Skip to main content
This section introduces how to develop and pacakge NodeJs Pulsar functions to use on StreamNative cloud. The NodeJs runtime is still in private preview stage, If you want to try it out or have any questions, please submit a ticket to the support team.

​
Develop

The StreamNative cloud will find the process method from the NodeJs function file, below is an example: 
async function process(message, context) {
  // 1. Get the logger to send logs to a Pulsar topic(when log-topic is set)
  const logger = context.getLogger()
  for (let word of message.split(' ')) {
    // 2. Incre and Get counter from state store
    await context.incrementCounter(word, 1)
    let count = await context.getCounter(word)
    logger.info(`got word: ${word} for ${count['value']} times`)
  }
  // 3. Get user-defined configurations
  let cfg = context.getUserConfigValue('configKey')
  // 4. Get secret configurations
  let sec = context.getSecret('secretKey')
  let result = `config: ${cfg}, secret: ${sec}`
  // 5. Publish messages to any Pulsar topic
  await context.publish(
    'persistent://public/default/test-node-package-serde-extra',
    result
  )
  // 6. Return the processing result
  return message.concat('!')
}
It can also handle Avro and Json schema record, for Avro scheme input and output: 
function process(params) {
  params['grade']['int'] = params['grade']['int'] + 1
  return params
}

const definitions = {
  name: 'Student',
  type: 'record',
  fields: [
    { name: 'name', type: ['null', 'string'] },
    { name: 'age', type: ['null', 'int'] },
    { name: 'grade', type: ['null', 'int'] },
  ],
}

module.exports.definitions = definitions
For Json input and output: 
function process(params) {
  params['grade'] = params['grade'] + 1
  return params
}
To get more examples, please refer to examples

​
Feature Matrix

The NodeJs runtime doesn’t support full features comparing to Java runtime, and it’s still in developing, below is the matrix:

​
Input Arguments

InputJavaGo(Pulsar)PythonNodeJs
Custom SerDe✅❌✅✅
Schema - Avro✅❌✅✅
Schema - JSON✅❌✅✅
Schema - Protobuf✅❌❌❌
Schema - KeyValue✅❌❌❌
Schema - AutoSchema✅❌❌❌
Scehma - Protobuf Native✅❌❌❌
e-2-e encryption✅❌✅✅
maxMessageRetries✅❌❌✅
dead-letter policy✅❌❌✅
SubscriptionName✅✅✅✅
SubscriptionType✅✅✅✅
SubscriptionInitialPosition✅❌✅✅
AutoAck✅✅✅✅

​
Output Arguments

OutputJavaGo(Pulsar)PythonNodeJs
Custom SerDe✅❌✅✅
Schema - Avro✅❌✅✅
Schema - JSON✅❌✅✅
Schema - Protobuf✅❌❌❌
Schema - KeyValue✅❌❌❌
Schema - AutoSchema✅❌❌❌
Schema - Protobuf Native✅❌❌❌
useThreadLocalProducers✅❌❌✅
Key-based Batcher✅✅✅✅
e-2-e encryption✅❌✅✅
Compression✅✅✅✅

​
Context

ContextJavaGo(Pulsar)PythonNodeJs
InputTopics✅✅✅✅
OutputTopic✅✅✅✅
CurrentRecord✅✅✅✅
OutputSchemaType✅❌✅✅
Tenant✅✅✅✅
Namespace✅✅✅✅
FunctionName✅✅✅✅
FunctionId✅✅✅✅
InstanceId✅✅✅✅
NumInstances✅❌✅✅
FunctionVersion✅✅✅✅
PulsarAdminClient✅❌❌❌
GetLogger✅❌✅✅
RecordMetrics✅✅✅❌
UserConfig✅✅✅✅
Secrets✅❌✅✅
State✅❌❌✅
Publish✅✅✅✅
ConsumerBuilder✅❌❌❌
Seek / Pause / Resume✅❌❌❌
PulsarClient✅❌❌❌

​
Other

OtherJavaGo(Pulsar)PythonNodeJs
Resources✅✅✅✅
At-most-once✅✅✅✅
At-least-once✅✅✅✅
Effectively-once✅❌✅❌

​
Package

You can either provide a single .js file or package your function into a .zip file when creating NodeJs functions. A zip file should contain:
  1. the pacakge.json file
  2. other source code file
Like: 
Archive:  pad.zip
 Length   Method    Size  Cmpr    Date    Time   CRC-32   Name
--------  ------  ------- ---- ---------- ----- --------  ----
     107  Defl:N       89  17% 2024-03-01 01:07 918cdb62  index.js
      94  Defl:N       78  17% 2024-03-01 01:07 2ed1db00  package.json
sidebarTitle: Develop NodeJs Functions
--------          -------  ---                            -------
     201              167  17%                            2 files
Below is an example of the package.json: 
{
  "name": "test",
  "main": "index.js",
  "dependencies": {
    "left-pad": "1.3.0"
  }
}
The “main” file is required for such package.json: 
function process(message) {
  const leftPad = require('left-pad')
  return leftPad(message, 20, 'x')
}
The StreamNative Cloud will install dependencies specified in the package.json and handle your function.

​
Deploy

After creating a cluster, set up your environment and develop&package your function, you can use the snctl, pulsarctl, pulsar-admin command, the REST API, or terraform to deploy a Pulsar function to your cluster. You can create a NodeJs Pulsar function by using a local .js or .zip file or an uploaded Pulsar functions package(recommend).

​
(Optional) Upload your function file to Pulsar

It’s recommended to upload your function file to Pulsar before you create a function. Since you can add a version suffix to the package.
Upload packages
snctl pulsar admin packages upload function://${tenant}/${namespace}/${package_name} \
--path ${file_path} \
--description "${description}" \
--properties fileName=${file_name}
You should see the following output:
The package 'function://${tenant}/${namespace}/${package_name}' uploaded from path '${file_path}' successfully

​
Create

snctl pulsar admin functions create \
--tenant public \
--namespace default \
--name function1 \
--inputs persistent://public/default/test-node-input \
--output persistent://public/default/test-node-output \
--classname exclamation \
--py function://public/default/node-exclamation@v0.1 \
--custom-runtime-options '{"genericKind": "nodejs"}' \
--sn-service-account $SERVICE_ACCOUNT
Only Pulsar 4.0.x clusters running version 4.0.10.6 or later, or Pulsar 4.2.x clusters running version 4.2.1.4 or later, support the --sn-service-account flag.You can use --as-service-account instead of --sn-service-account if you are using other versions of Pulsar clusters.
Since Pulsar doesn’t support NodeJs runtime, we need to use --py to specify the function file and specify the --custom-runtime-options '{"genericKind": "nodejs"}' to make it work.
You should see something like this:
Created function1 successfully
For details about Pulsar function configurations, see Pulsar function configurations.

​
What’s next?