What is event-driven programming?
Event-driven programming (EDP) is a programming paradigm where external events determine the flow of program execution. These events come in different shapes: user actions (e.g., button clicks, keyboard inputs), system events (like a finished file download), messages from other programs, sensor outputs, etc.
Unlike procedural and sequential programming (which generally follow predefined execution steps or sequences), in EDP, the program waits (listens) for events and reacts to them as they occur, by using appropriate event-handling procedures. This allows for greater flexibility and responsiveness in applications.
It’s worth noting that event-driven programming is often asynchronous, and it’s particularly popular in graphical user interfaces (GUIs), real-time apps, and IoT use cases.
EDP typically goes hand-in-hand with event-driven architectures (EDA). While EDA defines how multiple (and different) programs interact and communicate via events, EDP provides the means to implement the specific logic and behaviors within a program in response to events. That’s not to say that all components that comprise an EDA need to follow the EDP paradigm; however, adopting event-driven programming where suitable can lead to more efficient and responsive systems within the overarching EDA. I won’t go into more details here, but please see the “Event-oriented architecture: e-commerce example” section towards the end of this article to better understand how EDP works as a part of a wider EDA.
If you’re here because you’re planning to build an event-driven application, I recommend the “Guide to the Event-Driven, Event Streaming Stack,” which talks about all the components of EDA and walks you through a reference use case and decision tree to help you understand where each component fits in.
Key concepts of event-driven programming
This section walks you through the core components involved in event-driven programming.
An event refers to a significant occurrence or change in state within a system or application. In simpler words, an event records that something has happened. Here are some common examples of events:
- A user clicks a button in the UI.
- Someone makes an online payment.
- A user sends a chat message.
- A smartwatch detects an irregular heartbeat.
- A sensor reports a vehicle’s location and speed has changed.
- A record is inserted, updated, or deleted in a database.
- An application is starting or shutting down.
- A warning that disk space is running low.
Here’s an event that signifies a vibration recorded by a sensor installed on a machine used in manufacturing:
In this example:
- eventId is the unique identifier specific to this vibration recording.
- eventType specifies the type of event.
- timestamp indicates the exact date and time when the event occurred.
- source and sensorLocation indicate the origin of the event.
- data contains the actual readings from the sensor and other relevant metrics.
- metadata provides additional information about the sensor.
- systemMessage is a human-readable alert or notice about the event.
Collecting such telemetry events enables the manufacturing company to monitor its machines in real time and detect issues in production lines as soon as they occur. This way, the manufacturer can quickly react and follow the appropriate course of action (i.e., investigate the machine and perform maintenance, if needed).
Producers and consumers
Producers are sources of events in a system. They generate events to signify that something of interest has occurred. In the example provided in the previous section of this article, the sensor that measures vibrations is the event producer.
Meanwhile, consumers listen for events generated by producers. To continue our example, if the vibration sensor is the producer, the consumer could be, for instance:
- A monitoring system designed to analyze and visualize real-time vibration data, history, and trends.
- An alerting system that sends automated alerts via email, SMS, or other means to maintenance teams.
- A database or data storage system that stores the vibration readings for historical analysis, reporting, or compliance purposes.
- All of the above (you can have multiple and different types of components/systems consuming an event).
It’s also worth mentioning that, depending on the use case, an event producer can also be a consumer. Think, for example, of chat apps. If you send a chat message (an event), you are the producer. On the other hand, if you receive a chat message, you are the event consumer.
Event listeners, event handlers, and the event loop
In the snippet above, we’re attaching an event listener to a button (‘myButton‘). The addEventListener method listens for a ‘click‘ event and associates it with the handleButtonClick callback function.
handleButtonClick is the event handler — the actual code that gets executed in response to the ‘click‘ event. Here’s a possible implementation:
On button click, this trivial event handler will display an alert to the user saying ‘Hello world!‘
Note that you will often see these two terms (event listener and event handler) being used interchangeably. However, as we have just seen, there is a distinction between them:
- The event listener listens for events and specifies which function to call when an event occurs.
- The event handler is the actual callback function that gets executed in response to the event.
Being aware of this distinction is important for several reasons:
- More efficient troubleshooting. When debugging, it's essential to know if an issue lies in the event listening mechanism (the event isn't being captured) or the event handler (the response isn't as expected).
