Anyone using the full page cache



  • I am trying to test the integrated full page cache of ThirtyBees. I heavily tested a module called ExpressCache in the past and while it performed incredibly in terms of speed it had some issues causing me to never be able to update it.

    Having this feature integrated into the platform is an amazing idea so today I gave it a shot on the 1.0.4-RC1.

    I have enabled the File System option in Server Side cache and then I use the default settings for controllers and for testing I checked everything to green for the modules.

    The thing is that even after repeated tests, I am not seeing any difference. Lighthouse reports the exact same Performance score, load times there are unchanged. On webpagetest.org if anything the ttfb increased a bit compared to when I ran the tests without the full page cache activated.

    Anyone has any idea, experience from using this? The instructions on the blog are quite clear and in the backoffice it looks ok. I am testing on a vanilla installation running the tests on the product page with a sample product.

    0_1522242488118_Screen Shot 2018-03-28 at 15.07.07.png



  • maybe if there’s no traffic you can’t see differences.


  • Global Moderator

    You need to make everything red… Only the things which should be loaded dynamically should be green…

    @lesley please finally write this into the core. It always confuses merchants.



  • but the blog post says:
    "When a hook is pink that means it is not being cached. Clicking on the hook will turn it green and allow it to be cached. "


  • Global Moderator

    Well I am 99% sure, that the blog is wrong then… But you can test it out… What the blog writes, would be the logical way, but it isn’t like that. Thats why this issues come up every month.



  • hmm, ok I have to run but I will retry the test tomorrow then.



  • @wakabayashi so I have rerun the test unchecking everything so the module boxes are red/pink.

    I see a small change, about 100 msec in the TTFB on webpage test but that could be from natural fluctuations from day to day as well. The ttfb in general is a bit better (around 100msec) than my PS 1.5 installation and PS 1.7 but taking into account that my 1.5 installation is a live shop with much more things to load on the page, I am not sure the real difference in the end. The fastest ttfb I got for TB was around 450 msec and when I ran the ExpressCache module on the same hosting with my 1.5 shop I was below 100 msec (compared to around 550msec without the caching module).

    The lighthouse performance score is identical regardless if the boxes are green or red.

    Maybe I am missing something but I have reloaded the page many time navigating back and forth and I would assume this is what is causing the page to end up in the full page cache. Happy to test more if anyone has some ideas, I was just expecting to see a significant difference when activating this but again, perhaps I am doing something wrong.



  • @lesley Does this really sound correct?


  • administrators

    @wakabayashi Is right, I need to update the docs. Sorry about that.


Log in to reply
 

Looks like your connection to thirty bees forum was lost, please wait while we try to reconnect.