Need to do some maintenance on one or two of the crunchers, and I've noticed that on machines with 2 SMP clients when I shut down one the other becomes unstable and I get the dreaded error63 on both after that ...essentially losing both of them.
I know I can add a flag so that it will send a completed WU and not get more but would rather not wait hours (or days) for a current WU to finish.
Anyone have some good practices that work for you without losing a WU?
I know I can add a flag so that it will send a completed WU and not get more but would rather not wait hours (or days) for a current WU to finish.
Anyone have some good practices that work for you without losing a WU?