Prometheus disk full alerts needed for prometheus-federation cluster

[stephen-soltesz]

The PlatformCluster_PrometheusPersistentDiskTooFull & PlatformCluster_DataPartitionTooFull alerts were migrated to the k8s-support configuration for prometheus.

https://github.com/m-lab/k8s-support/blob/master/config/prometheus/alerts.yml#L594

These alerts are no longer present in the prometheus-federation config but we need this alert for that cluster as well as the data-processing clusters.