Partisan: Enabling Cloud-Scale Erlang Applications · HackerPapers discussions

Partisan: Enabling Cloud-Scale Erlang Applications — mentioned by hosh

Fri, 04 Aug 2023 09:30:47 GMT

There is a reason stuff like Partisan were conceived - https://arxiv.org/abs/1802.02652 If you do a lot of cross node talk with say, the Phoenix channels, then lowering the total number of nodes helps (n*(n-1) for a full mesh). And since the BEAM scheduler will try to use up all available cores by default, past a certain threshold of nodes, it makes more sense to use smaller numbers of nodes with more cores per node. When you also have say, an observaility agent taking up 1000 mcore, doubling the instance size makes a difference. In the last shop, we would idle at 4-cores, jump to 16-cores…

Paper and discussion history. A mention is not an endorsement.

Partisan: Enabling Cloud-Scale Erlang Applications — mentioned by zzzcpan

Sun, 02 Dec 2018 14:31:57 GMT

This is mostly about Partisan [1], touching a bit Hyparview [2] and a bit fault injection. [1] https://arxiv.org/pdf/1802.02652.pdf [2] http://asc.di.fct.unl.pt/~jleitao/pdf/dsn07-leitao.pdf

Paper and discussion history. A mention is not an endorsement.

Partisan: Enabling Cloud-Scale Erlang Applications — HN submission

Sat, 10 Feb 2018 02:28:24 GMT

Partisan: Enabling Cloud-Scale Erlang Applications

88 points; 7 discussion replies on HN. Paper and source history. A submission is not an endorsement.