Tango is a client and a server. The server runs in the cluster that Tango works with and is part of the "control plane" (TODO: fill in more about the control plane later), and you install it (and the rest of the control plane) with the CLI using tango install
.
The CLI talks to the control plane, and you can configure it by doing one of these:
- Using a Kubernetes config file (the same one that
kubectl
uses) - with this method,tango
will create a tunneling proxy
This is similar to the method that Helm version 2 uses to communicate with Tiller, its server-side component
- Connecting to a public endpoint that the Tango server listens on
By default,
tango install
will install the server so that it listens on a public endpoint, so that othertango
users don't need to have a Kubernetes configuration on their machine at all