Start a Local Cluster (Secure)

On this page Carat arrow pointing down
Warning:
CockroachDB v22.2 is no longer supported as of June 5, 2024. For more details, refer to the Release Support Policy.

Once you've installed CockroachDB, it's simple to run a secure multi-node cluster locally, using TLS certificates to encrypt network communication.

Tip:

To deploy a free CockroachDB Cloud cluster instead of running CockroachDB yourself, see the Quickstart.

Before you begin

Step 1. Generate certificates

You can use either cockroach cert commands or openssl commands to generate security certificates. This section features the cockroach cert commands.

  1. Create two directories:

    icon/buttons/copy
    $ mkdir certs my-safe-directory
    
    Directory Description
    certs You'll generate your CA certificate and all node and client certificates and keys in this directory.
    my-safe-directory You'll generate your CA key in this directory and then reference the key when generating node and client certificates.
  2. Create the CA (Certificate Authority) certificate and key pair:

    icon/buttons/copy
    $ cockroach cert create-ca \
    --certs-dir=certs \
    --ca-key=my-safe-directory/ca.key
    
  3. Create the certificate and key pair for your nodes:

    icon/buttons/copy
    $ cockroach cert create-node \
    localhost \
    $(hostname) \
    --certs-dir=certs \
    --ca-key=my-safe-directory/ca.key
    

    Because you're running a local cluster and all nodes use the same hostname (localhost), you only need a single node certificate. Note that this is different than running a production cluster, where you would need to generate a certificate and key for each node, issued to all common names and IP addresses you might use to refer to the node as well as to any load balancer instances.

  4. Create a client certificate and key pair for the root user:

    icon/buttons/copy
    $ cockroach cert create-client \
    root \
    --certs-dir=certs \
    --ca-key=my-safe-directory/ca.key
    

Step 2. Start the cluster

  1. Use the cockroach start command to start the first node:

    icon/buttons/copy
    $ cockroach start \
    --certs-dir=certs \
    --store=node1 \
    --listen-addr=localhost:26257 \
    --http-addr=localhost:8080 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    --background
    

    You'll see a message like the following:

    *
    * INFO: initial startup completed.
    * Node will now attempt to join a running cluster, or wait for `cockroach init`.
    * Client connections will be accepted after this completes successfully.
    * Check the log file(s) for progress.
    *
    
  2. Take a moment to understand the flags you used:

    • The --certs-dir directory points to the directory holding certificates and keys.
    • Since this is a purely local cluster, --listen-addr=localhost:26257 and --http-addr=localhost:8080 tell the node to listen only on localhost, with port 26257 used for internal and client traffic and port 8080 used for HTTP requests from the DB Console.
    • The --store flag indicates the location where the node's data and logs are stored.
    • The --join flag specifies the addresses and ports of the nodes that will initially comprise your cluster. You'll use this exact --join flag when starting other nodes as well.

      For a cluster in a single region, set 3-5 --join addresses. Each starting node will attempt to contact one of the join hosts. In case a join host cannot be reached, the node will try another address on the list until it can join the gossip network.

    • The --background flag starts the cockroach process in the background so you can continue using the same terminal for other operations.

  3. Start two more nodes:

    icon/buttons/copy
    $ cockroach start \
    --certs-dir=certs \
    --store=node2 \
    --listen-addr=localhost:26258 \
    --http-addr=localhost:8081 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    --background
    
    icon/buttons/copy
    $ cockroach start \
    --certs-dir=certs \
    --store=node3 \
    --listen-addr=localhost:26259 \
    --http-addr=localhost:8082 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    --background
    

    These commands are the same as before but with unique --store, --listen-addr, and --http-addr flags.

  4. Use the cockroach init command to perform a one-time initialization of the cluster, sending the request to any node on the --join list:

    icon/buttons/copy
    $ cockroach init --certs-dir=certs --host=localhost:26257
    

    You'll see the following message:

    Cluster successfully initialized
    

    At this point, each node also prints helpful startup details to its log. For example, the following command retrieves node 1's startup details:

    icon/buttons/copy
    $ grep 'node starting' node1/logs/cockroach.log -A 11
    

    The output will look something like this:

    CockroachDB node starting at 
    build:               CCL v22.2.19 @ 2024-02-26 00:00:00 (go1.12.6)
    webui:               https://localhost:8080
    sql:                 postgresql://root@localhost:26257?sslcert=certs%2Fclient.root.crt&sslkey=certs%2Fclient.root.key&sslmode=verify-full&sslrootcert=certs%2Fca.crt
    RPC client flags:    cockroach <client cmd> --host=localhost:26257 --certs-dir=certs
    logs:                /Users/<username>/node1/logs
    temp dir:            /Users/<username>/node1/cockroach-temp966687937
    external I/O path:   /Users/<username>/node1/extern
    store[0]:            path=/Users/<username>/node1
    status:              initialized new cluster
    clusterID:           b2537de3-166f-42c4-aae1-742e094b8349
    nodeID:              1
    

Step 3. Use the built-in SQL client

Now that your cluster is live, you can use any node as a SQL gateway. To test this out, let's use CockroachDB's built-in SQL client.

