Folding Forum

I'm using System tray client on Windows 7 32 bit. I use 2 clients installed using official guide for multi CPU system.

I got this error when client tried to send finished work:

[10:55:12] Completed 1500000 out of 1500000 steps (100%)
[10:55:13] DynamicWrapper: Finished Work Unit: sleep=10000
[10:55:23]
[10:55:23] Finished Work Unit:
[10:55:23] - Reading up to 2908800 from "work/wudata_01.trr": Read 2908800
[10:55:23] trr file hash check passed.
[10:55:23] - Reading up to 1554364 from "work/wudata_01.xtc": Read 1554364
[10:55:23] xtc file hash check passed.
[10:55:23] edr file hash check passed.
[10:55:23] logfile size: 75593
[10:55:23] Leaving Run
[10:55:28] - Writing 4543769 bytes of core data to disk...
[10:55:30] Done: 4543257 -> 4325284 (compressed to 95.2 percent)
[10:55:30] ... Done.
[10:55:31] - Shutting down core
[10:55:31]
[10:55:31] Folding@home Core Shutdown: FINISHED_UNIT
[10:55:35] CoreStatus = 64 (100)
[10:55:35] Sending work to server
[10:55:35] Project: 7809 (Run 3, Clone 469, Gen 2)


[10:55:35] + Attempting to send results [September 29 10:55:35 UTC]
[10:56:46] - Server reports problem with unit.
[10:56:46] - Preparing to get new work unit...
[10:56:46] + Attempting to get work packet

I've been working on this unit for nine days, so any help is appreciated. My username is Tomazyk.

Welcome to Folding@Home Support Forum, tomazyk. Our apologies for the delay in post visibility. This happens for new members and doesn't mean there is a problem with your post(s).

If you skipped the step about deleting the work unit out of the copied 2nd client, that work unit won't get credit as the first client probably did already.

Sometimes there is data corruption in the work unit, either caused by a transmission error while uploading, or more often, the local computer is overclocked just a bit too much, and the WU was corrupted that way.

Another possible cause is the work unit was simply malformed to begin with. It's less common, but a forum Mod could check to see if anyone has finished that specific work unit, or if others have errored out as well.

Thank you both. I checked connection to server and it is OK.
I also used -verbosity 9 switch wich showed: [18:18:23] + No unsent completed units remaining. That means my work is gone . (I still have wuresults_01.dat file in my work directory, though)

I also checked the whole log and found probable cause of error. When I created second work directory I removed all work data. Then I ran second client to change settings and client got this work unit. It was assigned to MachinedID1. After client reset the new configuration was in use - MachineID2. Server probably did not accept my work beacuse the work was assigned to MachineID 1 and resaults came from MachineID2 I guess I should delete work data twice (first after copying folder and then again after configuring client).

I can send file wuresults_01.dat if it is useful to you.

No need to send the file.

Keep an eye on the clients. Make sure the next work units upload okay.

So far there is nothing in the database for Project: 7809 (Run 3, Clone 469, Gen 2) for anyone.

sortofageek wrote:So far there is nothing in the database for Project: 7809 (Run 3, Clone 469, Gen 2) for anyone.

There seems to be an occasional WU that looks like this one does (finished ok but the server rejects it). Is it possible that this is an undiagonsed symptom of the so called "bad wu" problem? If so, we might expect the same WU to be assigned and fail repeatedly and never show up in the database. Has that condition ever been observed?

I have observed this, yes. The problem is I can't prove anything when nothing shows up in the database and I only have one folder's log showing the issue. We do mark WUs like this for followup and, if after enough time passes without database results, mark them bad. By that time, however, I suspect it has already been caught and stopped by a project manager.

Here is a situation with that error message, however, which definitely wasn't a bad WU as another folder completed the WU in question successfully. See this topic ---> viewtopic.php?f=19&t=19695&p=196176#p196176