-
Notifications
You must be signed in to change notification settings - Fork 83
Description
Quick summary
We have about 2,000 or so ActivityPub followers on our site at We Distribute. Whenever we post an article, the site briefly goes down for a little while. I suspect this is happening because the server needs to dispatch posts to 2,000 people?
Steps to reproduce
- Migrate followers from a decent-sized Fediverse account to WordPress.
- Make sure Enable Mastodon Apps and Friends are also installed.
- Post an article. Alternatively, post a podcast episode with Seriously Simple Podcasting.
- Watch as your site briefly goes down.
Some things have helped mitigate various problems with ActivityPub and WordPress, such as using Surge, Redis, and Jetpack. Curiously, the site still goes down pretty frequently, which is not ideal.
What's weird is that most parts of the server are still fine. Nginx, Redis, PHP-FPM, and various applications all seem to run just fine. It's not clear why this happens, just that the WordPress part of the site becomes unreachable until about 20 minutes pass.
Are there best practices for scaling up a WordPress installation that heavily makes use of ActivityPub?
Site owner impact
More than 60% of the total website/platform users
Severity
Critical
What other impact(s) does this issue have?
No response
If a workaround is available, please outline it here.
No response
Additional Context:
- OS: Ubuntu 22.04.4 LTS (GNU/Linux 5.15.0-134-generic x86_64)
- Web Server: nginx/1.18.0 (Ubuntu)
- PHP Version: 8.3.11