Vault with integrated storage deployment guide

12min
|
Vault

Vault Enterprise

As of version 1.12.0, Vault Enterprise will no longer start up if configured to use a storage backend other than Integrated Storage or Consul. See the Release Notes for more information.

This deployment guide outlines the required steps to manually install and configure a single HashiCorp Vault cluster as defined in the Vault with Integrated Storage Reference Architecture on virtual machines (VMs) or bare-metal servers running a Debian- or RedHat-based Linux distribution.

This guide includes general guidance as well as specific recommendations for popular cloud infrastructure platforms. These recommendations have also been encoded into official Terraform modules for AWS, Azure, and GCP.

Install Vault

Pre-built Vault packages are available from the HashiCorp Linux Repository. In addition to the installing the Vault binary, the official package also seeds Vault with a working configuration, a systemd service unit, and a local vault user account under which the service will be run.

Add the HashiCorp GPG key as a trusted package-signing key.

$ curl -fsSL https://apt.releases.hashicorp.com/gpg | sudo apt-key add -

Add the official HashiCorp Linux repository.

$ sudo apt-add-repository "deb [arch=amd64] https://apt.releases.hashicorp.com $(lsb_release -cs) main"

Update the package index.

$ sudo apt update

Install Vault.

$ sudo apt install vault-enterprise

Note

To install Vault Community Edition, replace vault-enterprise with vault.

Install yum-config-manager to manage your repositories.

$ sudo yum install -y yum-utils

Add the official HashiCorp Linux repository.

$ sudo yum-config-manager --add-repo https://rpm.releases.hashicorp.com/RHEL/hashicorp.repo

Install Vault.

$ sudo yum -y install vault-enterprise

Note

To install Vault Community Edition, replace vault-enterprise with vault.

Install dnf config-manager to manage your repositories.

$ sudo dnf install -y dnf-plugins-core

Add the official HashiCorp Linux repository.

$ sudo dnf config-manager --add-repo https://rpm.releases.hashicorp.com/fedora/hashicorp.repo

Install Vault.

$ sudo dnf -y install vault-enterprise

Note

To install Vault Community Edition, replace vault-enterprise with vault.

Install yum-config-manager to manage your repositories.

$ sudo yum install -y yum-utils

Add the official HashiCorp Linux repository.

$ sudo yum-config-manager --add-repo https://rpm.releases.hashicorp.com/AmazonLinux/hashicorp.repo

Install Vault.

$ sudo yum -y install vault-enterprise

Note

To install Vault Community, replace vault-enterprise with vault.

Prepare TLS certificates

Note

The installation package generates a self-signed TLS certificate for use by the Vault service to secure intra-cluster communication. While these certificates are usable for experimenting with getting Vault up and running, HashiCorp strongly recommends replacing them with certificates generated and signed by an appropriate CA.

You must have three files to configure TLS for Vault. Place them at these paths:

/opt/vault/tls/vault-cert.pem - The Vault TLS certificate itself, which a Common Name (CN) or Subject Alternative Name (SAN) that matches the value set for leader_tls_servername in the Raft configuration. If this file was signed by an intermediate CA, append the certificate of that CA (and any other chained CAs) to the end of this file.
/opt/vault/tls/vault-key.pem - The private key of the Vault TLS certificate.
/opt/vault/tls/vault-ca.pem - The certificate of the CA root which signed the Vault TLS certificate.

Set the file ownership of the CA and certificate files to be owned by root.

$ sudo chown root:root /opt/vault/tls/vault-cert.pem /opt/vault/tls/vault-ca.pem

Set the file group ownership of the private key to allow the Vault service to read the file.

$ sudo chown root:vault /opt/vault/tls/vault-key.pem

Set the file permissions of the CA and certificate files to be world-readable.

$ sudo chmod 0644 /opt/vault/tls/vault-cert.pem /opt/vault/tls/vault-ca.pem

Set the file permissions of the private key file to be readable only by the Vault service.

$ sudo chmod 0640 /opt/vault/tls/vault-key.pem

Configure Vault

Vault may be configured by editing the /etc/vault.d/vault.hcl file included with the installation package. Refer to the Vault Configuration Overview for additional details about each setting.

Base configuration

The core required configuration values for Vault are cluster_addr, api_addr, and listener. The configuration below tells vault to advertise its local IPv4 address for internal communication, and to listen to TCP port 8200 on all network interfaces using the TLS configuration files specified.

The configuration also sets disable_mlock to true, which is strongly recommended for the Integrated Storage backend to avoid out-of-memory errors.

/etc/vault.d/vault.hcl

cluster_addr  = "https://<LOCAL_IPV4_ADDRESS>:8201"
api_addr      = "https://<LOCAL_IPV4_ADDRESS>:8200"
disable_mlock = true

listener "tcp" {
  address            = "0.0.0.0:8200"
  tls_cert_file      = "/opt/vault/tls/vault-cert.pem"
  tls_key_file       = "/opt/vault/tls/vault-key.pem"
  tls_client_ca_file = "/opt/vault/tls/vault-ca.pem"
}

Raft configuration

Using Vault Integrated Storage requires configuring the Raft storage backend. Raft peers may be initialized manually, with hard-coded configuration values, or via the cloud auto-join feature on supported cloud providers. Refer to the Integrated Storage configuration documentation for additional details about each setting.

