open-vault/website/source/intro/getting-started/deploy.html.md

9.8 KiB

layout page_title sidebar_current description
intro Deploy Vault - Getting Started gettingstarted-deploy Learn how to deploy Vault into production, how to initialize it, configure it, etc.

Deploy Vault

Up to this point, we have been working with the "dev" server, which automatically authenticated us, setup in-memory storage, etc. Now that you know the basics of Vault, it is important to learn how to deploy Vault into a real environment.

On this page, we'll cover how to configure Vault, start Vault, the seal/unseal process, and scaling Vault.

Configuring Vault

Vault is configured using HCL files. The configuration file for Vault is relatively simple:

storage "consul" {
  address = "127.0.0.1:8500"
  path    = "vault/"
}

listener "tcp" {
 address     = "127.0.0.1:8200"
 tls_disable = 1
}

Within the configuration file, there are two primary configurations:

  • storage - This is the physical backend that Vault uses for storage. Up to this point the dev server has used "inmem" (in memory), but the example above uses Consul, a much more production-ready backend.

  • listener - One or more listeners determine how Vault listens for API requests. The example above listens on localhost port 8200 without TLS. In your environment set VAULT_ADDR=http://127.0.0.1:8200 so the Vault client will connect without TLS.

For now, copy and paste the configuration above to a file called config.hcl. It will configure Vault to expect an instance of Consul running locally.

Starting a local Consul instance takes only a few minutes. Just follow the Consul Getting Started Guide up to the point where you have installed Consul and started it with this command:

$ consul agent -dev

Starting the Server

With the configuration in place, starting the server is simple, as shown below. Modify the -config flag to point to the proper path where you saved the configuration above.

$ vault server -config=config.hcl
==> Vault server configuration:

         Log Level: info
           Storage: consul
        Listener 1: tcp (addr: "127.0.0.1:8200", tls: "disabled")

==> Vault server started! Log data will stream in below:

-> If you get a warning message about mlock not being supported, that is okay. However, you should run Vault on a system that supports mlock for maximum security.

Vault outputs some information about its configuration, and then blocks. This process should be run using a resource manager such as systemd or upstart.

You'll notice that you can't execute any commands. We don't have any auth information! When you first setup a Vault server, you have to start by initializing it.

On Linux, Vault may fail to start with the following error:

$ vault server -config=example.hcl
Error initializing core: Failed to lock memory: cannot allocate memory

This usually means that the mlock syscall is not available.
Vault uses mlock to prevent memory from being swapped to
disk. This requires root privileges as well as a machine
that supports mlock. Please enable mlock on your system or
disable Vault from using it. To disable Vault from using it,
set the `disable_mlock` configuration option in your configuration
file.

For guidance on dealing with this issue, see the discussion of disable_mlock in Server Configuration.

Initializing the Vault

Initialization is the process configuring the Vault. This only happens once when the server is started against a new backend that has never been used with Vault before. When running in HA mode, this happens once per cluster, not per server.

During initialization, the encryption keys are generated, unseal keys are created, and the initial root token is setup. To initialize Vault use vault operator init. This is an unauthenticated request, but it only works on brand new Vaults with no data:

$ vault operator init
Unseal Key 1: E4GnjX+VP9G50uWQNcwpCflzGAMKGR38BbQywgq4I6L8
Unseal Key 2: PYMxcCOswEYMNz7N6UW53Up6nu6y+SjAPwTJOTtkju3d
Unseal Key 3: yuJ5cSxC7tSBR5mMVJ/WJ9bfhhfGb+uwWw9FQR0JKILh
Unseal Key 4: 0vdvEFHM9PHEGMctJrl2ylHqoKQK8DLkfMU6ntmDz6jv
Unseal Key 5: cI8yglWJX+jPf/yQG7Sg6SPWzy0WyrBPvaFTOAYkPJTx

Initial Root Token: 62421926-81b9-b202-86f8-8850176c0cf3

Vault initialized with 5 key shares and a key threshold of 3. Please securely
distribute the key shares printed above. When the Vault is re-sealed,
restarted, or stopped, you must supply at least 3 of these keys to unseal it
before it can start servicing requests.

Vault does not store the generated master key. Without at least 3 key to
reconstruct the master key, Vault will remain permanently sealed!

