Getting Started
Three steps take you from nothing to Salesforce change events flowing into Kafka:
- Set up Salesforce — create a Connected App with the Client Credentials (or JWT Bearer) OAuth flow and enable Change Data Capture for the objects you want to stream.
- Install the plugins — download a connector ZIP from
GitHub Releases
(or build with
mvn clean package), unzip it onto your Connect worker'splugin.path, and restart the worker. - Run the quickstart — a docker-compose stack plus a ready-to-POST connector config.
Streaming Platform Events? The Platform Events walkthrough goes from creating the event in Salesforce to reading it in Kafka, with the records a real org produces and how replay, schema changes and restarts work.
Requirements
| Requirement | Version |
|---|---|
| Apache Kafka / Kafka Connect | 3.x (2.6+ for DLQ support) |
| Java (Connect worker) | 17+ |
| Salesforce edition | Enterprise, Performance, Unlimited, or Developer (Pub/Sub API) |
| Salesforce API version | v60.0 default; configurable (Bulk API 2.0 requires ≥ v47.0) |
Orgs without Pub/Sub API access (e.g. Government Cloud) can use the legacy Streaming source or the source connector's polling mode.
Verify the plugin installation
After restarting the worker, the connector classes should appear in the plugin list:
curl -s localhost:8083/connector-plugins | jq '.[].class' | grep salesforce
"sh.oso.salesforce.source.SalesforceSourceConnector"
"sh.oso.salesforce.sink.SalesforceSinkConnector"
"sh.oso.salesforce.pesink.SalesforcePlatformEventSinkConnector"
"sh.oso.salesforce.streaming.SalesforceStreamingSourceConnector"