Gravity wave wu finished in 4 and a half mins
6 May 2013 8:00:58 UTC
Topic 196943
(moderation:
I have this wu
It finished in 4 and a half minutes which is most unusual. It doesn't seem to think its an error, although validation might once its compared with another user.
Gravity wave wu finished in 4 and a half mins
)
If you actually click on the taskID for your task in that quorum and read through the stderr output (which is quite short) you will find the following line
2013-05-06 17:49:52.6899 (2556) [debug]: Successfully read checkpoint:725which seems to suggest that the task was restarted from a saved checkpoint (725) and that the time recorded was just the balance of time to the finish point. The accumulated time is not saved in a checkpoint - It's actually saved in the state file for each active task - towards the end of the state file inside .... tags. Inside those tags you will normally see items of data like.... 1 1304 1910.781000 1913.718750 0.300000 2112.453000 ....This particular example comes from the state file of one of my nearby hosts :-).
Maybe (for some odd reason) the time stats weren't saved properly in your state file when your checkpoint (725) was written to disk.
Was the machine restarted for some reason? Was there some sort of crash and automatic restart? You should be able to see this in the stderr.txt file in the BOINC data directory. I'm sure I've previously seen some sort of loss of time stats like this under certain crash conditions on some of my hosts.
Cheers,
Gary.