Jump to content

Beta 6 testing results.


Rhett

Recommended Posts

Posted

I have been testing IPS4 along the way for speed, using many different configurations, and wanted to share the end results with Beta 6 tonight. Of course this is a clean install, however the numbers are very promising. :)

 

 

test6.png

speed7.png

speed8.png

Posted

Hello,

Need to compare with previous versions

​Here is a test of ips4 beta5 vs 347, same server, same setup, same fresh install for both versions. (apples to apples)

I didn't post all of the prior testing, I just wanted to show the great results of beta6 ;)

 

 

 

 

ips4-347.jpg

Posted

FWIW, the time to load, ie initial load with no browser cache on a dialup connection (don't ask!) comparing 34x to 4x, 4x is slightly faster initially to load the forum index so tis all good.

Posted

Those test were done using nginx as a reverse proxy for apache 2.4, php 5.6 with Zend opcache, and mariadb 5.5.

I also did the same test using the same configs without nginx and just apache 2.4, there wasn't enough of a difference to matter with normal loads, however under stress testing while also running the performance test at the same time, nginx performed better. 

Here was the results during the stress test of 200 request per second while also running the performance test. (please note these performance testing sites are not 100% accurate, however it's a decent guide to go by for testing if you use them all and combine the results with some common sense.)

 

 

 

test-load.jpg

Posted

I personally don't care for websites like WebPageTest since they're all very finicky on the way they actually perform these tests. For example, the index page on my production forum actually loads in ~0.88 seconds according to WebPageTest, but because of some beacon and analytic scripts that run as delayed Ajax requests in the background, it doesn't recognize the page as being fully loaded until those all fire off and return a response, even though they don't impact actual page load times at all.

Have you ever used New Relic Rhett? If you haven't, I think you'll really like the insights it can provide you with. It actually measures the application performance for one thing. Browser performance can vary on so many factors that are almost completely irrelevant to the server software itself.

These are the performance metrics over the last 24 hours for one of my production IP.Board 3.4.7 forums to provide an example,
Screenshot_from_2015-01-29_23:22:20.thum

It would definitely be nice to see some comparison metrics like this instead.

Posted

I normally don't use any testing sites from third parties, I can tell what sites are doing from the back end and what is quick or not etc,  I go by real usage of the site, some people love test site results though, and why I posted these.  :) 

 

 

  • 2 weeks later...
Posted

Any performance improvement under beta 8?

 

​Yes it bumped up a few notches in performance and scores :)   Look up two posts  ;) 

  • 4 weeks later...
Posted

Optimize images, including add size to link them loaded.

Turn compression and cache, use htaccess

Use cookie-free domains

and will improve performance boost your site.

Posted

You don't have gzip compression enabled. There's no excuse for that. Enable it. Seriously.
http://www.feedthebot.com/pagespeed/enable-compression.html

You can add cache control headers to static resources by adding the following to your .htaccess file,

# Cache-Control Headers
<ifModule mod_headers.c>
  <filesMatch ".(ico|jpe?g|png|gif|swf|css|js)$">
    Header set Cache-Control "max-age=2592000, public"
  </filesMatch>
</ifModule>

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...