[wp-trac] [WordPress Trac] #56040: Port Persistent Object Cache Health Check from performance plugin to core
WordPress Trac
noreply at wordpress.org
Tue Nov 15 06:06:01 UTC 2022
#56040: Port Persistent Object Cache Health Check from performance plugin to core
-------------------------------------------------+-------------------------
Reporter: furi3r | Owner: furi3r
Type: defect (bug) | Status: closed
Priority: normal | Milestone: 6.1
Component: Site Health | Version:
Severity: normal | Resolution: fixed
Keywords: has-patch has-unit-tests has-dev- | Focuses:
note | performance
-------------------------------------------------+-------------------------
Changes (by KnowingArt_com):
* type: enhancement => defect (bug)
Comment:
As a WordPress host, there's nothing "healthy" about this new
recommendation.
Even the memecached documentation agrees with me:
"If you have a database host, give as much ram as possible to the
database. When cache misses do happen, you'll get more benefit from
ensuring your indexes and data are already in memory."
https://github.com/memcached/memcached/wiki/Hardware
Read the memcached documentation, and use common sense. In most cases, it
makes a more sense to allocate "extra" memory to other areas: Innodb,
database connections, more php-fpm children, etc. I've been hosting
WordPress for almost 20 years, that's how you do it.
More caching is rarely the solution, you're just asking for trouble.
To spin up an additional memcached server that will eat up all free
memory, that is a recipe for disaster, as the memcached documentation
itself warns about.
"Will take as much memory as you give it."
Sounds like a good way to end up with an unstable system. Oh wait, that's
not a bug, it's a feature LOL.
"Assign physical memory, with a few percent extra, to a memcached server.
Do not over-allocate memory and expect swap to save you. Performance will
be very, very poor. Take extra care to monitor if your server is using
swap, and tune if necessary."
https://github.com/memcached/memcached/wiki/Hardware
And you're actually recommending we use this? Do you really expect hosts
to have a separate memcached server, with special swap and network
requirements, just for memcached? Who is going to pay for and administer
that?
And will this make blogs faster? No, because adding more software and
unnecessary network traffic never makes things faster or more reliable.
Even the memcached documentation warns about it:
"Network requirements will vary greatly by the average size of your
memcached items. Your application should aim to keep them small, as it can
mean the difference between being fine with gigabit inter-switch uplinks,
or being completely toast."
Yeah, that's a "healthy" recommendation. Sarcasm off.
On top of that, the recommended plugins look to be almost experimental.
Do you think I am going to configure and manage TWO databases (Redis +
MySQL) to satisfy this little nag message? I already spent years of my
life getting MySQL to work well. There is no way I am going to reinvent
the wheel (as if WordPress hosting is not complicated enough) to add Redis
into the mix, when MySQL has all kinds of caches that work perfectly fine.
If people have slow WordPress performance, it's NOT because they lack an
object cache.
I am really sick of all the silly "health" recommendations that are shoved
in the face of my clients on the dashboard. I can't help but wonder if
there are any conflicts of interest here. Who really benefits from an
object cache? Not the blogger. Not the host. Who then?
Show me any proof memory is better allocated to an object cache and not
the database or php-fpm. And even if you can prove it on a poorly
configured system, we're talking about a few milliseconds.
--
Ticket URL: <https://core.trac.wordpress.org/ticket/56040#comment:33>
WordPress Trac <https://core.trac.wordpress.org/>
WordPress publishing platform
More information about the wp-trac
mailing list