Joined: 3 Oct 07
Earlier this year, LHC started sending some workunits that take ten times longer to complete than all the others. However, the deadline is the same, about a week, for these much longer units.
The Pentium-3-class machine on which I run LHC usually is not up for forty hours in one week that would be needed to return the unit before its deadline. Therefore, I've had to do a User Abort on these long units whenever they're sent. That's a shame.
Can LHC please improve this so that the long units receive deadlines of at least two weeks? How about a month?