About the community charts#
The chart depends on two charts that other projects maintain.
The PostgreSQL cluster comes from the cluster chart of the CloudNativePG project, and the Varnish cache comes from the varnish-cache chart of Varnish Software.
The chart installs them as dependencies, and it configures them through the postgresql and varnish values.
Why depend on them#
CloudNativePG is an operator, which runs and repairs PostgreSQL clusters in Kubernetes.
Its maintainers know the database, the failover, and the backups better than this chart can, and they release fixes for it.
The cluster chart is the way the CloudNativePG project describes a cluster, so the chart doesn’t need its own templates for one.
The varnish-cache chart is maintained by Varnish Software, the company behind Varnish.
The chart uses its Deployment, Service, and VCL handling, and only adds the defaults that Plone needs.
Depending on these charts means the chart does less of the work of maintaining them. It also means the chart has to follow their conventions, for the values and for the names of their resources.
What it means for you#
The options of each dependency are available under its key, such as postgresql.cluster.instances and varnish.server.replicas.
The chart’s defaults, and the names of the resources they create, follow the dependencies.
See Naming conventions for the names.
The CloudNativePG operator is not part of the chart. It must be installed in the cluster before the chart, once for the whole cluster, because it needs cluster-wide permissions. See How to install the CloudNativePG operator.
The chart replaces the subcharts of the previous charts of this repository, which had their own templates for the backend, the frontend, the ZEO server, nginx, Varnish, and PostgreSQL. See How to migrate from the previous charts.