Page 1 of 2
Project 3798
Posted: Tue Nov 25, 2008 8:01 pm
by anko1
When I first saw this in my log, I thought I had an error for sure. Then when I saw the finished unit, I thought this had to be the world's smallest WU. When I looked it up, I saw that they are testing code.
Code: Select all
[08:21:24] + Closed connections
[08:21:24]
[08:21:24] + Processing work unit
[08:21:24] Core required: FahCore_78.exe
[08:21:24] Core found.
[08:21:24] Working on Unit 06 [November 25 08:21:24]
[08:21:24] + Working ...
[08:21:24] - Calling 'FahCore_78.exe -dir work/ -suffix 06 -checkpoint 15 -verbose -lifeline 880 -version 504'
[08:21:24]
[08:21:24] *------------------------------*
[08:21:24] Folding@Home Gromacs Core
[08:21:24] Version 1.90 (March 8, 2006)
[08:21:24]
[08:21:24] Preparing to commence simulation
[08:21:24] - Looking at optimizations...
[08:21:24] - Created dyn
[08:21:24] - Files status OK
[08:21:25] - Expanded 237785 -> 1167708 (decompressed 491.0 percent)
[08:21:25] - Starting from initial work packet
[08:21:25]
[08:21:25] Project: 3798 (Run 30, Clone 4, Gen 1)
[08:21:25]
[08:21:25] Assembly optimizations on if available.
[08:21:25] Entering M.D.
[08:21:31] Protein: p3798
[08:21:31]
[08:21:31] Writing local files
[08:23:16] Extra SSE boost OK.
[08:23:16] Writing local files
[08:23:16] Completed 0 out of 1500 steps (0)
[08:24:02] Writing local files
[08:24:03] Completed 500 out of 1500 steps (33)
[08:24:51] Writing local files
[08:24:51] Completed 1000 out of 1500 steps (67)
[08:25:40] Writing local files
[08:25:40] Completed 1500 out of 1500 steps (100)
[08:25:40] Writing final coordinates.
[08:25:40] Past main M.D. loop
[08:26:40]
[08:26:40] Finished Work Unit:
[08:26:40] - Reading up to 550656 from "work/wudata_06.arc": Read 550656
[08:26:40] - Reading up to 0 from "work/wudata_06.xtc": Read 0
[08:26:40] goefile size: 0
[08:26:40] Leaving Run
[08:26:41] - Writing 573212 bytes of core data to disk...
[08:26:41] Done: 572700 -> 519640 (compressed to 90.7 percent)
[08:26:41] ... Done.
[08:26:41] - Shutting down core
[08:26:41]
[08:26:41] Folding@home Core Shutdown: FINISHED_UNIT
Re: Project 3798
Posted: Tue Nov 25, 2008 8:29 pm
by toTOW
Wow it's been completed 3 times
Re: Project 3798
Posted: Tue Nov 25, 2008 9:09 pm
by anko1
That's interesting. I've had three separate WUs so far:
Project: 3798 (Run 6, Clone 0, Gen 28)
Project: 3798 (Run 30, Clone 4, Gen 1)
Project: 3798 (Run 11, Clone 0, Gen 1)
Btw, are they giving points for this? It didn't say in the project description.
Re: Project 3798
Posted: Tue Nov 25, 2008 9:12 pm
by toTOW
No record under your name for Project: 3798 (Run 6, Clone 0, Gen 28), but two people completed it successfully.
Hi anko1 (team 47815),
Your WU (P3798 R30 C4 G1) was added to the stats database on 2008-11-25 02:21:08 for 5 points of credit.
There also two other full credits for this one.
Project: 3798 (Run 11, Clone 0, Gen 1) : two people got full credit for it (but not you).
Yes they're worth 5 points ... and I don't think you should get them ...
Re: Project 3798
Posted: Wed Nov 26, 2008 12:49 am
by anko1
"Yes they're worth 5 points ... and I don't think you should get them ..." At a point per minute I don't mind.
Thanks for checking.
Edit: thanks anandhanju for explaining toTOW's comment. Edit to reflect that.
Re: Project 3798
Posted: Wed Nov 26, 2008 10:09 pm
by Snyde
Having an issue uploading this one as well. From the time of download to the first upload attempt, about 17 minutes. Took about 6 minuted from writing local files to past main MD loop, completed with 100% time remaining.
