Navigation Menu

Skip to content

Latest commit

 

History

History
451 lines (305 loc) · 15.1 KB

index.mdx

File metadata and controls

451 lines (305 loc) · 15.1 KB
Raw
layout page_title description
docs
Commands (CLI)
In addition to a verbose HTTP API, Vault features a command-line interface that wraps common functionality and formats output. The Vault CLI is a single static binary. It is a thin wrapper around the HTTP API. Every CLI command maps directly to the HTTP API internally.

Vault commands (CLI)

~> Note: The Vault command-line interface (CLI) changed substantially in 0.9.2+ and may cause confusion while using older versions of Vault with this documentation. Read our upgrade guide for more information.

In addition to a verbose HTTP API, Vault features a command-line interface (CLI) that wraps common functionality and formats output. The Vault CLI is a single static binary. It is a thin wrapper around the HTTP API. Every CLI command maps directly to the HTTP API internally.

CLI command structure

Each command is represented as a command or subcommand, and there are a number of command and subcommand options available: HTTP options, output options, and command-specific options.

Construct your Vault CLI command such that the command options precede its path and arguments if any:

vault <command> [options] [path] [args]
  • options - Flags to specify additional settings
  • args - API arguments specific to the operation

-> NOTE: Use the command help to display available options and arguments.

Examples:

The following write command creates a new user (bob) in the userpass auth method. It passes the -address flag to specify the Vault server address which precedes the path (auth/userpass/users/bob) and its argument (password="long-password") at last.

$ vault write -address="http://127.0.0.1:8200" auth/userpass/users/bob password="long-password"

If multiple options (-address and -namespace) and arguments (password and policies) are specified, the command would look like:

$ vault write -address="http://127.0.0.1:8200" -namespace="my-organization" \
        auth/userpass/users/bob password="long-password" policies="admin"

The options (flags) come after the command (or subcommand) preceding the path, and the args always follow the path to set API parameter values.

The four most common operations in Vault are read, write, delete, and list. These operations work on most paths in Vault. Some paths will contain secrets while other paths may contain configuration. Whatever it is, the primary interface for reading and writing data to Vault is similar.

Print cURL command

To see the equivalent API call to perform the same operation, use the -output-curl-string flag after the subcommand.

$ vault write -output-curl-string  auth/userpass/users/bob password="long-password"

curl -X PUT -H "X-Vault-Request: true" -H "X-Vault-Token: $(vault print token)" -d '{"password":"long-password"}' http://127.0.0.1:8200/v1/auth/userpass/users/bob

Print policy requirements

To view the policy requirements to perform an operation, use the -output-policy flag after the subcommand.

$ vault kv put -output-policy kv/secret value=itsasecret

path "kv/data/secret" {
  capabilities = ["create", "update"]
}

Command help

There are two primary ways to get help in Vault: CLI help (help) and API help (path-help).

CLI help

Use help (or -h for shorthand) to see the CLI help output which corresponds to your Vault version.

To get CLI help:

$ vault help

Example: To get help on the kv command.

$ vault kv help

The help output displays available subcommands, parameters, and command flags.

API help

To invoke the Vault API paths, you can use the read (for HTTP GET), write (for HTTP PUT or POST), delete (for HTTP DELETE), and list (for HTTP LIST) commands.

Use path-help to get Vault API help:

$ vault path-help -h

The path-help retrieves API help on any API paths. Vault API paths provide built-in help in markdown format. This includes system paths, secret engines, and auth methods.

Example: API help on the sys/mounts/ path.

$ vault path-help sys/mounts
Request:        mounts
Matching Route: ^mounts$

List the currently mounted backends.

## DESCRIPTION

This path responds to the following HTTP methods.

    GET /
        Lists all the mounted secret backends.

    GET /<mount point>
        Get information about the mount at the specified path.

    POST /<mount point>
        Mount a new secret backend to the mount point in the URL.

    POST /<mount point>/tune
        Tune configuration parameters for the given mount point.

    DELETE /<mount point>
        Unmount the specified mount point.

