Contents
Contents

Event-Driven Architecture and Integration Broker

Because Integration Broker has terms like “queues,” “publishers,” and “subscribers,” it is easy to assume it can act as a general-purpose enterprise event bus.

I have seen teams make this assumption in design meetings, and it almost always leads to pain down the road when they try to scale.

Integration Broker’s async queues are highly effective for database-level staging and PeopleSoft-to-PeopleSoft data replication.

But they are not a replacement for Kafka, RabbitMQ, NATS, or cloud event meshes.

Let’s talk about where the line is between IB’s queue structure and a true event-driven architecture, and how to bridge PeopleSoft to a real message broker.


What Integration Broker’s Queues Actually Are

IB’s async machinery is designed to solve a very specific, narrow set of coordination problems:

  • PeopleSoft-to-PeopleSoft Data Replication: Pre-built services like USER_PROFILE or DEPT_SYNC exist to propagate changes between different PeopleSoft databases. The publisher emits a payload during a component save; the local queue guarantees delivery; the receiving node processes it asynchronously.
  • Inbound Webhook Staging: When an external system posts webhooks to PeopleSoft at high frequency, a synchronous REST handler can write the raw payload to a staging table and return a 200 OK immediately. An async handler drains the staging table later.
  • Explicit Routings to a Few Named Targets: You configure routings at the publisher side, mapping a service operation to specific node URLs. This works fine for a handful of known consumers.

These patterns succeed because the list of publishers and subscribers is small, static, and known up front. That is the environment IB was designed for.


Where You Need a Real Message Broker

An enterprise broker (like Kafka, NATS, or RabbitMQ) is designed for a completely different set of requirements:

  • High-Throughput Ingestion: Handling tens of thousands or millions of events per second (e.g., IoT telemetry or clickstream data). IB is sized for business transaction rates (hundreds or thousands of events per hour).
  • Event Retention and Replay: Kafka can retain messages on a topic for days. A new consumer can attach to the topic today and “replay” the last 24 hours of history. In IB, once a message is successfully processed, its lifecycle is complete; there is no archival replay log.
  • Decoupled Fan-Out: You publish an event once, and fifty independent downstream systems can subscribe to it. Each system tracks its own read index (offset) and processes at its own pace. In IB, if you add a new subscriber, you must modify the publisher’s routing configuration. True event buses invert this: subscribers attach themselves without the publisher’s awareness.
  • Consumer Load Balancing: Multiple instances of a service can form a “consumer group” to share the workload of a single topic. IB dispatches messages sequentially within a queue partition.
  • Schema Registries: Tools like Confluent Schema Registry validate event formats and manage contract versioning across independent teams. IB has no equivalent mechanism.
  flowchart TD
    subgraph event_mesh ["Event Mesh"]
        Publisher["Publisher"] -->|Sends Event| Topic["Message Broker\n(Kafka / RabbitMQ / NATS)"]
        Topic -->|Offset A| SubA["Consumer A\n(Independent)"]
        Topic -->|Offset B| SubB["Consumer B\n(Independent)"]
        Topic -->|Offset C| SubC["Consumer C\n(Independent)"]
    end

If your architecture requires any of these patterns, you need a dedicated message broker, not Integration Broker.


Bridging a Message Broker to PeopleSoft

When PeopleSoft needs to participate in an event-driven architecture, we do not try to make PeopleSoft speak the broker’s protocol directly. There are no native clients in PeopleTools for Kafka, AMQP, or MQTT.

Instead, we place a lightweight bridge component between the broker and PeopleSoft.

Inbound Events (Broker to PeopleSoft)

If PeopleSoft needs to consume events from a topic:

  • A Lightweight Microservice: Write a small Go, Node.js, or Python service that runs as a daemon. It subscribes to the broker’s topic, drains the messages, validates them, and posts them to a standard PeopleSoft REST/JSON endpoint.
  • Serverless Cloud Functions: In cloud environments, use a function (like AWS Lambda, Azure Functions, or GCP Cloud Functions) triggered directly by the queue. The function parses the event and calls PeopleSoft via HTTPS.
  • iPaaS / Middleware: Tools like MuleSoft or Boomi have native connectors for both message brokers and PeopleSoft. This is a reasonable choice if your organization already uses these platforms.

