F@H 6.20 kills mac mini, 10.5.4

Locked
ordinary
Posts: 1
Joined: Mon Aug 25, 2008 12:47 pm
Hardware configuration: 1) frederic: gigabyte M57SLI-S4 Rev2.0, Athlon 5600+, 4GB Ram, CentOS 5.2, F@H Linux Console Edition 6.02.
2) rita: HP Pavilion a1350n, Windows XP SP2 (32 bit) AMD Athlon 4200+ 2.19GHz, 2GB RAM F@H CPU Client 6.20
3) opal: Mac Mini 2GHz Core 2 Duo, 2GB RAM, OS X 10.5.4 F@H SMP Console Edition, 6.20
Location: Vermont

F@H 6.20 kills mac mini, 10.5.4

Post by ordinary » Mon Aug 25, 2008 2:50 pm

I have been folding for about a year now, with little problem. Mostly on Fedora, some on XP. I just acquired a Mac Mini, and put it to work, too.

My problem is this: F@H seems to kill the Mac, and I don't know how to troubleshoot it. On a Linux box, I'd at least know how to start looking at the problem, but I'm an OS X tyro.

The machine is a Mac Mini, Core 2 Duo, 2GHz, 2GB RAM, OS X 10.5.4.

The failure mode is that the Mac put up a window saying it must be restarted in about 5 languages. At this point, I can neither ping nor ssh into the mini, so I believe it is well and truly dead. When I restart the mini, it takes forever (running fsck, I think) and when it finally boots, it immediately demands to be restarted again. No ping or ssh response.

I was at a loss as to what was wrong with the mini, but the only suspect thing I had done was install the F@H s/w. I have installed the gimp and a couple of other macport packages, but none are running; then only non Mac s/w running was F@H. So, I booted into safe mode, disabled F@H, rebooted and things worked fine.

I have included a portion of FAHlog.txt here. It is from the time I restarted the machine until the client dies. The log contains error messages, but none at the moment that the machine and client die.

I recognize that my problem really may be unrelated to F@H, and I may be observing a coincidence. However, if the F@H client is enabled my machine won't completely boot; without the F@H client, the mac runs fine.

Lastly, the mac F@H client has been running fairly well, producing results right along, just this new work package seems to be creating problems.

What should I do?

Should I assume the problem isn't F@H, but elsewhere? Probably not, huh?
Should I report the issue to F@H in case it is a problem with the client?
How can I get back to folding productively?

Other thoughts are welcome.

Thanks,
Phil

Code: Select all

--- Opening Log file [August 24 03:19:45 UTC] 


# Mac OS X SMP Console Edition ################################################
###############################################################################

                       Folding@Home Client Version 6.20

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /Users/phil/Library/Folding@home
Executable: /Applications/Folding@home.app/fah6


[03:19:45] - Ask before connecting: No
[03:19:45] - User name: ordinary (Team 37988)
[03:19:45] - User ID: xxxxxxxxxxxxxxxx
[03:19:45] - Machine ID: 1
[03:19:45]
[03:19:45] Loaded queue successfully.
[03:19:45]
[03:19:45] + Processing work unit
[03:19:45] Core required: FahCore_a1.exe
[03:19:45] Core found.
[03:19:46] Working on queue slot 04 [August 24 03:19:46 UTC]
[03:19:46] + Working ...
[03:19:47] ---*
[03:19:47] Folding@Home Gromacs SMP Core
[03:19:47] Version 1.74 (September 24, 2007)
[03:19:47]
[03:19:47] Preparing to commence simulation
[03:19:47] - Ensuring status. Please wait.
[03:20:04] - Looking at optimizations...
[03:20:04] - Working with standard loops on this execution.
[03:20:04] Examination of work files indicates 8 consecutive improper terminations of core.
[03:20:07] - Expanded 2430169 -> 12908109 (decompressed 531.1 percent)
[03:20:07] - Error: Bad work unit. Digital signatures don't match
[03:20:07] Error: Could not open work file
[03:20:07] Finalizing output
[03:20:08]
[03:20:08] Project: 755630080 (Run 83886080, Clone 1073741824, Gen 1392508928)
[03:20:08]
[03:20:08] Error: Could not write local file.  Exiting.
[03:20:13] - Shutting down core
[03:22:12] Finalizing output
[03:22:17] CoreStatus = 12 (18)
[03:22:17] Client-core communications error: ERROR 0x12
[03:22:17] This is a sign of more serious problems, shutting down.


