Build from Dockerfile.php Contains a Laravel app that displays the page. Cache headers are set in routes/web.php and can be controlled programtically if its needed.
Build from Dockerfile.nginx. Nginx container used to serve the requests to PHP-FMP, and add static content cache header.
Config can be seen in "deploy/nginx/vhost.conf"
Build from Dockerfile.cache. Nginx cache proxy. This can be used inside the cluster to cache responses from BE, it also can be deployed outisde of the cluster and be used as edge node.
Config can be seen in "deploy/cache/cache.conf"
$ docker-compose up -d- On port 8080 - traffic is exposed via the proxy.
- On port 8081 - directly call the backend service.
Located in "deploy/helm/lorem"
$ helm install $APP_NAME deploy/helm/lorem/ # Installs the app
$ helm show values deploy/helm/lorem/ # Shows variablesJust an example deployment Jenkins pipeline ( Not tested :) )
The application is separated in two pods - one with PHP-FPM+Nginx ( called backend ), and one Cache proxy.
This give us the abillity to scale different parts of the app separately. For example if we have lots of requests on the same data - we can scale only cache proxy to handle the requests, otherwise if we have heavy MEM/CPU operations on the App - we can scale the application deployment.
Helm chart do not contain Autoscalers, but they can be added separately.