The importance of check-pointing

A simulation/benchmark that I’m running at work, and started mid-friday afternoon .. just crashed. It takes about 36 hours to run it, and because of the crash I have to go back into the office and restart it.

The 64-thousand-penny question is, now, did the test checkpoint (aka, save it’s state) before crashing? If it did, only the remaining tests have to be run. If it didn’t .. it has to be restarted from the top

ARGH!

AAAAND .. of course Murphy wins by a mile.   Everything crashed, and now I’ll have to spend the better part of three hours setting it all up again.   Darnit.