News for hackers. Stuff that matters.

A comment on Webserver Benchmark: Erlang vs. Go vs. Java vs. Node.js.

Re:Webserver Benchmark: Erlang vs. Go vs. Java vs. Node.js by jdlshore on Tuesday January 07, 2020 @10:40PM (#21988043)
I don't believe the built-in Node.js http server implements backpressure or 503 errors out of the box. Do the other systems tested do so? Could that explain the poor behavior and out-of-memory crash?
[ Parent | on HN ]
Re:Webserver Benchmark: Erlang vs. Go vs. Java vs. Node.js by kt315_ on Tuesday January 07, 2020 @11:05PM (#21988199)
None of the tested webservers were returning 503.
[ Parent | on HN ]
Re:Webserver Benchmark: Erlang vs. Go vs. Java vs. Node.js by kt315_ on Wednesday January 08, 2020 @10:47PM (#21998398)
While no 503s were observed in the test, there were timeouts. I updated the article with corresponding graph.
[ Parent | on HN ]