To use Raft auto-join on AWS, each Vault EC2 instance must be tagged with a key-value pair that is unique to its specific Vault cluster. The instances must also have appropriate permissions via an IAM role attached to their instance profile. See the Cloud auto-join documentation for more details about the format of the auto_join configuration value and the permissions required.

/etc/vault.d/vault.hcl

storage "raft" {
  path    = "/opt/vault/data"
  node_id = "<UNIQUE_ID_FOR_THIS_HOST>"

  retry_join {
    auto_join               = "provider=aws region=<AWS_REGION> tag_key=<VAULT_TAG_NAME> tag_value=<VAULT_TAG_VALUE>"
    auto_join_scheme        = "https"
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
}

To use Raft auto-join on Azure, each Vault VM may be tagged with a key-value pair that is unique to its specific Vault cluster or the VMs can be identified by their VM scale set ID if so provisioned. See the Cloud auto-join documentation for more details about the format of the auto_join configuration value and the permissions required.

/etc/vault.d/vault.hcl

storage "raft" {
  path    = "/opt/vault/data"
  node_id = "<UNIQUE_ID_FOR_THIS_HOST>"

  retry_join {
    auto_join               = "provider=azure subscription_id=<SUBSCRIPTION_ID> resource_group=<RESOURCE_GROUP_NAME> vm_scale_set=<VM_SCALE_SET_NAME>"
    auto_join_scheme        = "https"
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
}

To use Raft auto-join on GCP, each Vault compute instance must be tagged with a value that is unique to its specific Vault cluster. See the Cloud auto-join documentation for more details about the format of the auto_join configuration value and the permissions required.

/etc/vault.d/vault.hcl

storage "raft" {
  path    = "/opt/vault/data"
  node_id = "<UNIQUE_ID_FOR_THIS_HOST>"

  retry_join {
    auto_join               = "provider=gce tag_value=<UNIQUE_VAULT_TAG>"
    auto_join_scheme        = "https"
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
}

In cloud or on-premises environments, Vault can be pre-configured to join with its cluster peers directly via the configuration file.

/etc/vault.d/vault.hcl

storage "raft" {
  path    = "/opt/vault/data"
  node_id = "<UNIQUE_ID_FOR_THIS_HOST>"

  retry_join {
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_api_addr         = "https://<ADDRESS_OF_PEER_1>:8200"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
  retry_join {
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_api_addr         = "https://<ADDRESS_OF_PEER_2>:8200"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
  retry_join {
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_api_addr         = "https://<ADDRESS_OF_PEER_3>:8200"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
  retry_join {
    leader_tls_servername   = "<VALID_TLS_SERVER_NAME>"
    leader_api_addr         = "https://<ADDRESS_OF_PEER_4>:8200"
    leader_ca_cert_file     = "/opt/vault/tls/vault-ca.pem"
    leader_client_cert_file = "/opt/vault/tls/vault-cert.pem"
    leader_client_key_file  = "/opt/vault/tls/vault-key.pem"
  }
}

Auto-unseal

If you are using a supported cloud provider or have access to a supported hardware security module (HSM), you can enable auto-unseal using the seal configuration stanza.

To use AWS KMS for auto-unsealing securely, your EC2 instances should have appropriate permissions to a KMS key via their instance profile. See the awskms seal documentation for full details of the configuration options and necessary permissions.

/etc/vault.d/vault.hcl

seal "awskms" {
  region     = "<AWS_REGION>"
  kms_key_id = "<KMS_KEY_ARN>"
}

To use Azure Key Vault for auto-unsealing securely, your VMs should have a Managed Service Identity (MSI) with appropriate permissions to an Azure Key Vault. See the azurekeyvault seal documentation for full details of the configuration options and necessary permissions.

/etc/vault.d/vault.hcl

seal "azurekeyvault" {
  tenant_id  = "<AZURE_TENANT_ID>"
  vault_name = "<AZURE_VAULT_NAME>"
  key_name   = "<AZURE_KEY_NAME>"
}

To use GCP Cloud KMS for auto-unsealing securely, your compute instances should be bound to a service account with appropriate permissions to a GCP Cloud KMS Key Ring. See the gcpckms seal documentation for full details of the configuration options and necessary permissions.

/etc/vault.d/vault.hcl

seal "gcpckms" {
  project    = "<GCP_PROJECT_NAME>"
  region     = "<GCP_REGION>"
  key_ring   = "<GCP_KMS_KEY_RING>"
  crypto_key = "<GCP_KMS_KEY_NAME>"
}

Note

This section applies only to Vault Enterprise.

