About every 3rd work unit, I am getting one of the sidechain 10 step WU, and I get through maybe 4 steps and the client just hangs in a loop or something, my processor usage is still 100%, but nothing happens. I have to shutdown the client, delete the WU, and start over. It's starting to get really annoying. Anyone else seeing this behavior?
Gyro77
Yes.. I have seen it and yes it irritates the #$$%(% outta me...
I about to give the Stanford projects a long needed boot from my boxes
-Jim
PS: any other suggestions???
Seti, RC5, G@H, F@H, Ubero.... been there, done that...
how does D2OL behave in a work environment? ie: can you hide it?
Jim, I tried D2OL for a day. It has a hide feature. Once a day stats, and you would be the number one producer every day, and if I took my four boxes there, I would be #2.
It is a baby project. not many people on it.
Colin Thompson
Or you can try brand new distributed folding project (aims same way as G@H).
It features really nice ASCII graphics (console ver.) and client's very stable...
Give it a try: www.distributedfolding.org
Free-DC handle: 5sa2okpe
rm -Rf /
Okay, so it's not just me. VERY annoying though, is pointwood around here anywhere? If so, could we have you post to the yahoo group about this issue? Or maybe its there already? {Gyro77 lurks over tot he yahoo boards....}
Gyro77
If any of you are considering Ubero, I recommend against it. Their servers are overloaded now. Try something else for a while until Ubero gets some new rack mounted servers.
So Jim, where are you going? I might tag along so you can have someone to beat
Colin Thompson
I haven't had any problems with the sidechains. Is it a common problem? So far I've only done 9 of them.
Justin
Jim - did you try UD ?
Pretty client - very stable. Works well with FireDaemon. Has tons of stats, results - points - CPU time. We don't have much of a Free-DC team on there now, only ten of eleven members.
No native SMP or *nix support but other people have found ways around this.
Maybe this thread should be moved to Lounge or Lobby to get other opinions ?
me neither, are you sure you waited long enough?Originally posted by JustinFields
I haven't had any problems with the sidechains.
Justin
even though they are only ten frames, each frame is very loooooooooong
i pulled one up this morning on a celeron 600 and it was on frame 5
it took ~1-2 hours to move to frame 6, it is now on frame 8, so it is taking as long as any of them, just in fewer frames
Use the right tool for the right job!
On one machine, it is taking about 200 minutes per frame. That is on an Intel 450 mhz. So yeah, it is slow, but it hasn't hung the machines or the Folding core process. I've gone through 9 whole sidechain WUs, on various machines without problem. None of the machines are real screamers.
2 PII 266
1 PII 300
3 PII 350
2 PII 400
2 PII 450
1 PIII 667
1 PIII 900
2 Celeron 900
1 PIII 933
I'm about to get two more machines and bring them into F@H. They are Pentium III 933mhz boxes. Each has dual processors.They'll be cranking some WUs, compared to most of the other slow dogs I'm running. The two new machines will add 47% to my cumulative mhz in Folding.
Justin
Last edited by JustinFields; 02-07-2002 at 03:59 PM.
They seem to only be getting stuck on my XP1800+, my 1Ghz Celeron hasn't had a problem, and neither has my PIII-733, but my Athlon has gotten 3 of these 'bad' sidechains. And I know I am waiting long enough, I let it crunch the frame it was 'stuck' on overnight each time, if that proc can't do a frame in 7 hours, something isn't right. HAven't had any problems the last day and a half though, maybe it was just a few bad WU's? Hope so.
Gyro77
Posting Permissions
Reply With Quote