How did this HAPPEN?! (100k WU!)

[Eugenitor]

So a friend of mine starts folding and out of the clear blue sky he gets two 100k-point workunits. This is his history from Hardfolding's stats:

10a,04 0 0
7a,04 494 1
4a,04 100,611 2
1a,04 100,000 2
10p,03 587 1
7p,03 494 1
4p,03 494 1
1p,03 494 1

Name: themasterban
Team: 32461

Notes:

- He's running an Nvidia GPU and usually gets 494-point WUs. He's not running SMP.
- This happened overnight, absolutely nothing changed on his end
- Please don't subtract his points if you find/fix the bug

[7im]

It's actually 2 WUs for 100k (50k x 2), during 2 different periods (total of 4 WUs at 50k+) but still very strange.

And with only 1 client active in the last 7 days as shown on the Stanford stats, it can't be someone else using the same name by accident.

[Eugenitor]

I'm willing to bet that it's not 50k x 2, that it's more like his regular unit of 494 plus another 99506 for the first one and 100117 for the second (or thereabouts, depending on what his real workunit was). It feels like these things were just tacked-on somehow, but we're both clueless as to how.

[7im]

That's possible. Have themasterban check his fahlog.txt file to see if the client returned one or two WUs during those time frames, and also collect the PRCG numbers and post them here. That way a forum Mod might be able to help sort this out.

[artoar_11]

Yesterday also saw a similar comment in another forum for such a case. 100k for a WU. As I see this is a new member of F @ H. Really weird.

http://kakaostats.com/usum.php?u=1705997

[bruce]

Yesterday also saw a similar comment in another forum for such a case. 100k for a WU. As I see this is a new member of F @ H. Really weird.

http://kakaostats.com/usum.php?u=1705997

Major_Ricardo has a single NV client, though I can't tell if which model.

[Tobit]

Reading a few various team forums, there is definitely something sketchy going on as this has happened to a quite a few people.

[k1wi]

Signup bonus? lol :/

[VijayPande]

Thanks for reporting this. We are looking into it.

[rcranfield]

I am in a similar boat. I am running 2 smp with -advmethods, one uniprocessor, one GPU3 and 4 GPU2 clients. I average ~55000 PPD. Foraging for the culprit client and knowing I have given useable info might be hard.

[zodac]

Similar situation here. This member was using an E6400 and an 8800GT:

[bruce]

Can you tell us what project caused it? I don't know if this will help the Pande Group with the problem, but it would be interesting to us here on the forum.

[Xilikon]

I have a report of that 100k update in team 33 as well. Dunno exactly what project is the cause.

[zodac]

Can you tell us what project caused it? I don't know if this will help the Pande Group with the problem, but it would be interesting to us here on the forum.

Here's the Fahlog for the user who I posted above:

Code: Select all

[14:52:33] Loaded queue successfully.
[14:52:33] 
[14:52:33] + Processing work unit
[14:52:33] Core required: FahCore_a3.exe
[14:52:33] Core found.
[14:52:33] Working on queue slot 03 [October 4 14:52:33 UTC]
[14:52:33] + Working ...
[14:52:34] 
[14:52:34] *------------------------------*
[14:52:34] Folding@Home Gromacs SMP Core
[14:52:34] Version 2.22 (Mar 12, 2010)
[14:52:34] 
[14:52:34] Preparing to commence simulation
[14:52:34] - Looking at optimizations...
[14:52:34] - Files status OK
[14:52:35] - Expanded 1766113 -> 2254429 (decompressed 127.6 percent)
[14:52:35] Called DecompressByteArray: compressed_data_size=1766113 data_size=2254429, decompressed_data_size=2254429 diff=0
[14:52:35] - Digital signature verified
[14:52:35] 
[14:52:35] Project: 6062 (Run 0, Clone 90, Gen 89)
[14:52:35] 
[14:52:35] Assembly optimizations on if available.
[14:52:35] Entering M.D.
[14:52:41] Using Gromacs checkpoints
[14:52:43] Resuming from checkpoint
[14:52:44] Verified work/wudata_03.log
[14:52:44] Verified work/wudata_03.trr
[14:52:44] Verified work/wudata_03.edr
[14:52:44] Completed 381516 out of 500000 steps  (76%)
[15:04:59] Completed 385000 out of 500000 steps  (77%)
[18:05:25] Received EVENT= from OS -- shutting down w/ INTERRUPTED signal; events=[CtrlC= Close= Break= Shutdown= LogOff=]
[18:05:25] Folding@home Core Shutdown: INTERRUPTED