how_it_works.md

How it works

Kubeswitch consists of two components:

switch.sh - contains the shell function switch() which is the entry point to Kubeswitch. switcher - a go binary which is executed by the switch shell function and handles context selection and manipulation of the selected Kubeconfig file.

For a proper installation

• you have to source the switch.sh script
• make the switcher binary available in your $PATH (the switch.sh shell script looks for the switcher binary on the $PATH).

Flow

1. User executes $ switch (bash function available via the sourced switch.sh script)
2. The switch.sh script discovers and executes the switcher binary from the $PATH
3. The switcher binary searches for kubeconfigs in the kubeconfig stores configured in the SwitchConfig configuration file.
4. The switcher binary displays a fuzzy search for kubeconfig context names
5. The user selects on context name
6. The switcher binary creates a temporary copy of the selected kubeconfig file, sets the current-context and writes the kubeconfig to ~/.kube/switch_tmp
7. The switcher binary writes the filepath to the kubeconfig to STDOUT
8. The switch.sh script captures this filepath and executes export KUBECONFIG=</path/to/tmp/kubeconfig/file>

Each terminal window operates on its own copy of the kubeconfig file (terminal window isolation).