from Hacker News

Server-Side I/O Performance: Node vs. PHP vs. Java vs. Go

by vgallur on 12/13/17, 4:22 PM with 54 comments

  • by jwlake on 12/13/17, 5:32 PM

    The article talks about IO, but then benchmarks sha256. The node program doesn't work well because he's synchronously doing this:

      for (var i = 0; i < n; i++) { sha256(data); }
    He should be using crypto and the async versions. This benchmark actually is just measuring the speed of your sha256 implementation, which I would guess is equal on all 4 platforms if you actually do them correctly.

  • by glitcher on 12/13/17, 6:09 PM

    Pet peeve: tech articles/blogs which do not prominently display when the article was written. This is the single most important fact I search for before reading time sensitive tech info.

    I can't tell when this was written, am I missing something?

  • by ninkendo on 12/13/17, 5:30 PM

    I was reading the article until a giant fullscreen ad popped up and I had to close the tab. Sorry, no thanks.

  • by exabrial on 12/13/17, 5:36 PM

    I'm curious, which Java servers are creating a new Thread per request?

    Tomcat uses acceptors threads and a request executor pool. And, if available on your platform (which it probably is), it defaults to using non-blocking IO instead of polling.

    EDIT: It looks like he does acknowledge the executor threads are pooled. His main criticism is that "too many blocked threads are bad for a scheduler". But if Tomcat is using the NIO connector this doesn't apply, because your executor threads won't block on IO. And typically the pool size is limited to something manageable by a modern CPU (200 or so)

  • by pjmlp on 12/13/17, 6:05 PM

    So he acknowledges that has not benchmarked the best options for Java and then selects Go as a winner, what a joke.

  • by TheDong on 12/13/17, 5:03 PM

    The ease of use chart is a joke:

        Java: Requires callbacks
    nodejs: Requires Callbacks
    node has 'await', but there are other options as well than just plan callback hell.

    Java has pretty much whatever you want. Akka, futures/promises, etc.

  • by flavio81 on 12/13/17, 6:28 PM

    >Before I get into the section for Go, it’s appropriate for me to disclose that I am a Go fanboy

    Yes, it shows...

  • by mtgx on 12/13/17, 5:39 PM

    I visited the post fully expecting the author would benchmark the latest PHP 7.2. To my surprise, he wasn't even benchmarking 7.0, but the old 5.6. I can't take this test seriously.

  • by dingdingdang on 12/13/17, 7:06 PM

    PHP 5.6.x and old version of Node too - this article is too old to be of serious use.

  • by dullgiulio on 12/13/17, 5:06 PM

    The example of async I/O in Node.js is with a filesystem operation.

    How does that work? AFAIK Linux, unlike windows, doesn't have a proper async API for filesystem I/O.

    POSIX AIO is implemented using threads in libc and Linux AIO (io_submit, etc) only works with O_DIRECT (no kernel caching) and there are other issues depending on underlying filesystem driver.

    Is there any solution?

  • by ruffrey on 12/13/17, 6:43 PM

    While still suffering from the problem of benchmarks not reflecting real world performance, The Computer Language Benchmarks Game is an excellent resource.

    http://benchmarksgame.alioth.debian.org

  • by frugalmail on 12/13/17, 6:42 PM

    He's using the slowest and oldest framework on Java which still is preferable to use if you want things easy to read and maintain. But if you're going for raw I/O performance and already taking the readability and maintenance challenges with the other platforms, then he really should be using Netty, Vert.x or Akka for Java.

  • by politelemon on 12/13/17, 6:49 PM

    This post seems to be from about 7 months ago. Do the 'turbofan' changes of v8/v9 make any difference to the benchmarks?

    https://www.nearform.com/blog/node-js-is-getting-a-new-v8-wi...

  • by haglin on 12/18/17, 6:22 AM

    What a lousy article.

    "Most Java web servers work by starting a new thread of execution for each request that comes in "

    Nobody who has a performance critical application starts a new thread for every request.

    That would be beyond stupid.

  • by Neil44 on 12/13/17, 7:31 PM

    He’s using Apache in prefork mode with mod_php, which is a little bit 10 years ago. Worker or Event with php-fpm is standard now.

  • by bullen on 12/13/17, 9:11 PM

    Only Java can share memory between threads.

  • by callesgg on 12/13/17, 5:02 PM

    Line for line Go will be faster it is compiled.

    It is the only thing that matters line for line.