Skip to main content
Version: 最新版

本地配置单机 Pulsar

可以通过运行单机模式 Pulsar 进行本地开发和测试。 The standalone mode includes a Pulsar broker, the necessary RocksDB and BookKeeper components running inside of a single Java Virtual Machine (JVM) process.

Pulsar in production?
If you're looking to run a full production Pulsar installation, see the Deploying a Pulsar instance guide.

Install Pulsar standalone

This tutorial guides you through every step of installing Pulsar locally.

System requirements

Currently, Pulsar is available for 64-bit macOS, Linux, and Windows. To use Pulsar, you need to install 64-bit JRE/JDK. For the runtime Java version, please refer to Pulsar Runtime Java Version Recommendation according to your target Pulsar version.

tip

By default, Pulsar allocates 2G JVM heap memory to start. It can be changed in conf/pulsar_env.sh file under PULSAR_MEM. This is extra options passed into JVM.

note

Broker is only supported on 64-bit JVM.

Install JDK on M1

In the current version, Pulsar uses a BookKeeper version which in turn uses RocksDB. RocksDB is compiled to work on x86 architecture and not ARM. Therefore, Pulsar can only work with x86 JDK. This is planned to be fixed in future versions of Pulsar.

One of the ways to easily install an x86 JDK is to use SDKMan as outlined in the following steps:

  1. Install SDKMan.
  • Method 1: follow instructions on the SDKMan website.

  • Method 2: if you have Homebrew installed, enter the following command.


brew install sdkman

  1. Turn on Rosetta2 compatibility for SDKMan by editing ~/.sdkman/etc/config and changing the following property from false to true.

sdkman_rosetta2_compatible=true

  1. Close the current shell / terminal window and open a new one.
  2. Make sure you don't have any previously installed JVM of the same version by listing existing installed versions.

sdk list java|grep installed

Example output:


| >>> | 17.0.3.6.1 | amzn | installed | 17.0.3.6.1-amzn

If you have any Java 17 version installed, uninstall it.


sdk uinstall java 17.0.3.6.1

  1. Install any Java versions greater than Java 8.

sdk install java 17.0.3.6.1-amzn

Install Pulsar using binary release

To get started with Pulsar, download a binary tarball release in one of the following ways:

After you download the tarball, untar it and use the cd command to navigate to the resulting directory:


$ tar xvfz apache-pulsar-2.10.0-bin.tar.gz
$ cd apache-pulsar-2.10.0

What your package contains

The Pulsar binary package initially contains the following directories:

DirectoryContains
binPulsar's command-line tools, such as pulsar and pulsar-admin.
confConfiguration files for Pulsar, including broker configuration and more.
Note: Pulsar standalone uses RocksDB as the local metadata store and its configuration file path metadataStoreConfigPath is configurable in the standalone.conf file. For more information about the configurations of RocksDB, see here and related documentation.
examplesA Java JAR file containing Pulsar Functions example.
instancesArtifacts created for Pulsar Functions.
libThe JAR files used by Pulsar.
licensesLicense files, in the.txt form, for various components of the Pulsar codebase.

These directories are created once you begin running Pulsar.

DirectoryContains
dataThe data storage directory used by RocksDB and BookKeeper.
logsLogs created by the installation.
tip

If you want to use builtin connectors and tiered storage offloaders, you can install them according to the following instructions:

Install builtin connectors (optional)

Since 2.1.0-incubating release, Pulsar releases a separate binary distribution, containing all the builtin connectors. To enable those builtin connectors, you can download the connectors tarball release in one of the following ways:

After you download the nar file, copy the file to the connectors directory in the pulsar directory. For example, if you download the pulsar-io-aerospike-2.10.0.nar connector file, enter the following commands:


$ mkdir connectors
$ mv pulsar-io-aerospike-2.10.0.nar connectors

$ ls connectors
pulsar-io-aerospike-2.10.0.nar
...

note
  • If you are running Pulsar in a bare metal cluster, make sure connectors tarball is unzipped in every pulsar directory of the broker (or in every pulsar directory of function-worker if you are running a separate worker cluster for Pulsar Functions).
  • If you are running Pulsar in Docker or deploying Pulsar using a docker image (e.g. K8S or DC/OS, you can use the apachepulsar/pulsar-all image instead of the apachepulsar/pulsar image. apachepulsar/pulsar-all image has already bundled all builtin connectors.

