Web Framework Benchmarks Round 5 • Hacker News Board

Re: Web Framework Benchmarks Round 5

#11

Post by xt » Fri, May 17, 2013, 11:20 PM UTC

stefantalpalaru wrote: post #10

snaky wrote: post #2

OpenResty is twice as fast as raw Go at multiple queries DB test (dedicated hardware), huh.

A pure Go MySQL driver is being used for the Go test. OpenResty probably uses a thin wrapper around the C library.

The openresty mysql driver is in pure lua. It has a very efficient socket/pooling mechanism, using sockets from nginx.

Re: Web Framework Benchmarks Round 5

#12

Post by bhauer » Fri, May 17, 2013, 11:21 PM UTC

stefantalpalaru wrote: post #8

Please add Revel (the Go framework) to the dedicated hardware tests.

Hi stefantalpalaru. Good eye! We will get that added and patch it into Round 5 soon.

Re: Web Framework Benchmarks Round 5

#13

Post by bhauer » Fri, May 17, 2013, 11:22 PM UTC

cmircea wrote: post #9

bhauer wrote: post #7

Earlier quoted context omitted.

Thanks for the tip! I need to be clear that the ASP.NET implementation that you see in Round 5 was contributed by user @pdonald at Github. That said, we'd be happy to receive more pull requests with ASP.NET changes and improvements. We suspect there is a lot of room for improvement in the Windows numbers.

I've just submitted a pull request.

You may be the fastest pull-requester I've seen in the five rounds of this project. Good show.

Re: Web Framework Benchmarks Round 5

#14

Post by alberth » Fri, May 17, 2013, 11:22 PM UTC

Why is php one of the fastest on the database query test but one of the slowest on all other tests?

Re: Web Framework Benchmarks Round 5

#15

Post by FooBarWidget » Fri, May 17, 2013, 11:26 PM UTC

Please add Phusion Passenger (https://www.phusionpassenger.com/) to the benchmark for Ruby apps. Right now Unicorn is the only server in that benchmark for Ruby but it's far from the only available server.

Re: Web Framework Benchmarks Round 5

#16

Post by te_chris » Fri, May 17, 2013, 11:27 PM UTC

Why is the overhead of all the PHP frameworks so high? Is it because they have to evaluate all the framework code per request? The difference between raw PHP and symfony2 is massive!

Re: Web Framework Benchmarks Round 5

#17

Post by cmircea » Fri, May 17, 2013, 11:28 PM UTC

te_chris wrote: post #16

Why is the overhead of all the PHP frameworks so high? Is it because they have to evaluate all the framework code per request? The difference between raw PHP and symfony2 is massive!

Yeah, pretty much.

Re: Web Framework Benchmarks Round 5

#18

Post by aweb » Fri, May 17, 2013, 11:28 PM UTC

No metion of Yii? Too bad, it's the best PHP Framework I've ever used. Fast but complete at the same time, and fully OO.

Re: Web Framework Benchmarks Round 5

#19

Post by room271 » Fri, May 17, 2013, 11:29 PM UTC

I normally use Scalatra for Scala stuff, but interesting to see Spray's inclusion and results.

Re: Web Framework Benchmarks Round 5

#20

Post by FooBarWidget » Fri, May 17, 2013, 11:30 PM UTC

te_chris wrote: post #16

Why is the overhead of all the PHP frameworks so high? Is it because they have to evaluate all the framework code per request? The difference between raw PHP and symfony2 is massive!

I think it's because raw PHP doesn't do anything. You're literally benchmarking how fast you can do nothing. As soon as you add any kind of logic the number goes way down. It's like saying the performance of adding 4 numbers is a massive difference from the performance of adding 2 numbers.