Code: Select all
Project: 3798 (Run 30, Clone 0, Gen 6)
16:34:05] + Attempting to send results
[16:34:05] - Reading file work/wuresults_09.dat from core
[16:34:05] (Read 520061 bytes from disk)
[16:34:05] Connecting to http://171.64.122.139:8080/
[16:35:44] Writing local files
[16:35:44] Completed 325000 out of 2500000 steps (13)
[16:41:01] - Couldn't send HTTP request to server
[16:41:01] + Could not connect to Work Server (results)
[16:41:01] (171.64.122.139:8080)
[16:41:01] - Error: Could not transmit unit 09 (completed November 25) to work server.
[16:41:01] - 11 failed uploads of this unit.
[16:41:01] + Attempting to send results
[16:41:01] - Reading file work/wuresults_09.dat from core
[16:41:01] (Read 520061 bytes from disk)
[16:41:01] Connecting to http://171.64.122.135:8080/
[16:41:05] Posted data.
[16:41:05] Initial: 0000; - Uploaded at ~127 kB/s
[16:41:05] - Averaged speed for that direction ~131 kB/s
[16:41:05] - Server does not have record of this unit. Will try again later.
[16:41:05] Could not transmit unit 09 to Collection server; keeping in queue.
[16:41:05] + Sent 0 of 1 completed units to the server
Don't care about the points, just wondering how much longer this will try to upload. What is the expiry date and when will it auto delete? Or should I not sweat this?
Regards-
jb
Re: Project 3798
Posted: Wed Nov 26, 2008 10:14 pm
by anandhanju
I think the unit has an expiry period of 2 days, and as toTOW said, they shouldn't have been assigned outside of testing.
I had trouble uploading them as well, but they just went through about half an hour ago. Try stopping the client and restarting it or use the -send all switch.
Re: Project 3798
Posted: Wed Nov 26, 2008 10:16 pm
by Snyde
anandhanju wrote:I think the unit has an expiry period of 2 days, and as toTOW said, they shouldn't have been assigned outside of testing.
I had trouble uploading them as well, but they just went through about half an hour ago. Try stopping the client and restarting it or use the -send all switch.
Will give it a go, thanks-
Re: Project 3798
Posted: Thu Nov 27, 2008 3:54 am
by Snyde
Snyde wrote:anandhanju wrote:I think the unit has an expiry period of 2 days, and as toTOW said, they shouldn't have been assigned outside of testing.
I had trouble uploading them as well, but they just went through about half an hour ago. Try stopping the client and restarting it or use the -send all switch.
Will give it a go, thanks-
No go. Will it auto delete after expiration?
Regards-
Re: Project 3798
Posted: Thu Nov 27, 2008 6:30 am
by anko1
No go. Will it auto delete after expiration?
If it gets close to the deadline, I'd delete the unit. If autosend tries to send a past due unit, it kills the unit that's being worked on. At least that's been my experience.
http://foldingforum.org/viewtopic.php?f ... =15#p66245
Re: Project 3798
Posted: Fri Apr 17, 2009 8:53 pm
by GrGramps
Looks like project 3798 is back again. Downloaded 48 min ago; preferred deadline 34 min ago. I suppose I'm too slow?
Re: Project 3798
Posted: Fri Apr 17, 2009 9:15 pm
by John Naylor
lol... it reports that deadline but folds on straight through it... I dunno why.
Re: Project 3798
Posted: Sat Apr 18, 2009 1:13 am
by susato
I'll report it but since it's the weekend, please have extra patience. And it may be doing precisely what it's supposed to do.
IMHO this project is server test code again. Did you see there's no atom count for it in its psummary listing? The 0.01 day preferred deadline would make sure that it gets issued again every few minutes, slowly building up in number and in its ability to load up the server, while the 2 day final deadline gives everyone who gets it a chance to earn points on it.
Rather ingenious if you ask me!
Re: Project 3798
Posted: Sat Apr 18, 2009 3:01 am
by Grandpa_01
I think they were trying to load the servers today I started getting the about 3 AM this morning and kept getting them until about 1 hr ago 7 PM. I had around 30 of them today, made for allot of points and probably did a pretty good job of loading up the servers. The longest I had to wait for another WU was 7 attempts most connected and downloaded on the first attempt.
Re: Project 3798
Posted: Sat Apr 18, 2009 7:20 am
by jcoffland
susato wrote:IMHO this project is server test code again. Did you see there's no atom count for it in its psummary listing? The 0.01 day preferred deadline would make sure that it gets issued again every few minutes, slowly building up in number and in its ability to load up the server, while the 2 day final deadline gives everyone who gets it a chance to earn points on it.
That's correct. I am going to stop new assignments of project 3798 for now though to avoid wasting CPU cycles.