Configuring Splinter Daemon Database

Much of the runtime data maintained by a splinter daemon is persisted in a database instance. This includes things like circuit state and node registry information.

Splinter supports both SQLite and PostgreSQL databases. The former is useful for development situations, or very small testing deployments. The latter is well-suited for production deployments.

Services, such as scabbard, may provide alternative methods for managing their own state persistence. As a result, these are not necessarily stored in the same database.



Splinter supports SQLite version 3.6.8 or newer.


Splinter supports Postgres version 9.5 and above

Connection Strings

Connections to databases are established via a string appropriate for the database target.

For postgres, this string is a standard URI format


For SQLite, this format follows the basic connection string format, specifying either a file path, or memory only. For example:

  • splinter_state.db (the default SQLite database)
  • /var/lib/splinter/my_splinter.db
  • :memory:

Running Migrations

Over the lifetime of a Splinter node, the database schema may change across releases. The splinter CLI tool provides a command for running database migrations, in order to upgrade the installation.

This example command updates a PostgreSQL database:

$ splinter database migrate \
    -C postgres://admin:my_password@postgres-host/splinter

Running this command without a specific connection string will apply the migrations to the default database at $SPLINTER_STATE_DIR/splinter_state.db

Splinter Daemon Configuration

The Splinter daemon can be configured to connect to a specific database via a configuration option. This option can be set in several ways.

  • Configure the database connection via the --database option on the splinterd command:

    $ splinterd \
        --database postgres://admin:my_password@postgres-host/splinter
  • Configure the database via the toml configuration file:

    database = "postgres://admin:my_password@postgres-host/splinter"
  • For SQLite databases, configure the location of the database file via an environment variable.


    $ export SPLINTER_HOME=/my_splinter_home

    where the resulting SQLite database files will be found in /my_splinter_home/data.


    $ export SPLINTER_STATE_DIR=/my_splinter_data

    where the resulting SQLite database files will be found in /my_splinter_data.

    In either case, the database file itself can be customized specified using the --database option.

    $ splinterd --database my_splinter_state.db

    If no --database option is provided, the default value of splinter_state.db will be used.

Scabbard State Configuration

The Splinter daemon includes the scabbard service for use in circuits. Scabbard state is stored in the same database as the rest of the daemon. This includes data such as transaction receipts, the current commit hash, and transaction merkle state.

Scabbard may also be configured to store the transaction merkle state using LMDB. This is enabled using the flag --scabbard-state. For example

$ splinterd --scabbard-state lmdb

This setting maybe chosen either for using existing services data (after upgrading, for example) or for particularly performance-sensitive installations. Once this is used LMDB must be enabled for all restarts of a given Splinter daemon.