August 2015: Tasks Still Failing To Complete.
As of August, 2015, I am still routinely receiving tasks which fail to complete. I have over 15 computers running various BOINC projects, and several running Atlas, so I am not new to distributed computing and am fairly certain the issue is in the tasks themselves and not on my end. Does anyone in the Atlas project know with 100% certainty what the problem is with these tasks failing to complete? In actuality, I do not know that they don't, but when I find tasks stuck at 99.99% and at over 100 hours, I abort them, as they never seem to finish. Will they? And if not, at what point, in hours, should they be aborted? I hate to abort tasks which might actually finish, but it seems more likely a waste of processor time when they start hitting 25 or more hours on an FX processor, when other tasks finish in a few hours. Because of the number of computers I run, not every one gets checked every day, consequently, I see a fair number of tasks at a ridiculous number of hours and would like to see this problem resolved.
|ID: 2759 · Rating: 0 · rate: / Reply Quote|