3 ms·
Well until that major outage, I don't know how much infrastructure skype was running. I recall something about supernodes (or super super?) that they would be
by ohashi 15y ago
Well until that major outage, I don't know how much infrastructure skype was running. I recall something about supernodes (or super super?) that they would be hosting to prevent these cascading failures. The whole P2P nature was the clever trick, not sure how much more data centers will actually help when it's software bugs that seem to take the system down.
- bonzoesc 15y agoThey've spun up "mega-supernodes," when cascading failures have taken the network down before[1], but they don't need to run them full-time, just when traffic is such that the traditional supernodes can't handle the traffic of a new network. 1: http://news.ycombinator.com/item?id=2048630 http://news.ycombinator.com/item?id=2048630
- ohashi 15y agoThanks!