Have a UPS and had a fraction of second power drop on Sunday causing BOINC to go in pause and not resume. It was 5 seconds into a HDC. WXP btw. Closed BOINC completely, restarted upon which it resumed at 0 seconds. I have pre-empting on.

As to the 'Too Late', u were certainly still within the grace period of force majeure and should have gotten credit. If the recipient of the 4th copy reads his results status file, he will realize that it has become superfluous and cancel.

I'll have a look at BOINC dev forum if anyone else has observed this.

cheers

Ah, any throttle features on your MiniMac? I use Threadmaster, which i know if it gets interrupted, will hang BOINC.

Also did the crunching actually stop or had BOINCmgr lost contact with the Core client?

Could not locate anything mentioning for a process to just stop completely in its track....if the BOINC agent shows normal responsiveness, it might be a thing to report, with device ID, WU number BOINC version and message log file.

Some general and scheduling preferences can be set in the device profile in 'My Grid' http://www.worldcommunitygrid.org/ms/device/viewProfiles.do. Once done, hit the Update button in the projects tab of BOINC.

I'm a bit at odds with what you mean with preempting...language barrier. It's referred to if switching between projects after an fixed time frame of e.g. 4 hours and if BOINC is instructed to keep the interrupted job in memory when paused (see prefs u can set in previous para). Your first post suggests it was interrupted, your second post implies its running without your intervention?

BOINC has some mysterious ways......it froze again hanging on my firewall today requiring for latter to be closed temporarily.....no messages at all.....shortly i'll be away for an extended period.... little trust it will be running when i come back.

cheers

cheers