Redis Probleme? (Forum nach Upgrade defekt)

[Steven]

When you rebuild, you have a [FAILED] message at the end?

Could you show us the lines before this failed alert? (as much as you can)

We’ll find some clues in this.

[L30110]

I don’t think there’s a FAILED message at the end. I’ll run another build. Thanks!

[Steven]

If the rebuild goes to the end, my guess is a broken plugin or theme component.

Check the console in your browser (F12) to see if you have some red alerts.

You can also try [your-forum]/safe-mode to load your forum without plugins and themes

[L30110]

Yeah, no FAILURE. The build ends apparently fine and then the docker run appears fine as well.

[L30110]

Nothing obvious in the console. Attempting to run safe mode but get “site can’t be reached” error from Chrome. But I can access the (the normal way) the home page and static messages.

[L30110]

Here’s the tail of the current production log:

/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/redis-4.5.1/lib/redis.rb:72:in `block in synchronize'
/usr/local/lib/ruby/2.7.0/monitor.rb:202:in `synchronize'
/usr/local/lib/ruby/2.7.0/monitor.rb:202:in `mon_synchronize'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/redis-4.5.1/lib/redis.rb:72:in `synchronize'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/redis-4.5.1/lib/redis.rb:966:in `get'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/message_bus-3.3.8/lib/message_bus/backends/redis.rb:366:in `process_global_backlog'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/message_bus-3.3.8/lib/message_bus/backends/redis.rb:274:in `block in global_subscribe'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/message_bus-3.3.8/lib/message_bus/backends/redis.rb:287:in `global_subscribe'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/message_bus-3.3.8/lib/message_bus.rb:786:in `global_subscribe_thread'
/var/www/discourse/vendor/bundle/ruby/2.7.0/gems/message_bus-3.3.8/lib/message_bus.rb:734:in `block in new_subscriber_thread'

[Steven]

Could you share your plugin list? Or your forum url in DM so I can check?

Do you have something helpful in [your_forum]/logs? A list of errors in [your_forum]/sidekiq?

We don’t have a lot of informations to help you find the issue

[L30110]

To start with, here is the (very short) plugin list. These are all the plugins listed in app.yml:

          - git clone https://github.com/discourse/docker_manager.git
          - git clone https://github.com/discourse/discourse-push-notifications.git
          - git clone https://github.com/davidtaylorhq/discourse-whos-online.git

[JammyDodger]

I think the Who’s Online plugin was swopped to official in September. You could try using the updated repository and see if that helps?

https://github.com/discourse/discourse-whos-online

[L30110]

Update on this. Thanks everybody for your suggestions. The core problems (except for one) have apparently (more testing to do) been resolved. The obvious one remaining is that ssl certs are no longer renewing, and even the manual renew process recommended here doesn’t work. It appears that Let’s Encrypt is unable to access the validation file here from the “well known” directory. This had been working fine for almost 3 years previously. So this is still a problem, especially since HSTS blocks Chrome entirely in this situation.

[Steven]

Seems like a mess.

You could try running a discourse-doctor to see if it gives any idea. I have a few but without access to the server, it would take too much time from someone in my modest level to help you.

I probably would reinstall everything if I was in your shoes

1 Backup

cd /var/discourse
./launcher enter app
discourse backup

Retrieve the backup via sftp (sftp protocol, port 22, host : ip of your server, login : root, password : root password)

Make sure you have the uploads folder in the backup. If you don’t, make a local copy of the folder in your hardrive.

2 Delete, reinstall your vps, or use another vps

Use the discourse setup install, if everything is ok: log in, upload and restore a backup

(you can also restore the backup from command line, it’s quicker : Restore a backup from command line)

If you’re not sure about this, it’s best to wait from someone more qualified

[L30110]

Thanks. I don’t think I want to go that route just to solve the ssl problem. There must be a narrower solution, given that the other problems seem to have been resolved. But again, thanks.

[L30110]

So in /var/discourse/shared/standalone/ssl there appear to be newly written keys and certs from my last attempt. But the log clearly shows a validation error and the site isn’t actually using them. Also – maybe this is just cosmetic – the logs are not showing the correct paths in this case. For example it says installing key to:

/shared/ssl/[site]_ecc.key

But the actual path is /shared/standalone/ssl

Which is at least odd looking.

The typical error is:

[site]:Verify error:Fetching http://[site]/.well-known/acme-challenge/[challenge string]: Error getting validation data

I should note that the site requires login for all user access, but this has not been a problem for ssl cert renews (during the several years of operation).

[L30110]

Because other issues have been resolved, I’m moving the ssl question to a separate thread. Thanks all!