Remove support for component-prometheus based monitoring

We've ended up never using component-prometheus on OpenShift 4, and have
changed our strategy to build OpenShift 4 monitoring on the
cluster-monitoring and user-workload-monitoring stacks provided by
OpenShift 4. Therefore the logic to enable a component-prometheus
monitoring stack to scrape metrics from namespaces/workloads managed
through this component is no longer necessary.

This commit completely removes support for component-prometheus in this
component.

class/defaults.yml

@@ -9,10 +9,6 @@ parameters:
     defaultSourceNamespace: openshift-marketplace
     useCustomNamespace: false
 
-    monitoring:
-      enabled: true
-      instance: null
-
   openshift_operators:
     defaultSource: certified-operators
 

component/main.jsonnet

@@ -30,4 +30,4 @@ local namespace = operatorlib.validateInstance(params.namespace);
       },
     },
   ],
-} + (import 'monitoring.libsonnet')
+}

docs/modules/ROOT/pages/references/parameters.adoc

@@ -71,31 +71,3 @@ default:: `false`
 
 Whether to allow non-standard namespace names for the OperatorGroup.
 If this parameter is set to `true`, the component will allow using namespace names other than `openshift-operators` and `openshift-operators-redhat`.
-
-== `monitoring`
-
-This parameter allows users to enable the component's monitoring configuration.
-Currently the component has support for deploying custom `ServiceMonitors` on clusters which use component `prometheus` to manage a custom monitoring stack.
-
-[WARNING]
-The monitoring configuration is instance independent. It must not be changed between instances of this component.
-
-=== `enabled`
-
-[horizontal]
-type:: boolean
-default:: `true`
-
-Whether to deploy monitoring configurations.
-If this parameter is set to `true`, the component will check whether component `prometheus` is present on the cluster.
-If the component is missing, no configurations will be deployed regardless of the value of this parameter.
-
-=== `instance`
-
-[horizontal]
-type:: string
-default:: `null`
-
-This parameter can be used to indicate which custom Prometheus instance should pick up the configurations managed by the component.
-
-If the parameter is set to the empty string, the default instance configured for component `prometheus` will be used.