So I decided that I wanted folding to crunch heavily on my wifes idling computer during periods of the day when she isnt using it which is between say 2am and 2pm. I have the folding @ home service set to manual startup. I then added a shortcut to a startup in scheduled tasks to launch F@H and that worked fine. And at 2pm with no better way to shutdown the F@H to shut off for the next 12 hours, I use a Task Kill process in another scheduled task that stops it from running.
This worked ok for about 4 days and then I think I broke F@H, but it doesnt error out. Instead it launches now at 2am and doing so Firefox launches and it then comes up to the Web GUI showing the F@H activity and it says that its waiting and shows a count down of say 2 minutes. When the counter gets to 0, it then shows a count down of say 5 minutes and then when the count down gets to 0, it then keeps saying waiting with a different wait period. During this time the CPU is idle at like 5% in use. F@H no longer crunches proteins.
Also, if there is a more graceful way of stopping F@H at 2pm than a taskkill /T please let me know. I tried targeting the single service that is running and killing it caused it to end and then restart about 30 sec later. So I had to go with the Taskkill /T method to kill it.
*Note I can supply the actual kill command if needed to show which service name is targeted but from memory I cant remember the exact name of it, but can provide that info when I get home from work to copy/paste it here.
So I am wondering if I broke F@H by killing it in the middle of crunching through the Taskkill /T which kills F@H and all related subprocesses of it. I looked all over for a way to make it only run on a schedule and the only thing I can come up with is to start the service and kill the service to control when its run. In doing so its no longer happy and I figured I'd ask here to see if others who have run F@H have seen this before and the fix?
I was going to uninstall it and reinstall the Windows F@H client again and then configure for the group and set it crunching, but maybe this can be cleared up by wiping out a corrupt file or something that is related to the F@H startup vs a reinstall of the client. Maybe it was writing its results at the same time that the service was killed etc, which is a possibility with the write to disk I set to every 5 minutes so that data processed was check pointed more frequently to minimize wasted crunch time when the service was killed to where the maximum crunch time loss was less than 5 minutes, and then the following day at 2am it picks up where it left off.