problem sending results... for 3 days now.

A forum just for SPCR's folding team... by request.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

problem sending results... for 3 days now.

Post by riffst3r » Sat Jun 07, 2003 9:11 pm

i just can't seem to cash in my WU points. here's a portion of my logfile:
--- Opening Log file [June 8 04:10:57]


# Windows Console Edition #####################################################
###############################################################################

Folding@home Client Version 3.24

http://foldingathome.stanford.edu
email:[email protected]

###############################################################################
###############################################################################

Arguments: -forceasm

[04:10:57] - Ask before connecting: No
[04:10:57] - Use IE connection settings: Yes
[04:10:57] - User name: riff (Team 31574)
[04:10:57] - User ID = 100CF04F7E1582C2
[04:10:57] - Machine ID: 1
[04:10:57]
[04:10:57] Loaded queue successfully.
[04:10:57] + Benchmarking ...
[04:11:00]


[04:11:00] + Attempting to send results
[04:11:00] + Processing work unit
[04:11:00] Core required: FahCore_65.exe
[04:11:00] Core found.
[04:11:00] Working on Unit 02 [June 8 04:11:00]
[04:11:00] + Working ...
[04:11:00] Folding@Home Client Core Version 2.47 (June 14, 2002)
[04:11:00]
[04:11:00] Proj: work/wudata_02
[04:11:00] Done: 22738 -> 142972 (decompressed 628.7 percent)
[04:11:00] nsteps: 5000000 dt: 2.000000 dt_dump: 250.000000 temperature: 298.000000
[04:11:00] xyzfile:
[04:11:00] " 393 p639_L939_K12M_nat
[04:11:00] 1 N 5.512050 -9.046571 -4.470736 20..."
[04:11:00] keyfile:
[04:11:00] "parameters ./proj639.prm
[04:11:00] NOVERSION
[04:11:00] ARCHIVE
[04:11:00]
[04:11:00] cutoff 16.0
[04:11:00] taper 12...."
[04:11:00]
[04:11:00] Hashes matched on file work/wudata_02.dyn
[04:11:00] ARC file integrity verified
[04:11:00] Restarting from checkpointed files.
[04:11:00]
[04:11:00] Protein: p639_L939_K12M_nat
[04:11:00] - Run: 184 (Clone 33, Gen 1)
[04:11:00] - Frames Completed: 192, Remaining: 208
[04:11:00] - Dynamic steps required: 2600000
[04:11:00]
[04:11:00] Writing local files:
[04:11:00]
[04:11:00] parameters work/wudata_02.prm
[04:11:00] - Writing "work/wudata_02.key": (overwrite) successful.
[04:11:01] - Writing "work/wudata_02.xyz": (overwrite) successful.
[04:11:02] - Writing "work/wudata_02.prm": (overwrite) successful.
[04:11:02] - Writing "work/wudata_02.key": (append) successful.
[04:11:02]
[04:11:02] PROJECT="work/wudata_02", NSTEPS=2600000, DT=2.0000, DTDUMP=25.000000, TEMP=298.00
[04:11:03] Couldn't send HTTP request to server (wininet)
[04:11:03] + Could not connect to Work Server (results)
[04:11:03] - Error: Could not transmit unit 01 (completed June 5). Keeping unit in queue.

[04:11:03] TINKER: Software Tools for Molecular Design
[04:11:03] Version 3.8 October 2000
[04:11:03] Copyright (c) Jay William Ponder 1990-2000
[04:11:03] portions Copyright (c) Michael Shirts 2001
[04:11:03] portions Copyright (c) Vijay S Pande 2001
[04:23:23] Finished a frame (193)
[04:36:35] Finished a frame (194)
any help would be greatly appreciated.

miker
Patron of SPCR
Posts: 798
Joined: Sun Aug 11, 2002 3:26 pm
Location: Akron, OH (The Rubber Capital)
Contact:

Post by miker » Sun Jun 08, 2003 6:34 am

Are you behind a firewall or anything like that? Has this wroked before from the same PC? Is the Windows firewall active (if you have WinXP).

Seems like a connection problem.

Wrah
Patron of SPCR
Posts: 316
Joined: Thu Apr 10, 2003 1:56 am

Post by Wrah » Sun Jun 08, 2003 6:38 am

That WU is from project 639 and that means it's assigned to server .143. I've been having problems connecting to that server too the past few weeks, and not just me: http://forum.folding-community.org/view ... sc&start=8

After a few days of trying I suddenly do get the WU returned and everything works ok again for a week. And I only had problems at work (= clients are behind a proxy). Very strange. What I also noticed is that the client gets confused after an unsuccesfull attempt, at the first attempt it reported it was contacting the .143 server, after that it reports a bogus ip-address (something like 10.0.0.13). I had to restart the client periodically to make sure it would use the right ip address. I've read about someone else having the same problem and he tracked it down to the proxy server being the problem.
I don't know if this is helpfull to you. Are you behind a firewall or a proxy?

Edit: I must type faster next time.

riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

Post by riffst3r » Sun Jun 08, 2003 12:19 pm

miker, yes it has worked on the same computer before, but that was before i reinstalled windows xp. my firewall is inactive, and i don't use/have any proxies.

Wrah, i looked at the thread you linked, and i have a feeling my problem is isolated.
[04:11:03] Couldn't send HTTP request to server (wininet)
none of the other folders had the "wininet" error.

i'm going to try and fiddle around and see what happens. thanks for your input though :)

dukla2000
*Lifetime Patron*
Posts: 1465
Joined: Sun Mar 09, 2003 12:27 pm
Location: Reading.England.EU

Post by dukla2000 » Sun Jun 08, 2003 2:12 pm

riff, off topic but what processor do you have? If you have an Intel with SSE or AMD with 3DNow (or SSE) you want to add -advmethods to your start line as well as -forceasm you have already.

