way, I barely notice it now (command-line vs. gui)

[3.0charlie]

It depends which client you are using. GUI<command-line client<SMP Win<SMP Linux, and somewhere in there the GPU client. A major factor also is the CPU you are using, since the WUs assigned are different depending how many cores FAH recongnizes.

[Mr. Cipher]

I D/L the FAH504-console, I don't know if that is the SMP client or not.

I have an e2180 @ 3.4ghz at the moment.

[lcdguy]

Originally Posted by Mr. Cipher

I D/L the FAH504-console, I don't know if that is the SMP client or not.

I have an e2180 @ 3.4ghz at the moment.

That would be the regular console client. The SMP client identifies it self as an smp client in the name. Also the installation process is longer than the regular client.

[Cptn Vortex]

Hey all, my quad rig is taking a little break.... turns out my somewhat gimped Maximus board is now fully Gimped....It wont last a day running....whenever I come home from work it's still running, but I cant get a display... I feel so dumb now for when I forgot about that silly mobo standoff that was definitely in the wrong place..... Oh well, just means spending more money!