Skip to main content

Integrated storage (Raft) backend

The Integrated Storage backend is used to persist OpenBao's data. Unlike other storage backends, Integrated Storage does not operate from a single source of data. Instead all the nodes in an OpenBao cluster will have a replicated copy of OpenBao's data. Data gets replicated across all the nodes via the Raft Consensus Algorithm.

  • High Availability – the Integrated Storage backend supports high availability.

  • HashiCorp Supported – the Integrated Storage backend is officially supported by HashiCorp.

storage "raft" {
  path = "/path/to/raft/data"
  node_id = "raft_node_1"
}
cluster_addr = "http://127.0.0.1:8201"
warning

Note: When using the Integrated Storage backend, it is required to provide cluster_addr to indicate the address and port to be used for communication between the nodes in the Raft cluster.

warning

Note: When using the Integrated Storage backend, a separate ha_storage backend cannot be declared.

raft parameters​

  • path (string: "") – The file system path where all the OpenBao data gets stored. This value can be overridden by setting the VAULT_RAFT_PATH environment variable.

  • node_id (string: "") - The identifier for the node in the Raft cluster. This value can be overridden by setting the VAULT_RAFT_NODE_ID environment variable.

  • performance_multiplier (integer: 0) - An integer multiplier used by servers to scale key Raft timing parameters. Tuning this affects the time it takes OpenBao to detect leader failures and to perform leader elections, at the expense of requiring more network and CPU resources for better performance. Omitting this value or setting it to 0 uses default timing described below. Lower values are used to tighten timing and increase sensitivity while higher values relax timings and reduce sensitivity.

    By default, OpenBao will use a lower-performance timing that's suitable for minimal OpenBao servers, currently equivalent to setting this to a value of 5 (this default may be changed in future versions of OpenBao, depending if the target minimum server profile changes). Setting this to a value of 1 will configure Raft to its highest-performance mode and is recommended for production OpenBao servers. The maximum allowed value is 10.

  • trailing_logs (integer: 10000) - This controls how many log entries are left in the log store on disk after a snapshot is made. This should only be adjusted when followers cannot catch up to the leader due to a very large snapshot size and high write throughput causing log truncation before a snapshot can be fully installed. If you need to use this to recover a cluster, consider reducing write throughput or the amount of data stored on OpenBao. The default value is 10000 which is suitable for all normal workloads. The trailing_logs metric is not the same as max_trailing_logs.

  • snapshot_threshold (integer: 8192) - This controls the minimum number of Raft commit entries between snapshots that are saved to disk. This is a low-level parameter that should rarely need to be changed. Very busy clusters experiencing excessive disk IO may increase this value to reduce disk IO and minimize the chances of all servers taking snapshots at the same time. Increasing this trades off disk IO for disk space since the log will grow much larger and the space in the raft.db file can't be reclaimed till the next snapshot. Servers may take longer to recover from crashes or failover if this is increased significantly as more logs will need to be replayed.

  • snapshot_interval (integer: 120 seconds) - The snapshot interval controls how often Raft checks whether a snapshot operation is required. Raft randomly staggers snapshots between the configured interval and twice the configured interval to keep the entire cluster from performing a snapshot at once. The default snapshot interval is 120 seconds.

  • retry_join (list: []) - A set of connection details for another node in the cluster, which is used to help nodes locate a leader in order to join a cluster. There can be one or more retry_join stanzas.

    If the connection details for all nodes in the cluster are known in advance, you can include these stanzas to enable nodes to automatically join the Raft cluster. Once one of the nodes is initialized as the leader, the remaining nodes will use their retry_join configuration to locate the leader and join the cluster. Note that when using Shamir seal, the joined nodes will still need to be unsealed manually. See the section below for the parameters accepted by the retry_join stanza.

  • retry_join_as_non_voter (boolean: false) - If set, causes any retry_join config to join the Raft cluster as a non-voter. The node will not participate in the Raft quorum but will still receive the data replication stream, adding read scalability to a cluster. This option has the same effect as the -non-voter flag for the bao operator raft join command, but only affects voting status when joining via retry_join config. This setting can be overridden to true by setting the VAULT_RAFT_RETRY_JOIN_AS_NON_VOTER environment variable to any non-empty value. Only valid if there is at least one retry_join stanza.