Will tend to get you Gromacs work, which especially on a Pentium/SSE radically increases your points/day.

riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

Post by riffst3r » Sun Jun 08, 2003 2:54 pm

i have a p4 2ghz on a passive heat sink.

i knew that the -advmethods switch specifies only WU's on newer cores to be worked on (gromacs and such) but i didn't know that by using gromacs instead of tinkers, i'd be increasing my points/day. i was under the impression that both outputs equal points/day. owz, i'll be back on gromacs after my current WU.

thanks for the heads up :)

dukla2000
*Lifetime Patron*
Posts: 1465
Joined: Sun Mar 09, 2003 12:27 pm
Location: Reading.England.EU

Post by dukla2000 » Sun Jun 08, 2003 3:00 pm

riff wrote:i have a p4 2ghz on a passive heat sink.
Checking your stats seems you were averaging about 35 points a day (until this WU got stuck). My P3 1.2ghz averages about 50 points a day on Gromacs.

Wrah
Patron of SPCR
Posts: 316
Joined: Thu Apr 10, 2003 1:56 am

Post by Wrah » Sun Jun 08, 2003 3:07 pm

riff wrote:i was under the impression that both outputs equal points/day.
Yes, on the 333mhz Celeron stanford uses to benchmark their WU's. :) But that thing doesn't have SSE or 3DNow. And the gromacs core is built to use that while the tinker isn't. So gromacs run faster = more points/day. :)

riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

Post by riffst3r » Sun Jun 08, 2003 6:33 pm

dukla2000, my first few WU's were tinkers, and after that, they were all gromacs. only after i reinstalled windows xp a week ago, i started using tinkers again. owz, i've been going through my stats, and there were a few times my points/day went 60+ which kinda makes me feel better, since your p3 gives you 50 :P

Wrah, go gromacs! /me gets gf to do cheerleading chant

:D

riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

Post by riffst3r » Tue Jun 10, 2003 10:16 pm

well, the results ended up finding their way home, albeit, only after another WU was completed:

Code: Select all

[03:31:56] + Attempting to send results
[03:31:56] - Reading file work/wuresults_03.dat from core
[03:31:56]   (Read 835730 bytes from disk)
[03:33:47] + Results successfully sent
[03:33:47] Thank you for your contribution to Folding@home.
[03:33:47] + Starting local stats count at 1
[03:33:51] Trying to send all finished work units


[03:33:51] + Attempting to send results
[03:33:51] - Reading file work/wuresults_01.dat from core
[03:33:51]   (Read 609824 bytes from disk)
[03:35:13] + Results successfully sent
[03:35:13] Thank you for your contribution to Folding@home.
[03:35:13] + Number of Units Completed: 2
[03:35:13] + Sent 1 completed units to the server
Wrah, i've got a question for you, do these arguments seem alright:
Arguments: -forceasm -advmethods -verbosity 9 -local -service
because even with these, it takes longer than before to complete a frame for the same protien (p347). usually it was 10mins a frame for me, now its almost double. how long does a frame take on your machine?

Wrah
Patron of SPCR
Posts: 316
Joined: Thu Apr 10, 2003 1:56 am

Post by Wrah » Tue Jun 10, 2003 11:43 pm

I did a 347 yesterday and it took about 17 minutes per frame. That's on a tualatin-celly on 1.5 Ghz, with SSE support. The Tbird 1.4 (now running at 1.53Ghz) did a 340 and it needed 15 mins per frame, with 3DNow.
You haven't been folding for a while, have you? These new projects are all 53-pointers so they will take longer then the 542's from before.

I don't see anything wrong with the arguments that might influence performance. Though the -local doesn't have much use I think. Can't hurt either. But don't remove it until it has finished the WU, otherwise it will download a new one without finishing the other one. And with verbosity 9 you'll get less messages in your log, if that's your intention. And I take it you're running it as a service with firedaemon?

riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

Post by riffst3r » Wed Jun 11, 2003 10:33 am

EMIII reported me having 23:58 per frame on that WU. that was pretty disappointing, but last night i stopped firedaemon for a while (so i could do a quick benchmark) and after i restarted it, i got a SSE Boost OK in the log files, and now the logfile reports about 10mins a fram on the p544 i'm doing now. :)

i haven't been folding continuously for couple weeks now, but now i'm back 24/7 and noticing a lot of new things i've missed.

As for the arguments, the -local i because i installed the graphical version to see if it can send my results (there was a moment when i thought it a problem of the client) but it didn't, and I didn't get a chance to flush the registry yet. I thought verbosity 9 increases the messages?

yes, i'm running as a service now, earlier it was just a program in the taskbar getting in the way (and me accidently closing it). now, its a service, its a lot less bother.

Wrah
Patron of SPCR
Posts: 316
Joined: Thu Apr 10, 2003 1:56 am

Post by Wrah » Wed Jun 11, 2003 1:35 pm

That sounds familiar. There's an issue with the 3.24 client that sometimes it only uses SSE with the first WU but then 'forgets' about it with the rest. They fixed that in the 3.25 beta version which you can download here. I'm using this version too.

And you are right about verbosity.

riffst3r
Posts: 221
Joined: Tue Apr 08, 2003 8:43 am
Contact:

Post by riffst3r » Wed Jun 11, 2003 2:51 pm

thanks for the info about the version update, :)

Post Reply