Fedora 11 not finishing WUs

[gizmo]

Interesting problem. System's running, folding away. End of day comes, shut down my laptop and take it home:

Code: Select all

[13:03:14] Completed 100000 out of 250000 steps  (40%)
[13:19:45] Completed 102500 out of 250000 steps  (41%)
[13:36:42] Completed 105000 out of 250000 steps  (42%)
[13:38:48] ***** Got a SIGTERM signal (15)
[13:38:48] Killing all core threads

Folding@Home Client Shutdown.


All well and good. Get the machine home and start it up:

Code: Select all



--- Opening Log file [September 29 14:24:38 UTC]


# Linux SMP Console Edition ###################################################
###############################################################################

                       Folding@Home Client Version 6.24beta

                          http://folding.stanford.edu

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

Launch directory: /home/fah
Executable: /home/fah/fahclient
Arguments: -smp -verbosity 9

[14:24:38] - Ask before connecting: No
[14:24:38] - User name: ChrisRichards (Team 45)
[14:24:38] - User ID: 3D354B60823A780
[14:24:38] - Machine ID: 1
[14:24:38]
[14:24:38] Loaded queue successfully.
[14:24:38] - Autosending finished units... [September 29 14:24:38 UTC]
[14:24:38] Trying to send all finished work units
[14:24:38] + No unsent completed units remaining.
[14:24:38] - Autosend completed
[14:24:38] - Preparing to get new work unit...
[14:24:38] + Attempting to get work packet
[14:24:38] - Will indicate memory of 512 MB
[14:24:38] - Connecting to assignment server
[14:24:38] Connecting to http://assign.stanford.edu:8080/


Eh, what? I've already got a WU that's not complete. Why am I trying to get a new WU? What happened to the old one?

[gizmo]

Bump

I've backed down to the 6.02 client from the 6.24 beta client, and still, every time I reboot this machine, I'm downloading a new WU and starting over. Any ideas. Any thoughts on what to check?

[bruce]

You do NOT have a WU that's not complete, but I don't know what happened to it. It's trying to get a new WU because no WU is in queue.

Try this.
1) Run a WU for a while.
2) Shut down the client but do not reboot.
3) Restart the client? Does the WU resume?

FAH creates data in the current directory including queue.dat and a subdirectory called "work" and if that data is disturbed by something between runs, FAH will not restart. Does FAH have full read/write permissions to those locations?

[gizmo]

Sorry for the delayed reply; I didn't realize anyone had responded.

I really don't know what happened, but I redid my daemon script according to the instructions for the Linux client, cleared out everything for my fah user, redownloaded the client, and it all works, and has been for a couple of weeks.