To avoid a single point of failure with your HashiCorp Vault infrastructure, one would set up Vault in High Availability Mode (HA) with two identical Vault instances (one active and one standby) and use Keepalived to run VRRP between them.

VRRP provides a virtual IP address to the active Vault HA, and transfers the Virtual IP to the standby Vault HA in case of failure.

Network interface cards (NICs) typically bind to a single IP address in TCP/IP networks. However, you can also tell the NIC to listen on extra addresses. Such addresses are called virtual IP addresses or VIPs for short.

You can configure a VIP address by using a tracking script, that is a program that Keepalived runs at regular intervals, according to a vrrp_script definition:

global_defs {
  script_user keepalived
  enable_script_security
}

vrrp_script vault_active_node_script {
  script       "/etc/keepalived/vault_ha_active_node.py --timeout=1 --url='https://127.0.0.1:8200'"
  interval 2   # Run script every 2 seconds
  fall 1       # If script returns non-zero 2 times in succession, enter FAULT state
  rise 3       # If script returns zero r times in succession, exit FAULT state
  timeout 2    # Wait up to t seconds for script before assuming non-zero exit code
  weight 10    # Reduce priority by 10 on fall
}

In this example keepalived is the (non-root) user executing the script vault_ha_active_node.py.

You can use tracking scripts with a vrrp_instance section by specifying a track_script clause, for example:

vrrp_instance VIP_VAULT_HA {
  state MASTER
  interface ens192
  virtual_router_id 25
  priority 100
  advert_int 1
  authentication {
    auth_type PASS
    auth_pass n0ts0r4nd0m
  }
  virtual_ipaddress {
    10.0.0.100/24
  }
  track_script {
    vault_active_node_script
  }
}

The Python3 script vault_ha_active_node.py looks at the environment variable VAULT_ADDR and when set do ask the Vault REST API if the (local) Vault node is the active one. In that case will return 0. The script will return 1 otherwise (VAULT_ADDR unset, vault service not running or running but not unsealed, node in standby mode, a network error).

The script requires the Python library requests (provided by the package python3-requests in Debian and Fedora).

./vault_ha_active_node.py --timeout=1 --url='https://127.0.0.1:8200'

You can also do some debugging (check for messages with journald -f)

./vault_ha_active_node.py --debug