Project: 10505 (Run 38, Clone 91, Gen 2)-second time?

[Mactin]

I just started folding p10505-r38-c91-g2.
The problem is that I allready folded this WU on september 7 !

Cuurent log :

Code: Select all

[18:35:21] Working on queue slot 08 [October 17 18:35:21 UTC]
[18:35:21] + Working ...
[18:35:21] - Calling '.\FahCore_11.exe -dir work/ -suffix 08 -nice 19 -priority 96 -checkpoint 10 -forceasm -verbose -lifeline 6940 -version 630'

[18:35:21] 
[18:35:21] *------------------------------*
[18:35:21] Folding@Home GPU Core
[18:35:21] Version 1.31 (Tue Sep 15 10:57:42 PDT 2009)
[18:35:21] 
[18:35:21] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[18:35:21] Build host: amoeba
[18:35:21] Board Type: Nvidia
[18:35:21] Core      : 
[18:35:21] Preparing to commence simulation
[18:35:21] - Assembly optimizations manually forced on.
[18:35:21] - Not checking prior termination.
[18:35:21] - Expanded 17828 -> 125087 (decompressed 701.6 percent)
[18:35:21] Called DecompressByteArray: compressed_data_size=17828 data_size=125087, decompressed_data_size=125087 diff=0
[18:35:21] - Digital signature verified
[18:35:21] 
[18:35:21] Project: 10505 (Run 38, Clone 91, Gen 2)
[18:35:21] 
[18:35:21] Assembly optimizations on if available.
[18:35:21] Entering M.D.
[18:35:28] Tpr hash work/wudata_08.tpr:  1272432890 408746608 3317894883 3601823979 1651195687
[18:35:28] 
[18:35:28] Calling fah_main args: 14 usage=100
[18:35:28] 
[18:35:29] Working on Protein
[18:35:29] Client config found, loading data.
[18:35:29] Starting GUI Server
[18:36:17] Completed 1%
[18:37:13] Completed 2%
[18:37:59] Completed 3%
[18:38:52] Completed 4%
[18:39:39] Completed 5%
[18:40:26] Completed 6%
[18:41:14] Completed 7%
[18:42:01] Completed 8%

September 7 log :

Code: Select all

[01:13:55] Working on queue slot 04 [September 7 01:13:55 UTC]
[01:13:55] + Working ...
[01:13:55] - Calling '.\FahCore_11.exe -dir work/ -suffix 04 -nice 19 -priority 96 -checkpoint 10 -forceasm -verbose -lifeline 5860 -version 630'

[01:13:56] 
[01:13:56] *------------------------------*
[01:13:56] Folding@Home GPU Core
[01:13:56] Version 1.31 (Tue Sep 15 10:57:42 PDT 2009)
[01:13:56] 
[01:13:56] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[01:13:56] Build host: amoeba
[01:13:56] Board Type: Nvidia
[01:13:56] Core      : 
[01:13:56] Preparing to commence simulation
[01:13:56] - Assembly optimizations manually forced on.
[01:13:56] - Not checking prior termination.
[01:13:56] - Expanded 17828 -> 125087 (decompressed 701.6 percent)
[01:13:56] Called DecompressByteArray: compressed_data_size=17828 data_size=125087, decompressed_data_size=125087 diff=0
[01:13:56] - Digital signature verified
[01:13:56] 
[01:13:56] Project: 10505 (Run 38, Clone 91, Gen 2)
[01:13:56] 
[01:13:56] Assembly optimizations on if available.
[01:13:56] Entering M.D.
[01:14:03] Tpr hash work/wudata_04.tpr:  1272432890 408746608 3317894883 3601823979 1651195687
[01:14:03] 
[01:14:03] Calling fah_main args: 14 usage=100
[01:14:03] 
[01:14:03] Working on Protein
[01:14:03] Client config found, loading data.
[01:14:03] Starting GUI Server
[01:14:53] Completed 1%
[01:15:39] Completed 2%
...
[02:30:45] Completed 99%
[02:31:32] Completed 100%
[02:31:32] Successful run
[02:31:32] DynamicWrapper: Finished Work Unit: sleep=10000
[02:31:41] Reserved 55196 bytes for xtc file; Cosm status=0
[02:31:41] Allocated 55196 bytes for xtc file
[02:31:41] - Reading up to 55196 from "work/wudata_04.xtc": Read 55196
[02:31:41] Read 55196 bytes from xtc file; available packet space=786375268
[02:31:41] xtc file hash check passed.
[02:31:41] Reserved 5328 5328 786375268 bytes for arc file=<work/wudata_04.trr> Cosm status=0
[02:31:41] Allocated 5328 bytes for arc file
[02:31:41] - Reading up to 5328 from "work/wudata_04.trr": Read 5328
[02:31:41] Read 5328 bytes from arc file; available packet space=786369940
[02:31:41] trr file hash check passed.
[02:31:41] Allocated 564 bytes for edr file
[02:31:41] Read bedfile
[02:31:41] edr file hash check passed.
[02:31:41] Logfile not read.
[02:31:41] GuardedRun: success in DynamicWrapper
[02:31:41] GuardedRun: done
[02:31:41] Run: GuardedRun completed.
[02:31:44] + Opened results file
[02:31:44] - Writing 61600 bytes of core data to disk...
[02:31:44] Done: 61088 -> 57666 (compressed to 94.3 percent)
[02:31:44]   ... Done.
[02:31:44] DeleteFrameFiles: successfully deleted file=work/wudata_04.ckp
[02:31:44] Shutting down core 
[02:31:44] 
[02:31:44] Folding@home Core Shutdown: FINISHED_UNIT
[02:31:48] CoreStatus = 64 (100)
[02:31:48] Unit 4 finished with 99 percent of time to deadline remaining.
[02:31:48] Updated performance fraction: 0.986750
[02:31:48] Sending work to server
[02:31:48] Project: 10505 (Run 38, Clone 91, Gen 2)
[02:31:48] - Read packet limit of 540015616... Set to 524286976.


