websites slow? what module is to blame?

[Nicolaas Thiemen Francken]

Hi,

We have to two sites that load  in 200ms / 300ms respectively (on my dev machine). A fresh install of SS 3.5 loads in about 35ms.  I was wondering if anyone has worked out a quick and easy way to find the bottle necks.  

There are two modules that could help:

• https://github.com/heyday/silverstripe-xhprof
  
• https://github.com/titledk/silverstripe-xhprof

​Are those modules still in use / current?

Has anyone tried the module by module removal approach?

I would love some ideas.

Thank you

Nicolaas

[Simon Erkelens]

I've not tried anything like removing module-by-module, but it's obvious the more modules, the larger the application gets.

Installing XHProf is a good start, although you don't need a module for it

[swaiba]

Generally, I've just used the xdebug grind outputs and to identify bottlenecks that way.

There was a mention of some modules that give a function / query overview of this time taken (like the network tab in chrome dev tools) which would also be useful.

I would not see the adding / removal of modules to be a way to achieve this as it would surely take more time programming up the gaps where the module was each time and skew all results.

[Jonathon Menz]

Another factor may be that a fresh install won't be spending a lot of time on template rendering. You'll be looping over 3 pages to build a menu and spit out some content and that's about it. Once you have a lot of pages and dataobjects to iterate over I guess that could add a lot of load. I like to use a combination of partial caching and the dynamic cache module to improve that.

To test that theory I guess you could change your database details so a new DB is generated with just the default content, and see what the load times look like then.

The clockwork module can help you see bottlenecks too.

[Nicolaas Thiemen Francken]

Thank ya'll for the great replies!

@Jonathon 

What was super interesting is that when we changed the template from "lots of $%^&" to literally "test" (i.e. page just outputs the word test on the screen) there was hardly any difference (cant remember exact numbers, but it may have reduced the speed by 5-10%). That is despite there being some tricky menus, etc...

The 300ms website has hardly any caching and the 200ms site has lots of caching. 

--

You received this message because you are subscribed to the Google Groups "SilverStripe Core Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to silverstripe-dev+unsubscribe@googlegroups.com.
To post to this group, send email to silverstripe-dev@googlegroups.com.
Visit this group at https://groups.google.com/group/silverstripe-dev.
For more options, visit https://groups.google.com/d/optout.

--

Nicolaas Thiemen Francken
  www.sunnysideup.co.nz

  phone: +64221697577