• Sooo, I have a nice bug… When I activate this plugin, it seems to work (comment and reply on this post are from Mastodon) but after a while (around 10/30 min? Can’t really get a perfect timing on it), it crash the websiteS. Yes, with a S. EVERY SITE on my server (I’m renting my own machine) is down.

    504 Gateway Time-out for all!

    My admin and me are puzzled by this one. Deleting activitypub + webfinger + nodeinfo fixed the problem after cleaning a bunch of 0 byte .htaccess created in /wp-content/uploads/ for every site (no idea how…)

    Nginx, Varnish and Apache were “fine”, but nothing works after a while. I tried to reinstall just activityPub, everything went down again. I deleted the folder, everything works immediately.

    I would love to use this plugin on all my sites but obviously, right now something is very wrong with it and I have no clue if it’s our configuration or the plugin itself. oO

    The page I need help with: [log in to see the link]

Viewing 8 replies - 1 through 8 (of 8 total)
  • Plugin Author Matthias Pfefferle

    (@pfefferle)

    Does it no longer work because of high load?

    https://github.com/Automattic/wordpress-activitypub/issues/611

    Thread Starter Cafeine

    (@cafeine)

    I don’t think so but we’ll double check, we plan to do other tests because it was very weird (and we need to understand what created those weird .htaccess, could be unrelated of course).

    Do you have data about the CPU usage we “should” see for a new unknown account?

    Thanks again for your help and your work on this.

    Plugin Author Matthias Pfefferle

    (@pfefferle)

    The plugin does nothing with .htaccess !

    I have not experienced the issue by myself, but the issue states:

    When a post from WordPress is re-tooted, favorited, or replied to WordPress locks up. cPanel indicates all memory consumed (1 GBytes of 1 GBytes), and all I/O is consumed (10 MB/s of 10 MB/s)

    May I ask where you host your WordPress installation?

    I’ve experienced the same thing. I have three sites on a shared server operating in a virtual machine operated by my site hosting provider, whc.ca.

    Sometime after I installed the latest revision to fix the problem with the web finger endpoint problem, my server started to crash. All services on my server went down. Moving the ActivityPub plugin out of the active plugins directory allowed the server to recover. I’ve not tried to replicate the problem, as two of my sites are for my small business, and I want to avoid having them down during the work week.

    Plugin Author Matthias Pfefferle

    (@pfefferle)

    @macadk do you have some more infos for me? For example what version exactly, what hosting env, do you have an example of the error?

    Thread Starter Cafeine

    (@cafeine)

    May I ask where you host your WordPress installation?

    It’s on online.net / scaleway, on a dedicated server (Intel? Xeon E3 1245v5 / 32 GB of RAM, 3x250GB SSD). I host a discourse forum and around 7/8 WP, small ones. Very light traffic for the server.

    We’ll run more tests Wednesday evening. Thanks for the help, I’ll keep you informed!

    Thread Starter Cafeine

    (@cafeine)

    @macadk crash seems very close to what we had. I’ll try again tomorrow and we’ll see what kind of logs we can get. ??

    do you have some more infos for me? For example what version exactly, what hosting env, do you have an example of the error?

    @pfefferle

    I installed the latest version of ActivityPub, v2.0.1. I’m running it on a WordPress v6.4.2 site. It’s hosted at whc.ca, on a server running CentOS v7.9.2009 STANDARD. The web server is running Apache v2.4.58.

    As for the errors, the server would bog down until it would start to throw 503s. Many of the server functions would hang, like Apache, and lfd.

Viewing 8 replies - 1 through 8 (of 8 total)
  • The topic ‘ActivityPub crashing every site on my server… ??’ is closed to new replies.