[02:31:48] + Attempting to send results [September 7 02:31:48 UTC]
[02:31:48] - Reading file work/wuresults_04.dat from core
[02:31:48]   (Read 58178 bytes from disk)
[02:31:48] Gpu type=2 species=30.
[02:31:48] Connecting to http://171.67.108.21:8080/
[02:31:49] Posted data.
[02:31:49] Initial: 0000; - Uploaded at ~57 kB/s
[02:31:49] - Averaged speed for that direction ~66 kB/s
[02:31:49] + Results successfully sent

[gwildperson]

Perhaps this is a test WU designed to establish reliability statistics for the distributed supercomputer that we're all a part of. If so, there'd be no reason to require PRCG to change, just reassign the same WU unil a certain level of confidence is reached. There was another project like that a year or two ago.

That's just a guess. It could be a server error instead.

As long as you're sure you didn't configure two clients with the same machine-id and the same user-id, you should get credit for both of them. (But you knew that already.)

[sortofageek]

Thank you for your report. I can't tell you anything about your stats for this WU because I don't know your folding name. I can, however, tell you this particular WU has been successfully completed by a multitude of people and many of them have received full credit for it numerous times.

I also can't tell you why this is happening. If you give us your folding name, we can look it up based on your stats.

[Mactin]

Thank you for your report. I can't tell you anything about your stats for this WU because I don't know your folding name. I can, however, tell you this particular WU has been successfully completed by a multitude of people and many of them have received full credit for it numerous times.
I also can't tell you why this is happening. If you give us your folding name, we can look it up based on your stats

In this case : Martin_i7-875k_GTS250 (both times).

I realy like these p10505 they have the highest PPD on my GTS250.
But, is it not wastefull to fold again and again the same WU ?

[sortofageek]

Hi Martin_i7-875k_GTS250 (team 96377),
Your WU (P10505 R38 C91 G2) was added to the stats database on 2010-09-06 20:06:57 for 451 points of credit.

You got the credit before and it is most likely you'll get it again. Happy folding.

When a project manager allows a WU to be folded repeatedly, I suspect they have their reasons for the good of the program. No worries here.

[Mactin]

Thanks, but my apetite is still intact. Points is not the issue.
I'm wondering why redoing the same calculation over again is usefull.
FYI, the same is happening again with p10505-r47-c92-r2, that I am folding again after I folded it on Sept. 1st.

[sortofageek]

I don't have a specific answer for you. The possibility I can think of has already been mention in this topic, that of testing on as many types of hardware as possible. I suppose there might be times where they want it done on more than one client for verification of the results as well.

[gwildperson]

For what I'll call "normal" FAH projects, repeating the same calculation is wastful, but unless there's actually an error in your steup, it's not going to happen. If these happen to be reliablity tests as a surmised earlier, repeating the calculations wouldn't be wastful.

In any case, Stanford takes responsiblity for the assignment process and there is almost nothing you can do to change it. You've done the most appropriate thing you can do when you see something strange, namely report it. The message will get through to the Pande Group and they'll either fix something or they'll already know it was intentional due to some specific reason that we don't happen to know.

[Mactin]

It would be nice of PG to give official feedback on this, I have now re-folded twelve p10505 WUs in a row.
If this is intentional, please say so, I will stop worying.
If it is a malfunction, please say so, I will turn off my client and save electricity, I promise I will turn it back on when the problem is resolved.
Thank you for your speculation, assumpsions and experience, but ultimately only the real and true answer will help.