Switching from apache2 to nginx

hello… new to rt and sys adm stuffs.
Current RT setup use apache2 as web server and it uses

ScriptAlias / /opt/rt5/sbin/rt-server.fcgi/

I was assigned to use nginx instead and find out if it does provide better performance.
I’m reading through RT web deployment and manual installation from RT wiki. I want to use nginx as web server. It says that nginx requires that you start RT’s fastcgi process externally.
Here’s are the options I know of.
i) using spawn-fcgi as in web deployment RT doc
ii) using multiwatch as systemd service/socket as explained in rt.wiki manual installation.
iii) spawn-fcgi -s /tmp/fastcgi-rails.sock -n – /usr/bin/multiwatch -f 2 /home/rails/public/dispatch.fcgi
I want to know whats is preferred method and how to get best performance. Like when using multiwatch, I can spawn multiple instances but I don’t know how does it helps to provide better performance… Every advice is much appreciated. If anything confuses you… just ask.

I recommend systemd socket-activation with multiwatch.

It works well here, our setup also integrates podman so that multiwatch & RT itself are inside a container.

1 Like