Installing and Configuring calicoctl

This document outlines how to install and configure calicoctl which is the primary tool for viewing, manipulating, and creating Calico objects on the command line.

Where to run calicoctl

Calicoctl’s primary function is to read or manipulate state stored in the datastore. As such, it can run from any host with network access to the datastore. There are also the node sub-commands of calicoctl that are for starting and checking the calico/node container. To use this functionality the calicoctl tool must be run on the host where the container will run or is running.

Installing calicoctl

The calicoctl tool can be downloaded from the release page of the calicoctl repository, set executable, and then it is ready to run.

curl -O -L https://github.com/projectcalico/calicoctl/releases/download/v1.6.5/calicoctl
chmod +x calicoctl

Note: Move calicoctl to a directory in your PATH or add the directory it is in to your PATH to avoid prepending the path to invocations of calicoctl.

Datastore configuration

Datastore configuration may be as simple as using the defaults but in most cases the endpoint will need to be specified and possibly other settings too, all which depend on how your datastore is configured.

Here is a simple etcdv2 example.

ETCD_ENDPOINTS=http://etcd:2379 calicoctl get nodes

Here is a simple Kubernetes datastore example.

DATASTORE_TYPE=kubernetes KUBECONFIG=~/.kube/config calicoctl get nodes

For the possible options and configuration guidance see Configuring calicoctl.

Checking the configuration

Here is a simple command to check that the installation and configuration is correct.

calicoctl get nodes

A correct setup will yield a list of the nodes that have registered. If an empty list is returned you are either pointed at the wrong datastore or no nodes have registred. If an error is returned then attempt to correct the issue then try again.

Next steps

Now you are ready to read and configure most aspects of Calico. You can find the full list of commands in the Command Reference.

The full list of resources that can be managed, including a description of each, can be found in the Resource Definitions.