ContextQMD
Back to Influxdb

Node

content/shared/influxdb3-cli/stop/node.md

latest4.4 KB
Original Source

The influxdb3 stop node command marks a node as stopped in the catalog, freeing up the licensed cores it was using for other nodes.

[!Important] This command is designed for cleaning up the catalog after you have already stopped the physical instance. It does not shut down the running process - you must stop the instance first (for example, using kill or stopping the container).

Usage

<!--pytest.mark.skip-->
bash
influxdb3 stop node [OPTIONS] --node-id <NODE_ID>

Options

OptionDescription
--node-id({{< req >}}) The node ID to stop
--forceSkip confirmation prompt
-H--hostHost URL of the running {{< product-name >}} server (default is http://127.0.0.1:8181)
--tokenAuthentication token
--tls-caPath to a custom TLS certificate authority (for testing or self-signed certificates)
--tls-no-verifyDisable TLS certificate verification (Not recommended in production, useful for self-signed certificates)
-h--helpPrint help information

Option environment variables

You can use the following environment variables to set command options:

Environment VariableOption
INFLUXDB3_HOST_URL--host
INFLUXDB3_AUTH_TOKEN--token
INFLUXDB3_TLS_NO_VERIFY--tls-no-verify

Use case

Use this command when you have forcefully stopped a node instance (for example, using kill -9 or stopping a container) and need to update the catalog to reflect the change. This frees up the licensed cores from the stopped node so other nodes can use them.

Behavior

When you run this command:

  1. The command marks the specified node as stopped in the catalog
  2. Licensed cores from the stopped node are freed for reuse by other nodes
  3. Other nodes in the cluster see the update after their catalog sync interval (default 10 seconds)
  4. The command requires authentication if the server has auth enabled

[!Warning] Stop the instance first

This command only updates catalog metadata. Always stop the physical instance before running this command. If the instance is still running, it may cause inconsistencies in the cluster.

Examples

In the examples below, replace the following:

  • {{% code-placeholder-key %}}NODE_ID{{% /code-placeholder-key %}}: The node identifier for the stopped instance
  • {{% code-placeholder-key %}}AUTH_TOKEN{{% /code-placeholder-key %}}: Authentication token with sufficient privileges
  • {{% code-placeholder-key %}}INFLUXDB_HOST{{% /code-placeholder-key %}}: Host URL of the running {{< product-name >}} server

Clean up catalog after killing a node

This example shows the typical workflow: first stop the instance, then clean up the catalog.

<!--pytest.mark.skip-->
bash
# First, stop the physical instance (for example, using kill)
kill -9 <PID>

# Then, clean up the catalog
influxdb3 stop node --node-id NODE_ID

The command prompts for confirmation.

Clean up catalog without confirmation

<!--pytest.mark.skip-->
bash
influxdb3 stop node --node-id NODE_ID --force

Clean up catalog on a remote server

<!--pytest.mark.skip-->
bash
influxdb3 stop node \
  --host INFLUXDB_HOST \
  --node-id NODE_ID \
  --token AUTH_TOKEN

Verify node status

After stopping a node, verify its status using the influxdb3 show nodes command:

<!--pytest.mark.skip-->
bash
influxdb3 show nodes

The stopped node appears with state: "stopped" in the output.