The help output displays supported child-paths and available parameters if there are any.

Command input

To write data to Vault, the input can be a part of the command in key-value format.

$ vault kv put secret/password value=itsasecret

However, some Vault API require more advanced structures such as maps. You can use stdin or file input instead.

stdin

Some commands in Vault can read data from stdin using - as the value. If - is the entire argument, Vault expects to read a JSON object from stdin:

$ echo -n '{"value":"itsasecret"}' | vault kv put secret/password -

In addition to reading full JSON objects, Vault can read just a value from stdin:

$ echo -n "itsasecret" | vault kv put secret/password value=-

Files

Some commands can also read data from a file on disk. The usage is similar to stdin as documented above. If an argument starts with @, Vault will read it as a file:

$ vault kv put secret/password @data.json

Or specify the contents of a file as a value:

$ vault kv put secret/password value=@data.txt

To use values that begin with an @ character, escape the character with a backslash (\) to keep Vault from parsing the value as a file location:

$ vault login -method userpass username="mitchellh" password="\@foo"

Note that if an argument is supplied in a @key=value format, Vault will treat that as a kv pair with the key being @key, not a file called key=value. This also means that Vault does not support filenames with = in them.

Mount flag syntax (KV)

All kv commands can alternatively refer to the path to the KV secrets engine using a flag-based syntax like $ vault kv get -mount=secret password instead of $ vault kv get secret/password. The mount flag syntax was created to mitigate confusion caused by the fact that for KV v2 secrets, their full path (used in policies and raw API calls) actually contains a nested /data/ element (e.g. secret/data/password) which can be easily overlooked when using the above KV v1-like syntax secret/password. To avoid this confusion, all KV-specific docs pages will use the -mount flag.

Exit codes

The Vault CLI aims to be consistent and well-behaved unless documented otherwise.

  • Local errors such as incorrect flags, failed validations, or wrong numbers of arguments return an exit code of 1.

  • Any remote errors such as API failures, bad TLS, or incorrect API parameters return an exit status of 2

Some commands override this default where it makes sense. These commands document this anomaly.

Autocompletion

The vault command features opt-in autocompletion for flags, subcommands, and arguments (where supported).

Enable autocompletion by running:

$ vault -autocomplete-install

~> Be sure to restart your shell after installing autocompletion!

When you start typing a Vault command, press the <tab> character to show a list of available completions. Type -<tab> to show available flag completions.

If the VAULT_* environment variables are set, the autocompletion will automatically query the Vault server and return helpful argument suggestions.

Token helper

By default, the Vault CLI uses a "token helper" to cache the token after authentication. This is conceptually similar to how a website securely stores your session information as a cookie in the browser. Token helpers are customizable, and you can even build your own.

The default token helper stores the token in ~/.vault-token. You can delete this file at any time to "logout" of Vault.

Environment variables

The CLI reads the following environment variables to set behavioral defaults. This can alleviate the need to repetitively type a flag. Flags always take precedence over the environment variables. Each of the following environment variables must be set on the Vault process. In Vault 1.13+, all environment variables available to the Vault process will be logged during startup.

VAULT_TOKEN

Vault authentication token. Conceptually similar to a session token on a website, the VAULT_TOKEN environment variable holds the contents of the token. For more information, please see the token concepts page.

VAULT_ADDR

Address of the Vault server expressed as a URL and port, for example: https://127.0.0.1:8200/.

VAULT_CACERT

Path to a PEM-encoded CA certificate file on the local disk. This file is used to verify the Vault server's SSL certificate. This environment variable takes precedence over VAULT_CAPATH.

VAULT_CAPATH

Path to a directory of PEM-encoded CA certificate files on the local disk. These certificates are used to verify the Vault server's SSL certificate.

VAULT_CLIENT_CERT

Path to a PEM-encoded client certificate on the local disk. This file is used for TLS communication with the Vault server.

VAULT_CLIENT_KEY

Path to an unencrypted, PEM-encoded private key on disk which corresponds to the matching client certificate.

VAULT_CLIENT_TIMEOUT

Timeout variable. The default value is 60s.

