So I have 1 machine that is the only one in my cluster for boinc that is used 1/2 the time and suspended. Yesterday I noticed about 10 WU aborted and it states that the task was not started by deadline , this was then logged 4/27/2017 ( today being 4/22/2017 ) This was 4/20/2017 as I assume the time is correct 10:31am PST for example "
4/27/2017 10:31:17 AM | Amicable Numbers | Aborting task amicable_2_64_2091_1492699803.155143_895_0; not started and deadline has passed " and the log line before that " 4/27/2017 10:31:43 AM | Amicable Numbers | Scheduler request completed: got 24 new tasks " so , was this just me and a glitch in the AM matrix or did you redact some WU that went out? Anybody else?
Only 1 machine that sports an APU and 2 GPU's had this happen , all other machines nothing aborted.

This is that task: https://sech.me/boinc/Amicable/result.php?resultid=2229972

Outcome	Computation error
Client state	Aborted by user
Exit status	200 (0x000000C8) EXIT_UNSTARTED_LATE

Your BOINC client thought it was too late to start the task. I guess system clock glitched on your PC.