Mark Komarinski wrote: > And as for replacing hardware... > The container gets replaced after some percentage > of the nodes inside fail and are unrecoverable.
That's a good point. I've heard similar stories about how Google treats their data centers where failed computers are just powered down and left in the rack, only occasionally swapped out for new units. So sure, your Pi cluster could simply be overbuilt to allow margin for failed units. It just rubs me the wrong way to design a system that ignores ease of repair. -Tom _______________________________________________ Hardwarehacking mailing list [email protected] http://lists.blu.org/mailman/listinfo/hardwarehacking
