Today I had a power failure at home. With 4 PCs running either CPDN or Sulphur Alpha, only one had a problem upon reboot.

The problem, which has been posted about a number of times on the forum here, is that the PC with the problem went back to timestep 0, <b>of the same Work Unit</b>. It did not download a new model, simply forgot where it was and started over on the same one. This was after 300 some odd hours of processing, with only about 90 to go. Before starting the model back up, I ran a scandisk and it found no errors.

I did not have a recent backup of the boinc folder.

Is there anyway to ensure this doesn't happen if I were to startup a model after a power failure such as this, or some other uncontrolled shutdown/reboot? That is, is there any xml file I can change, or copy which would ensure that the model is starting at a decent point that was not corrupted on shutdown? Thanks.

George

I've had a number of stabs at getting to the bottom of this one without success, George.

I've run tests to eliminate a few of the symptoms that appear to go along with the problem and now suspect that the culprit may be a corrupt file in the dataout directory.

The one thing I haven't tried yet is force crashing of one of my systems, so I'll do a backup and try that when I've got a spare couple of hours ;)

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

&gt; The one thing I haven't tried yet is force crashing of one of my systems, so
&gt; I'll do a backup and try that when I've got a spare couple of hours ;)
&gt;
Good luck with the forced crashing, since rewind to 0 for a given WU is probably only one of the possible bad results. You are dedicated Thyme...and we appreciate it.