It is possible to generate new unseal keys, provided you have a quorum of
existing unseal keys shares. See "vault operator rekey" for more information.

Initialization outputs two incredibly important pieces of information: the unseal keys and the initial root token. This is the only time ever that all of this data is known by Vault, and also the only time that the unseal keys should ever be so close together.

For the purpose of this getting started guide, save all of these keys somewhere, and continue. In a real deployment scenario, you would never save these keys together. Instead, you would likely use Vault's PGP and Keybase.io support to encrypt each of these keys with the users' PGP keys. This prevents one single person from having all the unseal keys. Please see the documentation on using PGP, GPG, and Keybase for more information.

Seal/Unseal

Every initialized Vault server starts in the sealed state. From the configuration, Vault can access the physical storage, but it can't read any of it because it doesn't know how to decrypt it. The process of teaching Vault how to decrypt the data is known as unsealing the Vault.

Unsealing has to happen every time Vault starts. It can be done via the API and via the command line. To unseal the Vault, you must have the threshold number of unseal keys. In the output above, notice that the "key threshold" is 3. This means that to unseal the Vault, you need 3 of the 5 keys that were generated.

-> Note: Vault does not store any of the unseal key shards. Vault uses an algorithm known as Shamir's Secret Sharing to split the master key into shards. Only with the threshold number of keys can it be reconstructed and your data finally accessed.

Begin unsealing the Vault:

$ vault operator unseal
Unseal Key (will be hidden):
Key                Value
---                -----
Sealed             true
Total Shares       5
Unseal Progress    1/3
Unseal Nonce       786e7190-d1e2-84d2-520c-022efee5b71e
Version            (version unknown)
HA Enabled         true
HA Mode            sealed

After pasting in a valid key and confirming, you'll see that the Vault is still sealed, but progress is made. Vault knows it has 1 key out of 3. Due to the nature of the algorithm, Vault doesn't know if it has the correct key until the threshold is reached.

Also notice that the unseal process is stateful. You can go to another computer, use vault operator unseal, and as long as it's pointing to the same server, that other computer can continue the unseal process. This is incredibly important to the design of the unseal process: multiple people with multiple keys are required to unseal the Vault. The Vault can be unsealed from multiple computers and the keys should never be together. A single malicious operator does not have enough keys to be malicious.

Continue with vault operator unseal to complete unsealing the Vault. To unseal the vault you must use three different keys, the same key repeated will not work. As you use keys, as long as they are correct, you should soon see output like this:

$ vault operator unseal
Unseal Key (will be hidden):
# ...

$ vault operator unseal
Unseal Key (will be hidden):
# ...

When the value for Sealed changes to false, the Vault is unsealed:

Key             Value
---             -----
Sealed          false <--
Total Shares    5
Version         (version unknown)
Cluster Name    vault-cluster-8a8b2c36
Cluster ID      34e94a2e-2d8f-c7cc-271d-96fd438ccc6d
HA Enabled      true
HA Mode         standby
HA Cluster      n/a

Feel free to play around with entering invalid keys, keys in different orders, etc. in order to understand the unseal process.

Finally, authenticate as the initial root token (it was included in the output with the unseal keys):

$ vault login 14d1316e-78f6-910b-a4cc-9ba6697ec814
Success! You are now authenticated. The token information displayed below
is already stored in the token helper. You do NOT need to run "vault login"
again. Future Vault requests will automatically use this token.

Key                Value
---                -----
token              14d1316e-78f6-910b-a4cc-9ba6697ec814
token_accessor     a8bbcc57-9be6-6584-a7a6-46290962fd33
token_duration     ∞
token_renewable    false
token_policies     [root]

As a root user, you can reseal the Vault with vault operator seal. A single operator is allowed to do this. This lets a single operator lock down the Vault in an emergency without consulting other operators.

When the Vault is sealed again, it clears all of its state (including the encryption key) from memory. The Vault is secure and locked down from access.

Next

You now know how to configure, initialize, and unseal/seal Vault. This is the basic knowledge necessary to deploy Vault into a real environment. Once the Vault is unsealed, you access it as you have throughout this getting started guide (which worked with an unsealed Vault).

Next, we have a short tutorial on using the HTTP APIs to authenticate and access secrets.