Configure a multi-region Mission Control environment

You can configure a multi-region Mission Control environment to deploy Mission Control across two or more Kubernetes clusters in multiple regions and datacenters. For example, you might host the Mission Control control plane in one region and the data plane in another.

A multi-region environment offers several benefits, including better data locality, increased fault tolerance, improved disaster recovery, and higher availability for distributed applications.

Prerequisites

Two or more Kubernetes clusters running in separate regions
An understanding of the planning considerations for deploying Mission Control
An understanding of Kubernetes concepts and how to use kubectl

Configure network connectivity between multiple Kubernetes clusters for Mission Control

Before you can configure a multi-region Mission Control environment, you must establish network connectivity between the control plane and data plane instances running in different Kubernetes clusters.

This guide uses two clusters running in two different regions, CP_EAST and DP_WEST, which are registered as contexts in your local kubeconfig file.

Verify the contexts for your two Kubernetes clusters:
```
kubectl config get-contexts
```
Sample result
CURRENT NAME CLUSTER AUTHINFO NAMESPACE * **CP_EAST** **CP_EAST** **CP_EAST** default **DP_WEST** **DP_WEST** **DP_WEST** default
In this example, the CP_EAST Kubernetes cluster is the Mission Control control plane and also serves as a data plane, hosting database datacenters. The DP_WEST Kubernetes cluster is a Mission Control data plane, which hosts datacenters but is controlled by the control plane of the CP_EAST cluster.
Ensure the following ports are open between the two Kubernetes clusters to allow communication between the control plane and the data plane:
- tcp:7000: Internode communications. This port is used for direct node-to-node communication within the Mission Control infrastructure. It facilitates coordination between nodes in different clusters, ensuring data consistency and system reliability.
- tcp:7001: Encrypted internode communications. Similar to port 7000, this port is used for internode communication, but with encryption enabled. This ensures secure data transfer between the control plane and data plane, protecting sensitive information from interception.
- tcp:8080: Management API. This port allows access to the management API for configuring Mission Control. It allows administrators to perform operations such as scaling, monitoring, and managing resources.
- tcp:30600: Vector aggregator. The vector aggregator collects and processes telemetry data from different nodes. This port is essential for log aggregation, monitoring, and analytics, providing visibility into the health and performance of the system. A vector aggregator runs in each data plane datacenter. Each instance aggregates data locally and then forwards logs and metrics to the control plane aggregator.

Now that you have verified your contexts and opened the necessary ports for your existing Kubernetes clusters, you can install the Mission Control control plane and data planes.

Install the Mission Control control plane and data planes

Select your Mission Control installation method. You can either Bring your own Kubernetes cluster, use the embedded Kubernetes cluster, or use Helm to install Mission Control in your Kubernetes clusters.
Change the context to the CP_EAST cluster:
```
kubectl config use-context CP_EAST
```
Replace CP_EAST with the name of the context for your control plane cluster.

Install Mission Control in control plane mode in the CP_EAST cluster.

You must specify the URL of the Vector Aggregator service in the control plane. The control plane cluster exposes this service on all nodes on port 30600. Specify a hostname that can load balance between all nodes in the control plane cluster.

Change the context to the DP_WEST cluster:
```
kubectl config use-context DP_WEST
```
Replace DP_WEST with the name of the context for the data plane cluster.
Install Mission Control in data plane mode in the DP_WEST cluster.
Use the Mission Control CLI, mcctl, to register the data plane to the control plane:
```
mcctl register --source-context CP_EAST --dest-context DP_WEST
```
Replace the following:
- CP_EAST: The context of the control plane cluster
- DP_WEST: The context of the data plane cluster
  
  The Mission Control control plane recognizes the registered data plane. This allows you to manage resources across the two planes.
Optional: Validate the connection between the control plane and data plane by checking the clientconfig and secret resources in the mission-control namespace.
1. Check the clientconfig resource for the registered data plane:
  kubectl get -n mission-control clientconfig
  Result
  Returns a list of registered data planes.
  
  NAME AGE DP_WEST-dataplane1 392d
2. Check the secret resource for the registered data plane:
  kubectl get -n mission-control secret
  Result
  Returns a list of registered data planes.
  
  NAME TYPE DATA AGE DP_WEST-dataplane1-config Opaque 1 392d

You can also verify that the data plane context registration in the Mission Control UI. To do this, check the Data Plane Context field in the cluster creation and modification forms.

Next, create a multi-datacenter Mission Control cluster environment spanning both Kubernetes clusters in your two regions.

Create a multi-datacenter cluster across Kubernetes clusters

The MissionControlCluster object defines the configuration for the Mission Control control plane and data planes.

To configure your environment, do the following:

Create a MissionControlCluster object that spans both Kubernetes clusters:

This example creates the Mission Control cluster using the CLI. The Mission Control UI uses the Data Plane Context field to specify the Kubernetes context where the datacenter should be created. If you don’t set a value for the Data Plane Context, the datacenter is created in the control plane context.

apiVersion: missioncontrol.datastax.com/v1beta1
kind: MissionControlCluster
metadata:
  name: PROJECT_NAME
  namespace: PROJECT_NAMESPACE
spec:
  createIssuer: true
  encryption:
    internodeEncryption:
      certs:
        createCerts: true
      enabled: true
  k8ssandra:
    auth: true
    cassandra:
      config:
        jvmOptions:
          gc: G1GC
          heapSize: 1Gi
      datacenters:
        - config:
            jvmOptions:
              gc: G1GC
          metadata:
            name: dc1
          racks:
            - name: r1
          size: 2
          stopped: false
        - config:
            jvmOptions:
              gc: G1GC
          k8sContext: CP_EAST
          metadata:
            name: dc2
          racks:
            - name: r1
          size: 2
          stopped: false
      resources:
        requests:
          cpu: 100m
          memory: 2Gi
      serverType: cassandra
      serverVersion: 4.0.11
      storageConfig:
        cassandraDataVolumeClaimSpec:
          accessModes:
            - ReadWriteOnce
          resources:
            requests:
              storage: 2Gi
          storageClassName: standard

Replace the following:

PROJECT_NAME: The name of the MissionControlCluster object
PROJECT_NAMESPACE: The namespace where the MissionControlCluster object is created
CP_EAST: The name of the context where the data plane is running

This sample MissionControlCluster object defines two datacenters, the first created in the control plane directly, and the other using the Kubernetes context of the data plane registered in the control plane. This configuration ensures a distributed architecture while maintaining centralized control.

Apply the MissionControlCluster object to create the multi-datacenter Mission Control cluster:
```
kubectl apply -f PROJECT_NAME.yaml
```
Replace PROJECT_NAME.yaml with the name of the MissionControlCluster object manifest file.

When the cluster starts, the cluster overview screen in the Mission Control UI displays the datacenters in each region:

Configure a multi-region Mission Control environment

Prerequisites

Configure network connectivity between multiple Kubernetes clusters for Mission Control

Install the Mission Control control plane and data planes

Create a multi-datacenter cluster across Kubernetes clusters

Next steps

Was this helpful?

Give Feedback