  1. Run the cockroach sql command against node 1:

    icon/buttons/copy
    $ cockroach sql --certs-dir=certs --host=localhost:26257
    
  2. Run some basic CockroachDB SQL statements:

    icon/buttons/copy
    > CREATE DATABASE bank;
    
    icon/buttons/copy
    > CREATE TABLE bank.accounts (id INT PRIMARY KEY, balance DECIMAL);
    
    icon/buttons/copy
    > INSERT INTO bank.accounts VALUES (1, 1000.50);
    
    icon/buttons/copy
    > SELECT * FROM bank.accounts;
    
      id | balance
    +----+---------+
       1 | 1000.50
    (1 row)
    
  3. Now exit the SQL shell on node 1 and open a new shell on node 2:

    icon/buttons/copy
    > \q
    
    icon/buttons/copy
    $ cockroach sql --certs-dir=certs --host=localhost:26258
    
    Note:

    In a real deployment, all nodes would likely use the default port 26257, and so you wouldn't need to set the port portion of --host.

  4. Run the same SELECT query as before:

    icon/buttons/copy
    > SELECT * FROM bank.accounts;
    
      id | balance
    +----+---------+
       1 | 1000.50
    (1 row)
    

    As you can see, node 1 and node 2 behaved identically as SQL gateways.

  5. Now create a user with a password, which you will need to access the DB Console:

    icon/buttons/copy
    > CREATE USER max WITH PASSWORD 'roach';
    
  6. Exit the SQL shell on node 2:

    icon/buttons/copy
    > \q
    

Step 4. Run a sample workload

CockroachDB also comes with a number of built-in workloads for simulating client traffic. Let's run the workload based on CockroachDB's sample vehicle-sharing application, MovR.

  1. Load the initial dataset:

    icon/buttons/copy
    $ cockroach workload init movr \
    'postgresql://root@localhost:26257?sslcert=certs%2Fclient.root.crt&sslkey=certs%2Fclient.root.key&sslmode=verify-full&sslrootcert=certs%2Fca.crt'
    
    I190926 16:50:35.663708 1 workload/workloadsql/dataload.go:135  imported users (0s, 50 rows)
    I190926 16:50:35.682583 1 workload/workloadsql/dataload.go:135  imported vehicles (0s, 15 rows)
    I190926 16:50:35.769572 1 workload/workloadsql/dataload.go:135  imported rides (0s, 500 rows)
    I190926 16:50:35.836619 1 workload/workloadsql/dataload.go:135  imported vehicle_location_histories (0s, 1000 rows)
    I190926 16:50:35.915498 1 workload/workloadsql/dataload.go:135  imported promo_codes (0s, 1000 rows)
    
  2. Run the workload for 5 minutes:

    icon/buttons/copy
    $ cockroach workload run movr \
    --duration=5m \
    'postgresql://root@localhost:26257?sslcert=certs%2Fclient.root.crt&sslkey=certs%2Fclient.root.key&sslmode=verify-full&sslrootcert=certs%2Fca.crt'
    

Step 5. Access the DB Console

The CockroachDB DB Console gives you insight into the overall health of your cluster as well as the performance of the client workload.

  1. On secure clusters, certain pages of the DB Console can only be accessed by admin users.

    Run the cockroach sql command against node 1:

    icon/buttons/copy
    $ cockroach sql --certs-dir=certs --host=localhost:26257
    
  2. Assign max to the admin role (you only need to do this once):

    icon/buttons/copy
    > GRANT admin TO max;
    
  3. Exit the SQL shell:

    icon/buttons/copy
    > \q
    
  4. Go to https://localhost:8080. Note that your browser will consider the CockroachDB-created certificate invalid; you'll need to click through a warning message to get to the UI.

    Note:

    If you are using Google Chrome, and you are getting an error about not being able to reach localhost because its certificate has been revoked, go to chrome://flags/#allow-insecure-localhost, enable "Allow invalid certificates for resources loaded from localhost", and then restart the browser. Enabling this Chrome feature degrades security for all sites running on localhost, not just CockroachDB's DB Console, so be sure to enable the feature only temporarily.

  5. Log in with the username and password you created earlier (max/roach).

  6. On the Cluster Overview, notice that three nodes are live, with an identical replica count on each node:

    DB Console

    This demonstrates CockroachDB's automated replication of data via the Raft consensus protocol.

    Note:

    Capacity metrics can be incorrect when running multiple nodes on a single machine. For more details, see this limitation.

  7. Click Metrics to access a variety of time series dashboards, including graphs of SQL queries and service latency over time:

    DB Console

  8. Use the Databases, Statements, and Jobs pages to view details about your databases and tables, to assess the performance of specific queries, and to monitor the status of long-running operations like schema changes, respectively.

