web framework benchmarks

[António Ramos]

just wait for the graphs...

https://www.techempower.com/benchmarks/#section=data-r11&hw=peak&test=update

Can someone get Phoenix in the list ?

[Steve Domin]

I think it was added at some point but there was a couple of mistakes with the way it was setup, and that lead to inaccurate results.

Chris or someone else probably remembers better than I do.

--
You received this message because you are subscribed to the Google Groups "phoenix-talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to phoenix-talk...@googlegroups.com.
To post to this group, send email to phoeni...@googlegroups.com.
Visit this group at https://groups.google.com/group/phoenix-talk.
To view this discussion on the web visit https://groups.google.com/d/msgid/phoenix-talk/CAEM0BxOxh9zmXWLzvj%2BvHhk8PUd1haxm8BEKVNfFXgzGGyP-zQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

[Alexander]

re: techempower.com benchmarks:

Things fixed in their repo, new results going to be included in the next round (round 12)

ETA: March-April-? or something like that

[Steve Domin]

Yes, should be around March/April.

[Alexander]

Round 12 is out:  https://www.techempower.com/benchmarks/#section=data-r12

Congrats, Phoenix results completely screwed (look at number of errors)

[Chris McCord]

They had a very poor initial application doing csrf token generation thru the browser pipeline on the JSON benchmarks, as well as dev db pool sizes that were 10x less than other applications. We sent them a PR months ago and I was hoping for true results this time around. I'm not sure where the errors are from, and it's not something we see when we perform our own benchmarks. I wish there was a better approach to these benchmarks around collaboration but it seems some frameworks are given more TLC than others.

To view this discussion on the web visit https://groups.google.com/d/msgid/phoenix-talk/71356b77-8d52-4de9-a895-d145e7825100%40googlegroups.com.

[elegua]

They apparently offer a preview of results in order to allow problems
like this to be fixed before the final test run, but the hardware
platform they have been using was going away, so they rushed and skipped
the preview step this round.

https://www.techempower.com/blog/2016/02/25/framework-benchmarks-round-12/

Gilbert

[António Ramos]

Its in Portuguese 

http://pompeu.github.io/articles/nodevs-elixir

The guy concluded that clearly Elixir is good and fast compared with python ruby and php. But with big brothers like nodejs doesnt stand a chance!!!

any comments ? 

To view this discussion on the web visit https://groups.google.com/d/msgid/phoenix-talk/D484D715-CDA9-43EE-A3A0-29AABD241D44%40chrismccord.com.

[Alexander]

Re: nodevs-elixir article

benchmarking in its finest

* router vs framework (what's in your pipeline, again?)

* hash vs JSON.parse every request

* charts starting from non-zero value

--
You received this message because you are subscribed to a topic in the Google Groups "phoenix-talk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/phoenix-talk/6cjJx61FPBg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to phoenix-talk...@googlegroups.com.

To post to this group, send email to phoeni...@googlegroups.com.
Visit this group at https://groups.google.com/group/phoenix-talk.

To view this discussion on the web visit https://groups.google.com/d/msgid/phoenix-talk/CAEM0BxPB13xOoSWGZZWwaE9aSnaHzOSm5cpm3y8%2BwH_1EZLAAA%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Alexander

alex...@iron.io

[Theron Boerner]

It depends on the core count. Having many more cores means Elixir beats Node.js easily.

(Old, but still relevant) https://gist.github.com/omnibs/e5e72b31e6bd25caf39a#results

To view this discussion on the web visit https://groups.google.com/d/msgid/phoenix-talk/CA%2B%2BYLExw%2BC1UN1MNrPVj9w%2BqbSLKOT%2BvXZNK2D2nJEp7Wq309Q%40mail.gmail.com.

[Alexander]

Round 12 logs released, phoenix: https://github.com/TechEmpower/TFB-Round-12/tree/master/peak/linux/2016-02-22-final/latest/logs/phoenix

Nothing suspicious so far, which is weird

[Eugeniu Tambur]

Hi.

169,526 errors ???

phoenix

11,303

3.8%

Mcr

Eli

Erl

Cow

Lin

Pg

Lin

Ful

Rea

169,526

[Christian Kruse]

Hey,

On Mon, 7 Mar 2016 13:17:07 -0800 (PST)
Alexander <tozo...@gmail.com> wrote:

> Round 12 logs released,
> phoenix:
> https://github.com/TechEmpower/TFB-Round-12/tree/master/peak/linux/2016-02-22-final/latest/logs/phoenix
> Nothing suspicious so far, which is weird

Was logging to file configured at all? Not even request logs have been
written?

Best regards,
--
Christian Kruse
https://wwwtech.de/about