Home

Awesome

Deploying a MongoDB database replica set across clusters

This tutorial demonstrates how to share a MongoDB database across multiple Kubernetes clusters that are located in different public and private cloud providers.

In this tutorial, you will deploy a three-member MongoDB replica set in which each member is located in its own cluster. You will also create a Virtual Application Nework for the servers, which will enable the members to form the replica set and communicate with each other.

To complete this tutorial, do the following:

Prerequisites

The basis for the demonstration is to depict the operation of a MongoDB replica set across distributed clusters. You should have access to three independent clusters to operate and observe the distribution of services over a Virtual Application Network. As an example, the three cluster might be comprised of:

While the detailed steps are not included here, this demonstration can alternatively be performed with three separate namespaces on a single cluster.

Step 1: Set up the demo

  1. On your local machine, make a directory for this tutorial and clone the example repo into it:

    mkdir mongodb-demo
cd mongodb-demo
git clone https://github.com/skupperproject/skupper-example-mongodb-replica-set.git

  2. Prepare the target clusters.

    1. On your local machine, log in to each cluster in a separate terminal session.
    2. In each cluster, create a namespace to use for the demo.
    3. In each cluster, set the kubectl config context to use the demo namespace (see kubectl cheat sheet)

Step 2: Deploy the Virtual Application Network

On each cluster, using the skupper tool, define the Virtual Application Network and the connectivity for the peer clusters.

  1. In the terminal for the first public cluster, deploy the public1 application router. Create a connection token for connections from the public2 cluster and the private1 cluster:

    skupper init --site-name public1
skupper token create public1-token.yaml --uses 2

  2. In the terminal for the second public cluster, deploy the public2 application router. Create a connection token for connections from the private1 cluser and connect to the public1 cluster:

    skupper init --site-name public2
skupper token create public2-token.yaml
skupper link create public1-token.yaml

  3. In the terminal for the private cluster, deploy the private1 application router. Connect to the public1 and public2 clusters;

    skupper init --site-name private1
skupper link create public1-token.yaml
skupper link create public2-token.yaml

Step 3: Deploy the MongoDB servers

After creating the Skupper network, deploy the servers for the three-member MongoDB replica set. The member in the private cloud will be designated as the primary, and the members on the public cloud clusters will be redundant backups.

  1. In the terminal for the private1 cluster, deploy the primary MongoDB member:

    kubectl apply -f ~/mongodb-demo/skupper-example-mongodb-replica-set/deployment-mongo-a.yaml

  2. In the terminal for the public1 cluster, deploy the first backup MongoDB member:

    kubectl apply -f ~/mongodb-demo/skupper-example-mongodb-replica-set/deployment-mongo-b.yaml

  3. In the terminal for the public2 cluster, deploy the second backup MongoDB member:

    kubectl apply -f ~/mongodb-demo/skupper-example-mongodb-replica-set/deployment-mongo-c.yaml

Step 4: Create Skupper services for the Virtual Application Network

  1. In the terminal for the private1 cluster, create the mongo-a service:

    skupper service create mongo-a 27017

  2. In the terminal for the public1 cluster, create the mongo-b service:

    skupper service create mongo-b 27017

  3. In the terminal for the public2 cluster, create the mongo-c service:

    skupper service create mongo-c 27017

  4. In each of the cluster terminals, verify the services created are present:

    skupper service status

    Note that the mapping for the service address defaults to tcp.

Step 5: Bind the Skupper services to the deployment targets on the Virtual Application Network

  1. In the terminal for the private1 cluster, expose the mongo-a deployment:

    skupper service bind mongo-a deployment mongo-a

  2. In the terminal for the public1 cluster, annotate the mongo-b deployment:

    skupper service bind mongo-b deployment mongo-b

  3. In the terminal for the public2 cluster, annotate the mongo-c deployment:

    skupper service bind mongo-c deployment mongo-c

  4. In each of the cluster terminals, verify the services bind to the targets

    skupper service status

    Note that each cluster depicts the target it provides.

Step 6: Form the MongoDB replica set

After deploying the MongoDB members into the private and public cloud clusters, form them into a replica set. The application router network connects the members and enables them to form the replica set even though they are running in separate clusters.

  1. In the terminal for the private1 cluser, use the mongo shell to connect to the mongo-a instance and initiate the member set formation:

    1.1. Use this if you have the mongo (command-line tool) installed and you are running your private1 site locally

    $ cd ~/mongodb-demo/skupper-example-mongodb-replica-set
$ mongo --host $(kubectl get service mongo-a -o=jsonpath='{.spec.clusterIP}')
> load("replica.js")

    1.2. Alternatively you can initiate the member set running the mongo command-line tool inside your running pod

    $ kubectl exec -it deploy/mongo-a -- mongo --host mongo-a
> rs.initiate( {
     _id : "rs0",
     members: [
        { _id: 0, host: "mongo-a:27017" },
        { _id: 1, host: "mongo-b:27017" },
        { _id: 2, host: "mongo-c:27017" }
     ]
  })

  2. Verify the status of the members array.

    > rs.status()

Step 7: Insert documents and observe replication

Now that the MongoDB members have formed a replica set and are connected by the application router network, you can insert some documents on the primary member, and see them replicated to the backup members.

  1. While staying connected to the mongo-a shell, insert some documents:

    > use test
> for (i=0; i<1000; i++) {db.coll.insert({count: i})}
# make sure the docs are there:
> db.coll.count()

  2. Using the mongo shell, check the first backup member to verify that it has a copy of the documents that you inserted:

    $ kubectl exec -it deploy/mongo-a -- mongo --host mongo-b

    > use test
> db.setSecondaryOk()
> db.coll.count()
> db.coll.find()

  3. Using the mongo shell, check the second backup member to verify that it also has a copy of the documents that you inserted.

    $ kubectl exec -it deploy/mongo-a -- mongo --host mongo-c

    > use test
> db.setSecondaryOk()
> db.coll.count()
> db.coll.find()

Cleaning up

Restore your cluster environment by returning the resource created in the demonstration. On each cluster, delete the demo resources and the skupper network:

  1. In the terminal for the private1 cluster, delete the resources:

    $ skupper unexpose deployment mongo-a
$ kubectl delete -f ~/mongodb-demo/skupper-example-mongodb-replica-set/deployment-mongo-a.yaml
$ skupper delete

  2. In the terminal for the public1 cluster, delete the resources:

    $ skupper unexpose deployment mongo-b
$ kubectl delete -f ~/mongodb-demo/skupper-example-mongodb-replica-set/deployment-mongo-b.yaml
$ skupper delete

  3. In the terminal for the public2 cluster, delete the resources:

    $ skupper unexpose deployment mongo-c
$ kubectl delete -f ~/mongodb-demo/skupper-example-mongodb-replica-set/deployment-mongo-c.yaml
$ skupper delete

Next Steps