--- Opening Log file [August 24 03:22:17 UTC]


# Mac OS X SMP Console Edition ################################################
###############################################################################

                       Folding@Home Client Version 6.20

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /Users/phil/Library/Folding@home
Executable: /Applications/Folding@home.app/fah6


[03:22:17] - Ask before connecting: No
[03:22:17] - User name: ordinary (Team 37988)
[03:22:17] - User ID: xxxxxxxxxxxxxxxx
[03:22:17] - Machine ID: 1
[03:22:17]
[03:22:17] Loaded queue successfully.
[03:22:17]
[03:22:17] + Processing work unit
[03:22:17] Core required: FahCore_a1.exe
[03:22:17] Core found.
[03:22:17] Working on queue slot 04 [August 24 03:22:17 UTC]
[03:22:17] + Working ...
[03:22:17]
[03:22:17] *------------------------------*
[03:22:17] Folding@Home Gromacs SMP Core
[03:22:17] Version 1.74 (September 24, 2007)
[03:22:17]
[03:22:17] Preparing to commence simulation
[03:22:17] - Ensuring status. Please wait.
[03:22:34] - Looking at optimizations...
[03:22:34] - Working with standard loops on this execution.
[03:22:34] Examination of work files indicates 8 consecutive improper terminations of core.
[03:22:35] - Expanded 2430169 -> 12908109 (decompressed 531.1 percent)
[03:22:37]
[03:22:37] Project: 2605 (Run 5, Clone 64, Gen 83)
[03:22:37]
[03:22:37] Entering M.D.
[03:22:44] mdrunner cpfilename:
[03:22:44] Calling FAH init
[03:22:45] Read topology
[03:22:45] g local files
[03:22:45]  checkpoint)
[03:22:45] Read checkpoint
[03:22:46] Protein: Protein in POPC
[03:22:46] Writing local files
[03:22:46] Extra SSE boost OK.
[03:22:46] Writing local files
[03:22:47] Completed 0 out of 500000 steps  (0 percent)
[03:23:10]
[03:23:10] Folding@home CoreStatus = 1 (1)
[03:23:14] Client-core communications error: ERROR 0x1
[03:23:14] This is a sign of more serious problems, shutting down.


--- Opening Log file [August 24 03:23:14 UTC]


# Mac OS X SMP Console Edition ################################################
###############################################################################

                       Folding@Home Client Version 6.20

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /Users/phil/Library/Folding@home
Executable: /Applications/Folding@home.app/fah6


[03:23:14] - Ask before connecting: No
[03:23:14] - User name: ordinary (Team 37988)
[03:23:14] - User ID: xxxxxxxxxxxxxxxx
[03:23:14] - Machine ID: 1
[03:23:14]
[03:23:14] Loaded queue successfully.
[03:23:14]
[03:23:14] + Processing work unit
[03:23:14] Core required: FahCore_a1.exe
[03:23:14] Core found.
[03:23:14] Working on queue slot 04 [August 24 03:23:14 UTC]
[03:23:14] + Working ...
[03:23:14]
[03:23:14] *------------------------------*
[03:23:14] Folding@Home Gromacs SMP Core
[03:23:14] Version 1.74 (September 24, 2007)
[03:23:14]
[03:23:14] Preparing to commence simulation
[03:23:14] - Ensuring status. Please wait.
Your milage will NOT vary; it's just an urban legend!

Beberg
Pande Group Member
Posts: 566
Joined: Sat Dec 01, 2007 9:05 pm
Location: Stanford

Re: F@H 6.20 kills mac mini, 10.5.4

Post by Beberg » Mon Aug 25, 2008 3:52 pm

Sounds like the kernel panic screen... which F@h generally wouldn't be causing on it's own.

With a mini, you may have overheating problems, or some other hardware issue. You can try manually starting the client while you have some temperature monitoring (there are many widgets to do that) going. Another common cause is bad RAM, so run a RAM testing utility - I believe there is one with the utility disk that a mac comes with...

Locked

Return to “Mac OS X Beta”