Update control-plane.md to ensure it goes beyond API, and align with …

…new cluster definition etc.

content/en/docs/reference/glossary/control-plane.md

@@ -4,22 +4,26 @@ id: control-plane
 date: 2019-05-12
 full_link:
 short_description: >
-  The container orchestration layer that exposes the API and interfaces to define, deploy, and manage the lifecycle of containers.
+ The control plane provides API/Interface driven management of the lifecycles and service levels (performance, scale, availability etc,) of the containerized workloads that run on the cluster, and of the cluster itself. 
 
 aka:
 tags:
 - fundamental
 ---
- The container orchestration layer that exposes the API and interfaces to define, deploy, and manage the lifecycle of containers.
-
+ The control plane provides API/Interface driven management of the lifecycles and service levels (performance, scale, availability etc,) of the containerized {{< glossary_tooltip text="workloads" term_id="workload" >}} that run on the {{< glossary_tooltip text="cluster" term_id="cluster" >}}, and of the cluster itself. 
  <!--more--> 
 
- This layer is composed by many different components, such as (but not restricted to):
+The control plane is composed by many different components, including (but not restricted to):
 
  * {{< glossary_tooltip text="etcd" term_id="etcd" >}}
  * {{< glossary_tooltip text="API Server" term_id="kube-apiserver" >}}
  * {{< glossary_tooltip text="Scheduler" term_id="kube-scheduler" >}}
  * {{< glossary_tooltip text="Controller Manager" term_id="kube-controller-manager" >}}
  * {{< glossary_tooltip text="Cloud Controller Manager" term_id="cloud-controller-manager" >}}
 
- These components can be run as traditional operating system services (daemons) or as containers. The hosts running these components were historically called {{< glossary_tooltip text="masters" term_id="master" >}}.
+ These components can be run as traditional operating system services (daemons), as containers or within pods.  They may also be distributed across multiple machines for scale and/or resilience. Components of the control plane may also be co-located with containerized workloads, on the same machines, but this is *not* a best practice in production environments.
+
+  The hosts running these components were historically called {{< glossary_tooltip text="masters" term_id="master" >}}.
+
+
+