Project: 2620 (Run 64, Clone 11, Gen 36) 0x79 error

[synaptix]

Hello,

It seems that I am having quite a bit of trouble with this project. After running for quite a while, I get the following error in the log:

Code: Select all

[18:34:37] Loaded queue successfully.
[18:34:37] Initialization complete
[18:34:37] 
[18:34:37] + Processing work unit
[18:34:37] Core required: FahCore_78.exe
[18:34:37] Core found.
[18:34:37] Working on queue slot 05 [November 1 18:34:37 UTC]
[18:34:37] + Working ...
[18:34:38] 
[18:34:38] *------------------------------*
[18:34:38] Folding@Home Gromacs Core
[18:34:38] Version 1.90 (March 8, 2006)
[18:34:38] 
[18:34:38] Preparing to commence simulation
[18:34:38] - Ensuring status. Please wait.
[18:34:55] - Looking at optimizations...
[18:34:55] - Working with standard loops on this execution.
[18:34:55] - Created dyn
[18:34:55] - Files status OK
[18:34:57] - Expanded 3301509 -> 6015025 (decompressed 182.1 percent)
[18:34:58] - Starting from initial work packet
[18:34:58] 
[18:34:58] Project: 2620 (Run 64, Clone 11, Gen 36)
[18:34:58] 
[18:34:59] Entering M.D.
[18:35:05] Gromacs error.
[18:35:05] 
[18:35:05] Folding@home Core Shutdown: UNKNOWN_ERROR
[18:35:08] CoreStatus = 79 (121)
[18:35:08] Client-core communications error: ERROR 0x79
[18:35:08] This is a sign of more serious problems, shutting down.

I have already removed this project, core FahCore_78.exe, and all data files from both the queue and my computer -- twice. Then I thought that perhaps the F@H CPU client had become corrupt (I put enough stress on my machine to cause at least one or two lockups or reboots a week), so I uninstalled, removed everything but the config file, ran a chkdsk, and reinstalled the Win32 6.20 systray client. I still receive the same errors, now as soon as the client has started and engaged the core (< 20 seconds after starting the client). Worse still, the assignment servers seem bent on giving me this specific project no matter how many times I delete it from my queue, and I haven't found a way to block it from being downloaded/run. Any ideas?

Many thanks in advance.

[toTOW]

This is a bad WU that someone will have to remove manually.

[synaptix]

So.... Basically any of my machines that are encountering errors running this WU will simply have to wait until this WU is removed from the work assignment servers before they can continue folding? Is there not a way to block specific WU's from being downloaded/run? If not, how long would you estimate that I would have to wait before going through the trouble of removing the WU from the queue and data files from each of my machines?

[goodone121]

Although I am no expert, try sending in the results with qfix. If memory serves me correctly, I was told the servers should give you a new WU. Again, I am no expert and make no promises.

[anandhanju]

Hello synaptics, welcome to the forums! Do you notice your system memory mushrooming in the 5 seconds before it EUEs?

You might want to check this thread out. James encountered the same problem there and the only common thread I see is that both are on p262x WUs.

@Mods, There have been several reports like these on the p262x WUs. See: 1, 2, ... Ah, found this post by toTow that sums it up. Could it be a bad run or clone setting?

[synaptix]

Many thanks for the quick replies and insightful advice.

@goodone121 - I think that qfix only runs on linux machines (the only result I get running this is "unknown version of queue.dat"). Currently my linux box is down and will hopefully miss the chance to see this WU. Everything else is running XP x86-32 or 2003 Standard x86-32. If there is a Windows qfix, I must have missed it.

@anandhanju - My memory doesn't really change that much before a crash -- just the normal 16-32MB typically used by the CPU systray client during startup. Also, thank you for that thread. It turns out that the recommendation of just deleting all F@H data files other than "client.cfg" until the assignment server gave me a different WU did the trick.

I would have thought that the third time would have done the trick, but even after my initial six or so deletions of the work and queue, I still got the same exact WU. Just a couple more deletions, though, and now it seems that I am back up and running. Gotta love the fixes that would never work in any other area of computer troubleshooting...

Anyways, thanks to all that helped, and I will keep this trick in mind should I experience similar issues in the future.