LiteFS Config Reference

litefs mount -config /path/to/litefs.yml

litefs run -config /path/to/litefs.yml

${MY_VAR == "xyz"}

${MY_VAR != "xyz"}

# The FUSE section handles settings on the FUSE file system. FUSE
# provides a layer for intercepting SQLite transactions on the
# primary node so they can be shipped to replica nodes
# transparently.
fuse:
  # Required. This is the mount directory that applications will
  # use to access their SQLite databases.
  dir: "/litefs"

  # Enable mounting of the file system by non-root users.
  # You must enable the 'user_allow_other' option in /etc/fuse.conf as well.
  allow-other: false

  # The debug flag enables debug logging of all FUSE API calls.
  # This will produce a lot of logging. Not for general use.
  debug: false

# The data section specifies where internal LiteFS data is stored
# and how long to retain the transaction files.
# 
# Transaction files are used to ship changes to replica nodes so
# they should persist long enough for replicas to retrieve them,
# even in the face of a short network interruption or a redeploy.
# Under high load, these files can grow large so it's not advised
# to extend retention too long.
data:
  # Path to internal data storage.
  dir: "/var/lib/litefs"

  # If true, compresses LTX files using LZ4 compression. Enabled by default.
  compress: true

  # Duration to keep LTX files. Latest LTX file is always kept.
  retention: "10m"

  # Frequency with which to check for LTX files to delete.
  retention-monitor-interval: "1m"

# The exec field specifies a command to run as a subprocess of
# LiteFS. This command will be executed after LiteFS either
# becomes primary or is connected to the primary node. LiteFS
# will forward signals to the subprocess and LiteFS will
# automatically shut itself down when the subprocess stops.
#
# This can also be specified after a double-dash (--) on the
# command line invocation of the 'litefs mount' command.
exec: "myapp -addr :8080"

# This section defines settings for the LiteFS HTTP API server.
# This API server is how nodes communicate with each other.
http:
  # Specifies the bind address of the HTTP API server.
  addr: ":20202"

# This section defines settings for the option HTTP proxy.
# This proxy can handle primary forwarding & replica consistency
# for applications that use a single SQLite database.
proxy:
  # Specifies the bind address of the proxy server.
  addr: ":8080"

  # The hostport of the target application. If blank, proxy is disabled.
  target: "localhost:8081"

  # The name of the database used for TXID tracking.
  db: "my.db"

  # If true, enables verbose logging of requests by the proxy.
  debug: false

  # List of paths that are ignored by the proxy. The asterisk is
  # the only available wildcard. These requests are passed
  # through to the target as-is.
  passthrough: ["/debug/*", "*.png"]

  # The duration to hold write requests when no primary is available.
  # This is helpful when a primary switches over and there is a small window
  # where the new primary is unknown to the replica.
  primary-redirect-timeout: "5s"

# The lease section defines how LiteFS creates a cluster and
# implements leader election. For dynamic clusters, use the
# "consul". This allows the primary to change automatically when
# the current primary goes down. For a simpler setup, use
# "static" which assigns a single node to be the primary and does
# not failover.
lease:
  # Required. Must be either "consul" or "static".
  type: "consul"

  # Required. The URL for this node's LiteFS API.
  # Should match HTTP port.
  advertise-url: "http://$HOSTNAME:20202"

  # Sets the hostname that other nodes will use to reference this
  # node. Automatically assigned based on hostname(1) if not set.
  hostname: ""

  # Specifies whether the node can become the primary. If using
  # "static" leasing, this should be set to true on the primary
  # and false on the replicas.
  candidate: true

  # Filtered list of databases to replicate from the primary to this node.
  # If no databases are specified, all databases are replicated. You may only
  # filter databases on non-candidate nodes.
  databases: [ "my.db", "myother.db" ]

lease:
  # A Consul server provides leader election and ensures that the
  # responsibility of the primary node can be moved in the event
  # of a deployment or a failure.
  consul:
    # Required. The base URL of the Consul server.
    url: "http://myhost:8500"

    # Required. The key used for obtaining a lease by the primary.
    # This must be unique for each cluster of LiteFS servers
    key: ""

    # Length of time before a lease expires. The primary will
    # automatically renew the lease while it is alive, however,
    # if it fails to renew in time then a new primary may be
    # elected after the TTL. This only occurs for unexpected loss
    # of the leader as normal operation will allow the leader to
    # handoff the lease to another replica without downtime.
    #
    # Consul does not allow a TTL of less than 10 seconds.
    ttl: "10s"

    # Length of time after the lease expires before a candidate
    # can become leader. This buffer is intended to prevent
    # overlap in leadership due to clock skew or in-flight calls.
    lock-delay: "1s"

# If true, then LiteFS will not exit if there is a validation
# issue on startup. This can be useful for debugging issues as
# it avoids constantly restarting the node on ephemeral hosting.
exit-on-error: false

# If true, then LiteFS will not wait until the node becomes the
# primary or connects to the primary before starting the subprocess.
skip-sync: false

# The tracing section enables a rolling, on-disk tracing log.
# This records every operation to the database so it can be
# verbose and it can degrade performance. This is for debugging
# only and should not typically be enabled in production.
tracing:
  # Toggles tracing if 'path' is also set. Defaults to true.
  enabled: true

  # Output path on disk.
  path: "/var/log/lifefs/trace.log"

  # Maximum size of a single trace log before rolling.
  # Specified in megabytes.
  max-size: 64

  # Maximum number of trace logs to retain.
  max-count: 10

  # If true, historical logs will be compressed using gzip.
  compress: true