See the pkcs11 seal documentation for full details of the configuration options and necessary permissions.

/etc/vault.d/vault.hcl

seal "pkcs11" {
  lib            = "<PATH_TO_HSM_SHARED_LIBRARY>"
  slot           = "<HSM_SLOT_NUMBER>"
  pin            = "<HSM_PIN>"
  key_label      = "<HSM_KEY_LABEL>"
  hmac_key_label = "<HSM_HMAC_KEY_LABEL>"
}

Refer to the auto-unseal collection for a full walkthrough of setting up auto-unseal with Vault.

Enterprise license

Note

This section applies only to Vault Enterprise.

Beginning with Vault 1.8, an Enterprise license file must exist on disk and be readable by the Vault service.

Copy your Vault Enterprise license file to /opt/vault/vault.hclic. Then set the file's group ownership to allow the Vault service to read the file.

$ sudo chown root:vault /opt/vault/vault.hclic

Set the file permissions to allow the Vault service to read the file, but no other users to access the file.

$ sudo chmod 0640 /opt/vault/vault.hclic

Add the path to the Enterprise license file to /etc/vault.d/vault.hcl.

/etc/vault.d/vault.hcl

license_path = "/opt/vault/vault.hclic"

Vault web UI

To enable the Vault web UI, set the ui configuration value to true.

/etc/vault.d/vault.hcl

ui = true

Enable and start the Vault service

Enable the systemd vault.service unit to allow the Vault service to start.

$ sudo systemctl enable vault.service

Start the Vault service.

$ sudo systemctl start vault.service

Check the status of the Vault service to ensure it is running.

$ sudo systemctl status vault.service

Check the status of Vault

To ensure Vault is running, you can use the vault status command. To ensure TLS connection can be validate, first set the VAULT_CACERT environment variable to the path of the CA root certificate.

$ export VAULT_CACERT=/opt/vault/tls/vault-ca.pem

Then use vault status to see the status of the local Vault server.

$ vault status
Key                     Value
---                     -----
[...]
Storage Type            raft
[...]

Initialize Vault

If this is the first time you are running this cluster, Vault will need to be initialized on one host before Raft clients can be joined.

$ vault operator init
Unseal Key 1: <omitted>
[...]
Initial Root Token: s.AJ8gE4odM92zW1clYbPPXpLu
[...]

Take note of these values and store them securely. You will need them to manually unseal Vault and to perform initial setup.

Unseal Vault

If you configured auto-unseal, Vault should unseal automatically. Use the vault status command to check whether Vault is sealed.

$ vault status
Key                     Value
---                     -----
[...]
Sealed                  false
[...]

If Vault is sealed, use the vault operator unseal command on each host in the cluster with the unseal keys you received when initializing Vault to unseal it.

$ vault operator unseal
Unseal Key (will be hidden): ******************************
Key                Value
---                -----
[...]
Unseal Progress    1/3
[...]

Repeat the unseal command with new keys until the unseal progress is complete and Vault is unsealed.

Additional configuration

Audit Logging

Vault provides audit logging via Audit devices which serialize details of each request and response Vault handles. These logs may be stored to a file on disk, or sent to syslog or any socket endpoint.

Learn more about using audit logs to monitor and secure your Vault deployment with the Monitor Telemetry & Audit Device Log Data tutorial.

Autopilot

Integrated Storage Autopilot may be configured through the Vault API. If you are using an autoscaling cloud service to manage your Vault instances, it's important to set the Cleanup Dead Servers setting to true and to tune the Min Quorum, Dead Server Last Contact Threshold, and Server Stabilization Time values for your deployment environment. See the Autopilot tutorial for more information on each configuration value.

Replication

Vault Enterprise supports replication functionality to support performance and disaster recovery use cases. There are a number of tutorials and other documentation to explain the concepts and provide details about configuring these replication modes.

Resource Quotas

Vault provides tools for establishing lease count limits (Enterprise only) and request rate limits to avoid systemwide denial-of-service. Learn more about the concepts and capabilities behind Vault resource quotas in the Resource Quotas concepts documentation, and discover how to enable and manage resource quotas in the Protecting Vault with Resource Quotas tutorial.

Telemetry

The telemetry stanza specifies various configurations for Vault to publish metrics to upstream systems. Vault supports sending telemetry to a number of systems including Circonus, DogStatsD, Prometheus, Stackdriver, Statsite, and any statsd-compatible endpoint.

Review the telemetry configuration section of our documentation for details on configuring Vault telemetry for your provider, and learn more about Vault's telemetry collection and which metrics are collected in the telemetry internals documentation.

You can also learn more about using telemetry to monitor Vault on our Monitor Telemetry & Audit Device Log Data tutorial.

Next Steps

Additional references

Integrated Storage reference

Multi-cluster architecture guide

Harden server deployments

This tutorial also appears in:

9 tutorials

Integrated Storage
Operational tasks associated with integrated storage to persist Vault data rather than using external storage.
- Vault