feat(network): enable internal ingress network policy

[andrewazores]

Related to #205

To test on OpenShift:

1. helm install --set core.route.enabled=true --set authentication.openshift.enabled=true cryostat ./charts/cryostat
2. Wait for the deployment to become ready.
3. oc run runner --image=registry.access.redhat.com/ubi8/ubi --rm -it /bin/bash. This runs an additional Pod in the same installation namespace, but one which is not part of Cryostat and does not have the same selector labels. Once this comes up and you have a shell, curl -v http://cryostat:8181 or curl -v http://cryostat-storage:8333. These should time out.
4. Open Route (echo https://$(oc get route -n cryostat cryostat -o jsonpath="{.status.ingress[0].host}")) and ensure Cryostat UI behaves as usual. Create a localhost:0 custom target, start and archive a recording, etc.
5. helm uninstall cryostat
6. helm install --set core.route.enabled=true --set authentication.openshift.enabled=true --set networkPolicy.ingress.enabled=false cryostat ./charts/cryostat
7. Wait for the deployment to become ready.
8. oc run runner --image=registry.access.redhat.com/ubi8/ubi --rm -it /bin/bash. Once this comes up and you have a shell, curl -v http://cryostat:8181 or curl -v http://cryostat-storage:8333. These should succeed and return HTTP responses quickly.
9. Repeat the test from step 4 to ensure Cryostat is working as expected from the user's POV.

Similar testing on other types of cluster (kind, minikube) should also work, with the usual adjustments (don't use core.route.enabled or authentication.openshift.enabled, etc.).

---

I also spent some time trying to define Egress policies. I thought this would be interesting along the same lines as cryostatio/cryostat-agent#242 and cryostatio/cryostat#323 - we could add network-level restrictions (firewall rules) to prevent Cryostat from being made to open network connections to unexpected destinations, ie. namespaces outside of the admin's chosen list of target namespaces. However, I ran into issues where this interrupted Cryostat's ability to connect to its own database deployment, and I was also not completely sure the right approach to allow Cryostat and the openshift-oauth-proxy to have traffic egress to the k8s API server (for doing RBAC checks, or Endpoints discovery). I'm holding off on that idea for now, but it may be worth following up with again later.

[ebaron]

When testing on OpenShift, I seem to be getting 503 errors when trying to access Cryostat through its Route. When the Network Policy is disabled, the route works fine. Not sure why