Realtime & Change Data Capture (CDC)
EntityTS includes a built-in Change Data Capture (CDC) engine (CdcEmitter) to stream database mutations in real time across services, websockets, and event consumers.
What is CDC?
Change Data Capture intercepts data modifications at the database or ORM lifecycle layer and emits structured change events containing:
- Operation:
INSERT,UPDATE, orDELETE - Table / Collection: The entity table being modified
- Before State: The previous state of the row/document (for updates and deletes)
- After State: The updated or created row/document
- Timestamp & Transaction ID: For ordering and idempotency
Subscribing to CDC Events
Use CdcEmitter to subscribe to table events or bridge mutations to an EntityEventBus:
import { CdcEmitter, EntityEventBus } from 'entityts';
const eventBus = new EntityEventBus();
const cdc = new CdcEmitter(eventBus);
// Listen to all INSERTs into the 'orders' table
cdc.subscribe({ table: 'orders', operation: 'INSERT' }, async event => {
console.log(`New order placed: ${event.after.orderNumber}`);
await sendConfirmationEmail(event.after);
});
// Listen to all UPDATEs across all tables
cdc.subscribe({ operation: 'UPDATE' }, async event => {
console.log(`Table ${event.table} updated:`, {
before: event.before,
after: event.after,
});
});
Publishing CDC Events
When performing mutations, publish CDC events directly or configure your DbContext hooks / Unit of Work to emit them automatically:
await cdc.publish({
table: 'orders',
operation: 'UPDATE',
before: { id: 101, status: 'pending' },
after: { id: 101, status: 'shipped' },
timestamp: new Date(),
txId: 'tx_8892',
});
Integration with Transactional Outbox
For distributed microservices, pair CdcEmitter with the Transactional Outbox to guarantee at-least-once event publishing without dual-write inconsistencies.