The Calico datastore

Calico stores the data about the operational and configuration state of your cluster in a central datastore. If the datastore is unavailable your Calico network continues operating, but cannot be updated (no new pods can be networked, no policy changes can be applied, etc.).

Calico has two datastore drivers you can choose from

  • etcd - for direct connection to an etcd cluster
  • Kubernetes - for connection to a Kubernetes API server

Using Kubernetes as the datastore

This guide uses the Kubernetes API datastore driver. The advantages of this driver when using Calico on Kubernetes are

  • Doesn’t require an extra datastore, so is simpler to manage
  • You can use Kubernetes RBAC to control access to Calico resources
  • You can use Kubernetes audit logging to generate audit logs of changes to Calico resources

For completeness, the advantages of the etcd driver are

  • Allows you to run Calico on non-Kubernetes platforms (e.g. OpenStack)
  • Allows separation of concerns between Kubernetes and Calico resources, for example allowing you to scale the datastores independently
  • Allows you to run a Calico cluster that contains more than just a single Kubernetes cluster, for example, bare metal servers with Calico host protection interworking with a Kubernetes cluster; or multiple Kubernetes clusters.

Custom Resources

When using the Kubernetes API datastore driver, most Calico resources are stored as Kubernetes custom resources.

A few Calico resources are not stored as custom resources and instead are backed by corresponding native Kubernetes resources. For example, workload endpoints are Kubernetes pods.

In order to use Kubernetes as the Calico datastore, we need to define the custom resources Calico uses.

Download and examine the list of Calico custom resource definitions, and open it in a file editor.

wget https://docs.projectcalico.org/v3.10/manifests/crds.yaml

Create the custom resource definitions in Kubernetes.

kubectl apply -f crds.yaml

calicoctl

To interact directly with the Calico datastore, use the calicoctl client tool.

Install

  1. Download the calicoctl binary to a Linux host with access to Kubernetes.

    wget https://github.com/projectcalico/calicoctl/releases/download/v3.8.0/calicoctl
    chmod +x calicoctl
    sudo mv calicoctl /usr/local/bin/
    
  2. Configure calicoctl to access Kubernetes.

    export KUBECONFIG=/path/to/your/kubeconfig
    export DATASTORE_TYPE=kubernetes
    

    On most systems, kubeconfig is located at ~/.kube/config. You may wish to add the export lines to your ~/.bashrc so they will persist when you log in next time.

Test

Verify calicoctl can reach your datastore by running

calicoctl get nodes

You should see output similar to

NAME
ip-172-31-37-123
ip-172-31-40-217
ip-172-31-40-30
ip-172-31-42-47
ip-172-31-45-29

Nodes are backed by the Kubernetes node object, so you should see names that match kubectl get nodes.

Try to get an object backed by a custom resource

calicoctl get ippools

You should see an empty result

NAME   CIDR   SELECTOR


Next

Configure IP pools