For a Production setup we had to configure an nginx container plus the FPM container, let em talk, Configure some more Commands to enable/disable xdebug (that requires a restart of the containers since it needs a php estension) and i think that’s all. Other langs we use are literally “run the app” then kube will handle the rest.
Yw, not really hard but since we’re not a php shop it required a lot of research, trial and error, complexity as in “need to know something not related to Application or domain” and ops are unhappy at having to support this one of a kind setup. There might be a better way tho