Corrupt WU ???
17 Aug 2006 14:26:19 UTC
Topic 191693
(moderation:
There was a power failure in our area last night and, when it hit, my computer was in the middle of crunching this WU. When the power came back on and I restarted the computer, BOINC appeared to pick-up where it left off. Now, I am not so sure. That is, progress is now (only) about 50% but it has already crunched longer than the "usual" amount of time needed to complete this type unit. I am wondering if I should abort it or if there might be something else to try. Any suggestions??
Corrupt WU ???
)
When the checkpoint file gets corrupted or unreadable (which might occure after a power failure) the computation starts at the beginning thus leading to the 'normal' time plus the time spent until the power failure occured.
Have a look (or copy it to this thread) at the file 'stderr.txt' in subfolder 'slots'...
Udo
Udo
RE: When the checkpoint
)
Udo,
Thank you! The stderr.txt file shows exactly that:
2006-08-16 18:14:33.7500 [normal]: No usable checkpoint found, starting from beginning. (first time it started the WU); and, again
2006-08-17 05:07:35.7031 [normal]: No usable checkpoint found, starting from beginning. (at start-up after the power failure).
Unfortunately, the WU was nearly finished when the power failure hit - a result of Murphy's Law no doubt.
Stick