VAULT_CLUSTER_ADDR

Address that should be used for other cluster members to connect to this node when in High Availability mode.

VAULT_FORMAT

Provide Vault output (read/status/write) in the specified format. Valid formats are "table", "json", or "yaml".

VAULT_LICENSE

[Enterprise, Server only] Specify a license to use for this node. This takes precedence over #VAULT_LICENSE_PATH and license_path in config.

VAULT_LICENSE_PATH

[Enterprise, Server only] Specify a path to a license on disk to use for this node. This takes precedence over license_path in config.

VAULT_LOG_LEVEL

Set the Vault server's log level. The supported values are trace, debug, info, warn, and err. The default log leve is info.

VAULT_MAX_RETRIES

Maximum number of retries when certain error codes are encountered. The default is 2, for three total attempts. Set this to 0 or less to disable retrying.

Error codes that are retried are 412 (client consistency requirement not satisfied) and all 5xx except for 501 (not implemented).

VAULT_REDIRECT_ADDR

Address that should be used when clients are redirected to this node when in High Availability mode.

VAULT_SKIP_VERIFY

Do not verify Vault's presented certificate before communicating with it. Setting this variable is not recommended and voids Vault's security model.

VAULT_TLS_SERVER_NAME

Name to use as the SNI host when connecting via TLS.

VAULT_CLI_NO_COLOR

If provided, Vault output will not include ANSI color escape sequence characters.

VAULT_RATE_LIMIT

This environment variable will limit the rate at which the vault command sends requests to Vault.

This environment variable has the format rate[:burst] (where items in [] are optional). If not specified, the burst value defaults to rate. Both rate and burst are specified in "operations per second". If the environment variable is not specified, then the rate and burst will be unlimited i.e. rate limiting is off by default.

Note: The rate is limited for each invocation of the vault CLI. Since each invocation of the vault CLI typically only makes a few requests, this environment variable is most useful when using the Go Vault client API.

VAULT_NAMESPACE

The namespace to use for the command. Setting this is not necessary but allows using relative paths.

VAULT_SRV_LOOKUP

Enables the client to lookup the host through DNS SRV look up as described in this draft. This is not designed for high-availability, just discovery. The draft specifies that the SRV record lookup is ignored if a port is given.

VAULT_MFA

ENTERPRISE ONLY

MFA credentials in the format mfa_method_name[:key[=value]] (items in [] are optional). Note that when using the environment variable, only one credential can be supplied. If a MFA method expects multiple credential values, or if there are multiple MFA methods specified on a path, then the CLI flag -mfa should be used.

VAULT_HTTP_PROXY

HTTP or HTTPS proxy location which should be used by all requests to access Vault. When present, this overrides the default proxy resolution behavior. Format should be http://server:port or https://server:port.

(See VAULT_PROXY_ADDR below).

VAULT_PROXY_ADDR

HTTP or HTTPS proxy location which should be used by all requests to access Vault. When present, this overrides the default proxy resolution behavior. Format should be http://server:port or https://server:port.

~> Note: When using VAULT_HTTP_PROXY or VAULT_PROXY_ADDR any of the standard proxy variables found in the environment will be ignored. Specifically HTTP_PROXY, HTTPS_PROXY and NO_PROXY. All requests will resolve the specified proxy; there is no way to exclude domains from consulting the proxy server.

~> Note: If both VAULT_HTTP_PROXY and VAULT_PROXY_ADDR environment variables are supplied, VAULT_PROXY_ADDR will be prioritized and preferred.

VAULT_DISABLE_REDIRECTS

Prevents the Vault client from following redirects. By default, the Vault client will automatically follow a single redirect.

~> Note: Disabling redirect following behavior could cause issues with commands such as 'vault operator raft snapshot' as this command redirects the request to the cluster's primary node.

Flags

There are different CLI flags that are available depending on subcommands. Some flags, such as those used for setting HTTP and output options, are available globally, while others are specific to a particular subcommand. For a complete list of available flags, run:

$ vault <subcommand> -h