Install tiered storage offloaders (optional)

tip
  • Since 2.2.0 release, Pulsar releases a separate binary distribution, containing the tiered storage offloaders.
  • To enable tiered storage feature, follow the instructions below; otherwise skip this section.

To get started with tiered storage offloaders, you need to download the offloaders tarball release on every broker node in one of the following ways:

After you download the tarball, untar the offloaders package and copy the offloaders as offloaders in the pulsar directory:


$ tar xvfz apache-pulsar-offloaders-2.10.0-bin.tar.gz

// you will find a directory named `apache-pulsar-offloaders-2.10.0` in the pulsar directory
// then copy the offloaders

$ mv apache-pulsar-offloaders-2.10.0/offloaders offloaders

$ ls offloaders
tiered-storage-jcloud-2.10.0.nar

For more information on how to configure tiered storage, see Tiered storage cookbook.

note
  • If you are running Pulsar in a bare metal cluster, make sure that offloaders tarball is unzipped in every broker's pulsar directory.
  • If you are running Pulsar in Docker or deploying Pulsar using a docker image (e.g. K8S or DC/OS), you can use the apachepulsar/pulsar-all image instead of the apachepulsar/pulsar image. apachepulsar/pulsar-all image has already bundled tiered storage offloaders.

Start Pulsar standalone

Once you have an up-to-date local copy of the release, you can start a local cluster using the pulsar command, which is stored in the bin directory, and specifying that you want to start Pulsar in standalone mode.


$ bin/pulsar standalone

If you have started Pulsar successfully, you will see INFO-level log messages like this:


21:59:29.327 [DLM-/stream/storage-OrderedScheduler-3-0] INFO org.apache.bookkeeper.stream.storage.impl.sc.StorageContainerImpl - Successfully started storage container (0).
21:59:34.576 [main] INFO org.apache.pulsar.broker.authentication.AuthenticationService - Authentication is disabled
21:59:34.576 [main] INFO org.apache.pulsar.websocket.WebSocketService - Pulsar WebSocket Service started

tip
  • The service is running on your terminal, which is under your direct control. If you need to run other commands, open a new terminal window.

You can also run the service as a background process using the bin/pulsar-daemon start standalone command. For more information, see pulsar-daemon.

  • By default, there is no encryption, authentication, or authorization configured. Apache Pulsar can be accessed from remote server without any authorization. Please do check Security Overview document to secure your deployment.

  • When you start a local standalone cluster, a public/default namespace is created automatically. The namespace is used for development purposes. All Pulsar topics are managed within namespaces. For more information, see Topics.

Use Pulsar standalone

Pulsar provides a CLI tool called pulsar-client. The pulsar-client tool enables you to consume and produce messages to a Pulsar topic in a running cluster.

Consume a message

The following command consumes a message with the subscription name first-subscription to the my-topic topic:


$ bin/pulsar-client consume my-topic -s "first-subscription"

If the message has been successfully consumed, you will see a confirmation like the following in the pulsar-client logs:


22:17:16.781 [main] INFO org.apache.pulsar.client.cli.PulsarClientTool - 1 messages successfully consumed

tip

As you have noticed that we do not explicitly create the my-topic topic, to which we consume the message. When you consume a message to a topic that does not yet exist, Pulsar creates that topic for you automatically. Producing a message to a topic that does not exist will automatically create that topic for you as well.

Produce a message

The following command produces a message saying hello-pulsar to the my-topic topic:


$ bin/pulsar-client produce my-topic --messages "hello-pulsar"

If the message has been successfully published to the topic, you will see a confirmation like the following in the pulsar-client logs:


22:21:08.693 [main] INFO org.apache.pulsar.client.cli.PulsarClientTool - 1 messages successfully produced

Stop Pulsar standalone

Press Ctrl+C to stop a local standalone Pulsar.

tip

If the service runs as a background process using the bin/pulsar-daemon start standalone command, then use the bin/pulsar-daemon stop standalone command to stop the service. For more information, see pulsar-daemon.