5 ms·
I don't know what the formalisms here are, but as someone intermittently (thankfully, at this point, rarely) responsible for some fairly large-scale distributed
by tptacek 2y ago
I don't know what the formalisms here are, but as someone intermittently (thankfully, at this point, rarely) responsible for some fairly large-scale distributed systems: the "network that is really slow" case is much scarier to me than the "total network partition". Routine partial failure is a big part of what makes distributed systems hard to reason about.
- rubiquity 2y agoEven more fun are asymmetric network degradations or partitions.