vaultpal

README

vaultpal 🖖

{ your pal for using vault }

At DB Schenker, we use HashiCorp Vault extensively to automate access to secrets (e.g. application credentials), and systems such as Kubernetes or AWS.

Even though the official Vault CLI and the HTTP API provide the necessary means to access all Vault features programmatically, their usage can become overly complex for end-users who simply want to manage the accesses they need for their daily work.

This is where vaultpal comes into play: It wraps common access management functions into higher level (sub)commands that don't require deep Vault know-how or excessive configuration. As an example, vaultpal can empower users to obtain temporary, limited-privilege credentials from the AWS Security Token Service (STS) through HashiCorp Vault with a single command call.

Releases

Check the releases section for the most recent binaries that are suitable for your operating system.

Please note that vaultpal has been thoroughly tested on macOS (*darwin binaries) and Linux, but not on Windows. The binary should execute without issues, but there may be subtle differences, e.g. in the handling of file locations. Alternatively, consider Windows Subsystem for Linux.

Setup

1. Extract the go binary, place it on your $PATH and make it executable.

     echo $PATH
     chmod a+x /path/to/vaultpal_binary
     cp /path/to/vaultpal_binary /element/of/your/$PATH
   

   Please note that $PATH output will look like this:

     /Users/your_user/.local/bin:/Users/your_user/.pyenv/plugins/pyenv-virtualenv/shims:
   

   That is, each element is separated from others by : . So in this example: /element/of/your/$PATH could be: /Users/your_user/.local/bin . You may need to use this command with sudo.

2. Check the installation with:

    vaultpal version
   

3. (optional) install completion (bash/zsh).

   echo 'source <(vaultpal completion bash)' >>~/.bashrc
   

4. (optional) define an alias

   echo 'alias vp=vaultpal' >>~/.bashrc
   echo 'complete -F __start_vaultpal vp' >>~/.bashrc
   

Docker Container Images

Released vaultpal versions are build for multiple architectures and pushed to the public GitHub Container Registry (https://ghcr.io).

$ docker pull ghcr.io/dbschenker/vaultpal:1.6.0
$ docker run --rm dbschenker/vaultpal:1.6.0 version
v1.6.0 (commit: 319f6c3)

Usage

• perform login with vault cli since vaultpal relies on a valid existing token, e.g.:

  vault login -method=oidc
  

• just launch without arguments to get an overview of available commands and flags

   vaultpal
         { vault~Pal 👍 }
        °- (🕶 ) v1.5.2 -°
  
  vault~Pal 👍 will help you using vault in your daily work.
  
  Usage:
    vaultpal [flags]
    vaultpal [command]
  
  Available Commands:
    completion  Generate shell completion scripts
    export      Export various types of resources to shell
    help        Help about any command
    switch      Switch between roles
    timer       Display the remaining TTL of your vault token
    version     Print version of vaultpal
    write       Write various types of resources
  
  Flags:
        --config string      config file (default is $HOME/.vaultpal.yaml)
    -h, --help               help for vaultpal
    -v, --verbosity string   Log level (debug, info, warn, error, fatal, panic (default "info")
  
  Use "vaultpal [command] --help" for more information about a command.
  

Features

Kubeconfig

1. Call vaultpal to create a kubeconfig file

   vaultpal write kubeconfig sandbox master
   

2. vaultpal will write kubeconfig file in home directory:

   ~/.vaultpal/kube/config
   

3. Enable usage of kubeconfig file with:

   export KUBECONFIG=~/.vaultpal/kube/config
   

4. Note that vaultpal will store a kubeconfig for each cluster with the cluster name as context name. This enables the usage of different clusters at the same time

Switch Role

1. Call vaultpal to switch to a token role

   vaultpal switch role k8s-admin
   

2. vaultpal will create a new token for given role and write it to vault token file
3. Use vault cli with the role token

Export AWS STS Credentials

1. Use vaultpal to create AWS STS credentials with vault

   vaultpal export awssts mytopic-prod-admin
   

2. vaultpal will use vault aws secret engine to create AWS STS credentials. The default secret engine path is "aws"
3. The credentials will be printed as bash export commands.

Use Alias function

vaultpal provides a bash alias function to wrap the vaultpal command with direct export of the credentials to the current shell.

1. Get the alias function

   vaultpal export awssts -a
   

2. Or use it directly in alias definition, e.g.:

   alias vpalsts="$(vaultpal export awssts -a)"
   

   Then you will get the credentials exported to current shell and can use it directly, e.g.:

   vpalsts mytopic-prod-admin
   

Export AWS web console URL to shell

1. Use vaultpal to create temporary sign-in URLs to access the AWS Web Console with a single click

   vaultpal -v warn export awsconsole myapp-prod-admin
   https://signin.aws.amazon.com/federation?Action=login&Issuer=https://(...)
   

Configuration

The following section describes central configurations, that are required for vaultpal kubeconfig functions

Kubeconfig

In order to render kubeconfig files, vaultpal requires meta information about the kubernetes cluster. Therefore, a cluster configuration object must be stored in vault providing the required information. The configuration object must be stored in a kv secret engine version 2 at mount path "kv". The configuration must be accessible for all vaultpal users

Example: Configuration for a kubernetes cluster called "bibi" must be stored at vault path kv/vaultpal/k8s/clusters/bibi with data:

{
  "name":   "bibi",
  "pki":    "k8s-bibi-pki-kube",
  "server": "https://api.bibi.mytopic.com"
}

Cluster Alias

vaultpal supports the definition of an alias to a kubernetes cluster. This is useful if you want to use a generic endpoint like "int" or "prod" pointing to a cluster.

Example: Configuration for an alias named "int" pointing to a kubernetes cluster called "bibi" must be stored at vault path kv/vaultpal/k8s/clusters/int with data:

{
  "name":   "int",
  "alias":  "bibi",
  "server": "https://api.int.mytopic.com"
}

Based on the alias value "bibi", vaultpal will read the configuration for cluster "bibi" in order to render the required certs and keys (pki).

Environment Variables

Variable	Usage
VAULTPAL_NP_URL	URL of Vault production environment, used for prompt label
VAULTPAL_PR_URL	URL of Vault non-production environment, used for prompt label
VAULTPAL_KUBECONFIG_FILE	Custom location of kubeconfig file

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss your idea.

License

MIT