Thousands of hits a day

[merdj]

Hello to All,

My friends in our home country will develop a big system (electric billing system) using PHP. I recommended F3 as their framework to use. Now I am worried because I never asked before if F3 can handle thousands of hits a day.

kindly someone enlighten me if F3 can handle such?

secondly, they need their source to be encoded/encrypted. Do you think F3 could be encrypted?

thank  you so much.

Bes regards,

[Doonge]

Nothing in particular hinges F3 in regards to the number of hits a day compared to raw PHP or any php framework.

Nothing in particular prevents F3, which is php code, to be encryted/encoded compared to raw PHP or any php framework.


I do not understand your question?

[Eric Fernance]

Those reports were really interesting. The results from multiple queries was I thought particularly relevant for framework performance.

Sometimes its good to get reassurance when you're going against the trend. And results like that certainly justify using F3.

Eric.

On March 23, 2015 6:22:32 PM AEST, xfra35 via Fat-Free Framework <f3-framework+APn2wQdllyYRCbrJQTu...@googlegroups.com> wrote:

As of performance, F3 performs very well so it clearly won't be the bottleneck. See the latest TechEmpower benchmarks: http://www.techempower.com/benchmarks/previews/round10/#section=data-r10&hw=peak&test=json&l=sg

If to exclude Yaf and Phalcon from the results (which are C-compiled extensions), then F3 ranks as the 2nd most performant PHP framework behind MicroMVC.

As of source encryption, I'd advise to directly ask the framework's author: Bong Cosca (e-mail address available here).

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

[merdj]

hello,

Now this is the peace of mind. The Benchmarks is sufficient enough to convince my friends to start using F3.

for the second question, I will email Mr. Bong.

thank you so much, your answers are highly appreciated.

[Steve Cove]

Good to see F3 doing well, but im finding it hard to understand how the ORM implementation beat the raw DB one, and more importantly how either beat raw php (in the multiple queries run).

Is this just an anomaly, or some kind of internal caching?

[xfra35]

[xfra35]

Oh you're right. I've read the results a bit too quickly. Actually, if to exclude Phalcon and Yaf, F3 performs best in 4 categories out of 6!

As you pointed it out, it performs particularly well in the "multiple queries" category, even better than the C-compiled extensions!

[ikkez]

well it is still just preview data. i could imagine that the results maybe change again. I checked the php-raw multi-db test but did not find anything that differs a lot from the F3 test. https://github.com/TechEmpower/FrameworkBenchmarks/blob/master/frameworks/PHP/php/dbraw.php vs. https://github.com/F3Community/FrameworkBenchmarks/blob/master/frameworks/PHP/php-fatfree/index.php#L21
I was also suprised, but caching only effects table schema caching, which isn't that important in these tests. Table rows/results are not cached here, and would not be allowed in the benchmark rules. So maybe it's really just an anomaly or some fat-free magic i cannot expain.

[Nikolaos Giannopoulos]

Anyone know why these latest TechEmpower Fat-Free tests are not completing?  I can see the output and error logs but hard to tell....

Also... any way to let anyone know how to fix this.  I guess bottom line is it would be nice to keep seeing recent results from Fat-Free vs. other frameworks.....

[ikkez]

this was already fixed.. we're just waiting for the next round taking place.

[Emmanuel]

Thousand hits a day with F3 ?

you  know the NF3 portal get up to 100 users "at the same time" (they say concurrent right?  thus they dont even know each other right?)

so 100 hits X 36 HTTP requests for home page (yes many photos on the landpage)

& 100 hits X 6 request for cached "details pages"

at the same time, roughly , peak momebnts too

HOW   TO    TEST  MOCK  FAKE  TRAFFIC  to  the F3 entry door (usually index.php but  to also please to others entry doors dedicated to categories and location ,since not only index.php get the traffic from the server )   

Thanks in advance F3::Wizards in there

Emmanuel

P.S.

[ved]

You can use a performance/stress testing tool.

Some common ones are:

Apache JMeter

Apache Benchmarking Tool

httperf

siege

httpload

There are also online services that can do that, but they're usually paid services. Here are some:

loadimpact

flood.io

I'm sure you can find more by searching google for "http stress test", "http benchmark", "webserver load testing", etc.

[Emmanuel]

WOW  @ved  !!! it-s like whe\n you have many phoen numbers on a ID card .  

• ? What one to pick up when the time has come ?
  
• The 1st  ,  right?

thanks, gonna prepare those tests now :))

On Sunday, March 22, 2015 at 12:22:29 PM UTC+5:30, merdj wrote:

[ved]

No problem.

JMeter is the most comprehensive of them all, although maybe not the easiest or quickest to setup and use.

But it does comes with a full java GUI that you can use to configure pretty much any type of http requests (you can setup cookies, headers, post variables, etc).

It's very flexible.

For quick tests from a terminal, ab or siege should work pretty well and don't have so many dependencies.

Good luck.

[ikkez]

I think the user exodus4d from slack told once that he was able to handle around 5000 concurrent users in his pretty complex mapping tool with some optimisations and redis cache.

[Jérôme Barth]

Hello,

Experience feedback :

In our company, we developed an invoice's manager with F3 (we sell an invoice management software SAAS).

It's work very well (several thousand invoices per day and several tens of thousand invoice's lines per day).

We used F3 for the backend and Angular for the frontend.

Best regards,

Jérôme