Step 6. Simulate node maintenance

  1. In a new terminal, gracefully shut down a node. This is normally done prior to node maintenance:

    Get the process IDs of the nodes:

    icon/buttons/copy
    ps -ef | grep cockroach | grep -v grep
    
      501  4482     1   0  2:41PM ttys000    0:09.78 cockroach start --certs-dir=certs --store=node1 --listen-addr=localhost:26257 --http-addr=localhost:8080 --join=localhost:26257,localhost:26258,localhost:26259
      501  4497     1   0  2:41PM ttys000    0:08.54 cockroach start --certs-dir=certs --store=node2 --listen-addr=localhost:26258 --http-addr=localhost:8081 --join=localhost:26257,localhost:26258,localhost:26259
      501  4503     1   0  2:41PM ttys000    0:08.54 cockroach start --certs-dir=certs --store=node3 --listen-addr=localhost:26259 --http-addr=localhost:8082 --join=localhost:26257,localhost:26258,localhost:26259
    

    Gracefully shut down node 3, specifying its process ID:

    icon/buttons/copy
    kill -TERM 4503
    
  2. Back in the DB Console, despite one node being "suspect", notice the continued SQL traffic:

    DB Console

  3. Restart node 3:

    icon/buttons/copy
    $ cockroach start \
    --certs-dir=certs \
    --store=node3 \
    --listen-addr=localhost:26259 \
    --http-addr=localhost:8082 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    --background
    

Step 7. Scale the cluster

Adding capacity is as simple as starting more nodes with cockroach start.

  1. Start 2 more nodes:

    icon/buttons/copy
    $ cockroach start \
    --certs-dir=certs \
    --store=node4 \
    --listen-addr=localhost:26260 \
    --http-addr=localhost:8083 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    --background
    
    icon/buttons/copy
    $ cockroach start \
    --certs-dir=certs \
    --store=node5 \
    --listen-addr=localhost:26261 \
    --http-addr=localhost:8084 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    --background
    

    Again, these commands are the same as before but with unique --store, --listen-addr, and --http-addr flags.

  2. Back on the Cluster Overview in the DB Console, you'll now see 5 nodes listed:

    DB Console

    At first, the replica count will be lower for nodes 4 and 5. Very soon, however, you'll see those numbers even out across all nodes, indicating that data is being automatically rebalanced to utilize the additional capacity of the new nodes.

Step 8. Stop the cluster

  1. When you're done with your test cluster, stop the nodes.

    Get the process IDs of the nodes:

    icon/buttons/copy
    ps -ef | grep cockroach | grep -v grep
    
      501  4482     1   0  2:41PM ttys000    0:09.78 cockroach start --certs-dir=certs --store=node1 --listen-addr=localhost:26257 --http-addr=localhost:8080 --join=localhost:26257,localhost:26258,localhost:26259
      501  4497     1   0  2:41PM ttys000    0:08.54 cockroach start --certs-dir=certs --store=node2 --listen-addr=localhost:26258 --http-addr=localhost:8081 --join=localhost:26257,localhost:26258,localhost:26259
      501  4503     1   0  2:41PM ttys000    0:08.54 cockroach start --certs-dir=certs --store=node3 --listen-addr=localhost:26259 --http-addr=localhost:8082 --join=localhost:26257,localhost:26258,localhost:26259
      501  4510     1   0  2:42PM ttys000    0:08.46 cockroach start --certs-dir=certs --store=node4 --listen-addr=localhost:26260 --http-addr=localhost:8083 --join=localhost:26257,localhost:26258,localhost:26259
      501  4622     1   0  2:43PM ttys000    0:02.51 cockroach start --certs-dir=certs --store=node5 --listen-addr=localhost:26261 --http-addr=localhost:8084 --join=localhost:26257,localhost:26258,localhost:26259
    

    Gracefully shut down each node, specifying its process ID:

    icon/buttons/copy
    kill -TERM 4482
    
    icon/buttons/copy
    kill -TERM 4497
    
    icon/buttons/copy
    kill -TERM 4503
    
    Note:

    For nodes 4 and 5, the shutdown process will take longer (about a minute each) and will eventually force the nodes to stop. This is because, with only 2 of 5 nodes left, a majority of replicas are not available, and so the cluster is no longer operational.

    icon/buttons/copy
    kill -TERM 4510
    
    icon/buttons/copy
    kill -TERM 4622
    
  2. To restart the cluster at a later time, run the same cockroach start commands as earlier from the directory containing the nodes' data stores.

    If you do not plan to restart the cluster, you may want to remove the nodes' data stores and the certificate directories:

    icon/buttons/copy
    $ rm -rf node1 node2 node3 node4 node5 certs my-safe-directory
    

What's next?

You might also be interested in the following pages:


Yes No
On this page

Yes No