Migrate to a Chronograf HA configuration

Use chronoctl to migrate your Chronograf configuration store from BoltDB to a shared etcd data store used for Chronograf high-availability (HA) clusters.

Note: Migrating Chronograf to a shared data source creates new source IDs for each resource. Update external links to Chronograf dashboards to reflect new source IDs.

  1. Stop the Chronograf server by killing the chronograf process.
  2. To prevent data loss, we strongly recommend that you back up your Chronograf data store before migrating to a Chronograf cluster.
  3. Install and start etcd.
  4. Run the following command, specifying the local BoltDB file and the etcd endpoint beginning with etcd://. (We recommend adding the prefix bolt:// to an absolute path. To specify a relative path to the BoltDB file, the prefix cannot be used.)

      $ chronoctl migrate -f bolt:///path/to/chronograf-v1.db -t etcd://localhost:2379

    Note: If you have authentication on etcd, use the standard URI format to define a username and password. For example, etcd://user:pass@localhost:2379

  5. Update links to Chronograf (for example, from external sources) to reflect your new URLs:

  6. Set up a load balancer for Chronograf.

  7. Start Chronograf.

This documentation is open source. See a typo? Please, open an issue.


Need help getting up and running? Get Support