  • max_entry_size (integer: 1048576) - This configures the maximum number of bytes for a Raft entry. It applies to both Put operations and transactions. Any put or transaction operation exceeding this configuration value will cause the respective operation to fail. Raft has a suggested max size of data in a Raft log entry. This is based on current architecture, default timing, etc. Integrated Storage also uses a chunk size that is the threshold used for breaking a large value into chunks. By default, the chunk size is the same as Raft's max size log entry. The default value for this configuration is 1048576 -- two times the chunking size.

  • autopilot_reconcile_interval (string: "10s") - This is the interval after which autopilot will pick up any state changes. State change could mean multiple things; for example a newly joined voter node, initially added as non-voter to the Raft cluster by autopilot has successfully completed the stabilization period thereby qualifying for being promoted as a voter, a node that has become unhealthy and needs to be shown as such in the state API, a node has been marked as dead needing eviction from Raft configuration, etc.

  • autopilot_update_interval (string: "2s") - This is the interval after which autopilot will poll OpenBao for any updates to the information it cares about. This includes things like the autopilot configuration, current autopilot state, raft configuration, known servers, latest raft index, and stats for all the known servers. The information that autopilot receives will be used to calculate its next state.

retry_join stanza​

  • leader_api_addr (string: "") - Address of a possible leader node.

  • auto_join (string: "") - Cloud auto-join configuration, using go-discover syntax.

  • auto_join_scheme (string: "") - The optional URI protocol scheme for addresses discovered via auto-join. Available values are http or https.

  • auto_join_port (uint: "") - The optional port used for addressed discovered via auto-join.

  • leader_tls_servername (string: "") - The TLS server name to use when connecting with HTTPS. Should match one of the names in the DNS SANs of the remote server certificate. See also Integrated Storage and TLS

  • leader_ca_cert_file (string: "") - File path to the CA cert of the possible leader node.

  • leader_client_cert_file (string: "") - File path to the client certificate for the follower node to establish client authentication with the possible leader node.

  • leader_client_key_file (string: "") - File path to the client key for the follower node to establish client authentication with the possible leader node.

  • leader_ca_cert (string: "") - CA cert of the possible leader node.

  • leader_client_cert (string: "") - Client certificate for the follower node to establish client authentication with the possible leader node.

  • leader_client_key (string: "") - Client key for the follower node to establish client authentication with the possible leader node.

Each retry_join block may provide TLS certificates via file paths or as a single-line certificate string value with newlines delimited by \n, but not a combination of both. Each retry_join stanza may contain either a leader_api_addr value or a cloud auto_join configuration value, but not both. When an auto_join value is provided, OpenBao will automatically attempt to discover and resolve potential Raft leader addresses using go-discover. See the go-discover README for details on the format of the auto_join value.

By default, OpenBao will attempt to reach discovered peers using HTTPS and port 8200. Operators may override these through the auto_join_scheme and auto_join_port fields respectively.

Example Configuration:

storage "raft" {
  path    = "/Users/foo/raft/"
  node_id = "node1"

  retry_join {
    leader_api_addr = "http://127.0.0.2:8200"
    leader_ca_cert_file = "/path/to/ca1"
    leader_client_cert_file = "/path/to/client/cert1"
    leader_client_key_file = "/path/to/client/key1"
  }
  retry_join {
    leader_api_addr = "http://127.0.0.3:8200"
    leader_ca_cert_file = "/path/to/ca2"
    leader_client_cert_file = "/path/to/client/cert2"
    leader_client_key_file = "/path/to/client/key2"
  }
  retry_join {
    leader_api_addr = "http://127.0.0.4:8200"
    leader_ca_cert_file = "/path/to/ca3"
    leader_client_cert_file = "/path/to/client/cert3"
    leader_client_key_file = "/path/to/client/key3"
  }
  retry_join {
    auto_join = "provider=aws region=eu-west-1 tag_key=openbao tag_value=... access_key_id=... secret_access_key=..."
  }
}