This page explains the cockroach init command, which you use to perform a one-time initialization of a new multi-node cluster. For a full tutorial of the cluster startup and initialization process, see one of the Manual Deployment tutorials.
When starting a single-node cluster with cockroach start-single-node, you do not need to use the cockroach init command.
Synopsis
Perform a one-time initialization of a cluster:
cockroach init <flags>
View help:
cockroach init --help
Flags
The cockroach init command supports the following client connection and logging flags.
cockroach init must target one of the nodes that was listed with --join when starting the cluster. Otherwise, the command will not initialize the cluster correctly.
Client connection
See Client Connection Parameters for details.
Logging
By default, this command logs messages to stderr. This includes events with WARNING severity and higher.
If you need to troubleshoot this command's behavior, you can customize its logging behavior.
Examples
Usage of cockroach init assumes that nodes have already been started with cockroach start and are waiting to be initialized as a new cluster. For a more detailed tutorial, see one of the Manual Deployment tutorials.
Initialize a Cluster on a Node's Machine
SSH to the machine where the node has been started. This must be a node that was listed with
--joinwhen starting the cluster.Make sure the
client.root.crtandclient.root.keyfiles for therootuser are on the machine.Run the
cockroach initcommand with the--certs-dirflag set to the directory containing theca.crtfile and the files for therootuser, and with the--hostflag set to the address of the current node:$ cockroach init --certs-dir=certs --host=<address of this node>At this point, all the nodes complete startup and print helpful details to the standard output, such as the CockroachDB version, the URL for the DB Console, and the SQL URL for clients.
SSH to the machine where the node has been started. This must be a node that was listed with
--joinwhen starting the cluster.Run the
cockroach initcommand with the--hostflag set to the address of the current node:$ cockroach init --insecure --host=<address of this node>At this point, all the nodes complete startup and print helpful details to the standard output, such as the CockroachDB version, the URL for the DB Console, and the SQL URL for clients.
Initialize a cluster from another machine
Install the
cockroachbinary on a machine separate from the node.Create a
certsdirectory and copy the CA certificate and the client certificate and key for therootuser into the directory.Run the
cockroach initcommand with the--certs-dirflag set to the directory containing theca.crtfile and the files for therootuser, and with the--hostflag set to the address of the node. This must be a node that was listed with--joinwhen starting the cluster:$ cockroach init --certs-dir=certs --host=<address of any node on --join list>At this point, all the nodes complete startup and print helpful details to the standard output, such as the CockroachDB version, the URL for the DB Console, and the SQL URL for clients.
Install the
cockroachbinary on a machine separate from the node.Run the
cockroach initcommand with the--hostflag set to the address of the node. This must be a node that was listed with--joinwhen starting the cluster:$ cockroach init --insecure --host=<address of any node on --join list>At this point, all the nodes complete startup and print helpful details to the standard output, such as the CockroachDB version, the URL for the DB Console, and the SQL URL for clients.