In all three setups, PeopleSoft is isolated from the broker mechanics. It simply processes standard inbound REST calls.

Outbound Events (PeopleSoft to Broker)

If external systems need PeopleSoft data published to a broker:

  • REST to Proxy: Have PeopleSoft make a standard outbound REST call to a small edge proxy service, which then publishes the message to the broker.
  • iPaaS Integration: Configure a local-to-local Integration Broker routing to an iPaaS node, and let the iPaaS relay the message to the broker.

Do not attempt to import Java client libraries for Kafka or NATS directly into PeopleCode. Even if you get the JARs to load, the transactional thread model of the application server does not match the long-lived connection pools these client libraries expect.


A Practical Rule of Thumb

If you are debating whether to use Integration Broker or a dedicated event bus, ask this question:

Will anything other than PeopleSoft need to consume this event?

  • Only PeopleSoft: Use Integration Broker’s local-to-local async queues. It is simple, transactional, and requires no extra infrastructure.
  • A couple of known external systems: Standard outbound REST calls or simple IB routings are fine.
  • Many independent consumers, or unknown future consumers: You need a real message broker. Build a bridge service to publish PeopleSoft events to the broker, and let the downstream systems subscribe there.

Building for a simple request/reply pattern and trying to turn it into an event-streaming mesh six months later is incredibly expensive. Ask this question early in your design process.


Author Profile
Chris Malek

Chris Malek is a PeopleTools® Technical Consultant with over two decades of experience. He is available for consulting engagements.

Work with Chris
Subscribe to Updates
SWS Bolt-On
PeopleSoft Simple Web Services

SWS turns SQL into production REST APIs — ready for AI, modern apps, and partner integrations. One install, unlimited potential.

  • Configuration-driven, no coding required
  • JSON, XML, and CSV output
  • Works across all PeopleSoft pillars
  • Built on 25+ years of PeopleSoft expertise
Read More & Purchase
SWS Bolt-On
PeopleSoft Simple Web Services

A powerful PeopleSoft bolt-on that makes REST web services easy. You bring the SQL, SWS handles the rest.

  • Go from idea to production in minutes
  • Zero code migrations after install
  • JSON, XML, and CSV output supported
  • No PeopleCode or Integration Broker expertise required
Read More & Purchase
SWS Bolt-On
PeopleSoft Simple Web Services

Traditional PeopleSoft web services cost $3,600–$13,000 each to develop. SWS deploys production REST APIs in under 5 minutes through configuration alone.

  • No PeopleCode or Integration Broker expertise required
  • Works across Campus Solutions, HCM, and Financials
  • Built-in pagination, caching, and nested data structures
  • Trusted by institutions across higher education and government
Read More & Purchase
SWS Bolt-On
PeopleSoft Simple Web Services

Turn PeopleSoft data into clean REST APIs for AI integrations, modern applications, and vendor data feeds. Configuration-driven — no PeopleCode required.

  • Deploy production APIs in under 5 minutes
  • AI and LLM ready (RAG, chatbots, intelligent search)
  • JSON, XML, and CSV output
  • Zero modifications to delivered PeopleSoft objects
Read More & Purchase
psLens Platform
psLens Operations & Intelligence

Look up any record, field, page, or component, audit security, and monitor Integration Broker across every database — in seconds.

  • 30+ object types browsable
  • 16 real-time alert types
  • Read-only by design
  • No App Designer or SQL required
Learn More
psLens Platform
psLens Operations & Intelligence

A web console built for the PeopleSoft community — operational monitoring, security auditing, and metadata browsing in one tool.

  • Sub-second object search
  • Catch stuck IB messages before users do
  • Audit service permissions from one screen
  • Works in any browser
Learn More
psLens Platform
psLens Operations & Intelligence

On-demand security and operational reports for your PeopleSoft environment — no client install required.

  • 14 on-demand reports
  • Markdown export for AI/LLM workflows
  • No shared tenancy
  • Built on 25+ years of PeopleSoft expertise
Learn More
psLens Platform
psLens Operations & Intelligence

Research any PeopleSoft object and monitor system health from a single browser tab — no App Designer, no SQL.

  • 30+ PeopleSoft object types browsable
  • Real-time alerts before users report problems
  • Read-only and secure
  • Private alpha — early access now
Learn More