[wp-trac] [WordPress Trac] #57916: Explore and assess 6.2 server-side performance regressions

[WordPress Trac]

#57916: Explore and assess 6.2 server-side performance regressions
----------------------------+--------------------------
 Reporter:  flixos90        |       Owner:  flixos90
     Type:  task (blessed)  |      Status:  assigned
 Priority:  normal          |   Milestone:  6.2
Component:  General         |     Version:  trunk
 Severity:  normal          |  Resolution:
 Keywords:  has-patch       |     Focuses:  performance
----------------------------+--------------------------

Comment (by flixos90):

 @spacedmonkey Can you clarify how you tested this, in which environment
 etc.? It looks like you tested with a block theme for example.

 I just manually added the code from your PR directly into the regular ZIP
 production build code for WP 6.2-RC2 and tested like that, also doing a
 fresh comparison with 6.1.1. I used a LocalWP site, where I "hacked" the
 `wp-includes/blocks.php` file with the code change from
 https://github.com/WordPress/wordpress-develop/pull/4243.

 Here are my numbers for TT1 theme:

 WP 6.1.1:
 * `wp-before-template`: 37.89ms
     * `wp-init` (part of the above): 18.66ms
 * `wp-template`: 25.89ms
 * `wp-total`: 63.88ms

 WP 6.2-RC2:
 * `wp-before-template`: 37.99ms
     * `wp-init` (part of the above): 19.46ms
 * `wp-template`: 23.3ms
 * `wp-total`: 61.39ms

 WP 6.2-RC2 with code from the PR:
 * `wp-before-template`: 38.11ms
     * `wp-init` (part of the above): 19.55ms
 * `wp-template`: 23.38ms
 * `wp-total`: 61.62ms

 Two main observations:
 * For some reason, now my numbers again show that 6.2 is faster than
 6.1.1, which is great and covers with the numbers I had before. Still not
 sure what went wrong earlier this week, again I haven't even touched my
 setup. Very confusing.
 * The PR does not seem to have a notable effect on performance when using
 TT1, so based on my data I wouldn't think it's what is causing the `init`
 regression (which, again, is still there regardless of which data we
 trust, whether WP 6.2 overall is faster or not). I see the major
 improvement on your tests though, but again I wonder if this is just a
 separate performance enhancement worth making rather than being the cause
 for the regression.

-- 
Ticket URL: <https://core.trac.wordpress.org/ticket/57916#comment:9>
WordPress Trac <https://core.trac.wordpress.org/>
WordPress publishing platform