Skip to main content

The Pulsar admin interface

The Pulsar admin interface enables you to manage all of the important entities in a Pulsar instance, such as tenants, topics, and namespaces.

You can currently interact with the admin interface via:

  • Making HTTP calls against the admin REST API provided by Pulsar brokers. For some restful apis, they might be redirected to topic owner brokers for serving with 307 Temporary Redirect, hence the HTTP callers should handle 307 Temporary Redirect. If you are using curl, you should specify -L to handle redirections.

  • The pulsar-admin CLI tool, which is available in the bin folder of your Pulsar installation:

    $ bin/pulsar-admin

    For the complete commands and descriptions of pulsar-admin, see here.

  • A Java client interface.

The REST API is the admin interface​

Under the hood, both the pulsar-admin CLI tool and the Java client both use the REST API. If you’d like to implement your own admin interface client, you should use the REST API as well. Full documentation can be found here.

In this document, examples from each of the three available interfaces will be shown.

Admin setup​

Each of Pulsar's three admin interfaces---the pulsar-admin CLI tool, the Java admin API, and the REST API ---requires some special setup if you have authentication enabled in your Pulsar instance.

If you have authentication enabled, you will need to provide an auth configuration to use the pulsar-admin tool. By default, the configuration for the pulsar-admin tool is found in the conf/client.conf file. Here are the available parameters:

webServiceUrlThe web URL for the cluster.http://localhost:8080/
brokerServiceUrlThe Pulsar protocol URL for the cluster.pulsar://localhost:6650/
authPluginThe authentication plugin.
authParamsThe authentication parameters for the cluster, as a comma-separated string.
useTlsWhether or not TLS authentication will be enforced in the cluster.false
tlsAllowInsecureConnectionAccept untrusted TLS certificate from client.false
tlsTrustCertsFilePathPath for the trusted TLS certificate file.

How to define Pulsar resource names when running Pulsar in Kubernetes​

If you run Pulsar Functions or connectors on Kubernetes, you need to follow Kubernetes naming convention to define the names of your Pulsar resources, whichever admin interface you use.

Kubernetes requires a name that can be used as a DNS subdomain name as defined in RFC 1123. Pulsar supports more legal characters than Kubernetes naming convention. If you create a Pulsar resource name with special characters that are not supported by Kubernetes (for example, including colons in a Pulsar namespace name), Kubernetes runtime translates the Pulsar object names into Kubernetes resource labels which are in RFC 1123-compliant forms. Consequently, you can run functions or connectors using Kubernetes runtime. The rules for translating Pulsar object names into Kubernetes resource labels are as below:

  • Truncate to 63 characters

  • Replace the following characters with dashes (-):

    • Non-alphanumeric characters

    • Underscores (_)

    • Dots (.)

  • Replace beginning and ending non-alphanumeric characters with 0

  • If you get an error in translating Pulsar object names into Kubernetes resource labels (for example, you may have a naming collision if your Pulsar object name is too long) or want to customize the translating rules, see customize Kubernetes runtime.
  • For how to configure Kubernetes runtime, see here.