- Flexibility and modularity. Multiple event listeners can trigger the same event handler, and vice versa, a single event listener can trigger different event handlers based on certain conditions. Grasping the distinction between handlers and listeners allows for more flexible and modular coding practices.
- Clarity in collaboration. It’s crucial to have clear, shared terminology at a team level. Misunderstandings can arise if one person is discussing a handler while someone else believes they’re talking about a listener.
- Event listeners detect incoming events and place them in an event queue.
- The event loop mechanism continually checks for queued events and calls the corresponding event handler for each event.
One of the main benefits of the event loop is that it enables concurrent execution and non-blocking I/O operations when you’re working with single-threaded programming languages and runtimes.
Queues and topics
When dealing with high volumes of data and numerous producers and consumers in a distributed environment, it’s common practice to use some sort of messaging middleware to intermediate the flow of events between different components. We’ll talk about the different types of messaging middleware later in this article; for now, suffice to say that messaging middleware technologies generally use topics and message queues to route events from producers to consumers:
- Message queues. A message queue temporarily stores messages (events) generated by producers, ensuring they are processed in a specific order, often first-in-first-out (FIFO) by a consumer. As events occur, they are added to the back of the queue. Once a consumer processes the event, it’s deleted from the queue. There can be multiple producers publishing events to a queue; however, on the consumer side, there is a single entity consuming each message (point-to-point messaging). Note: do not confuse message queues with event queues I mentioned earlier when talking about the event loop. While an event queue helps manage and execute tasks (like functions or callbacks) in an event-driven application, a message queue helps manage and transmit events between different instances, applications, and systems.
- Topics. A topic is a communication channel specific to publish-subscribe (pub/sub) messaging systems. Producers publish events to topics, and multiple consumers can subscribe to topics to receive events. Unlike queues, events sent to a topic can be consumed by multiple consumers, (one-to-many communication). Some messaging middleware solutions allow you to retain events in topics for long periods of time, even after they were delivered to consumers (this is not generally the case with queues).
Message queues and topics bring many benefits, such as:
- Decoupling. They introduce a level of decoupling between producers and consumers. This means changes in one won't directly impact the other. Decoupling allows different parts of the system to scale, evolve, and be maintained independently.
- Parallel processing. Multiple consumers can process events concurrently, optimizing throughput, reducing processing times, and improving responsiveness.
- Asynchronous communication. Queues and topics allow non-blocking data exchange between producers and consumers. This means producers aren't held up waiting for a response, which leads to efficient workflows and improved system performance.
- Load leveling. Queues and topics act as buffers, preventing sudden bursts from overwhelming processing units, and ensuring smooth operation.
- Reliable event delivery. Queues generally guarantee the order of delivery, ensuring that sequences of events are processed in the correct chronological order. Meanwhile, topics can guarantee ordering within specific conditions or configurations (e.g., at a partition level, or across events that share the same key). Additionally, topics and queues offer guarantees around event delivery, even during system failures — unprocessed messages can be resent, which helps prevent data loss.
Advantages and challenges of event-driven programming
EDP brings significant advantages to developers and organizations. Here are the key ones:
- Responsive applications. When an event occurs, predefined handlers or callbacks are invoked instantly. This ensures that user inputs or system triggers receive prompt feedback, in real time. In addition, many EDP apps use asynchronous, non-blocking operations, especially for I/O tasks. In other words, instead of waiting for a task to complete, the app can continue processing other events. This ensures the app remains responsive even if it’s handling potentially time-consuming operations.
- Concurrency and scalability. Event-driven systems can process more than one event at a time, which allows them to deal with numerous simultaneous user interactions or system events. Events can be distributed and processed by multiple consumers in parallel. This means it’s possible to handle large volumes of events without relying on a single consumer, who could be a bottleneck that slows down the system.
- Modularity. Event-driven programming creates a framework where software components are modular, with clearly defined responsibilities (for instance, each type of event is usually handled by a corresponding event handler). Due to this separation of concerns, you can add, remove, modify, and reuse components without affecting other parts of the system.
- Efficient resource usage. By reacting only when events occur, event-driven systems avoid the wastefulness and overhead of constant polling. This leads to more efficient CPU and memory utilization (on-demand resource usage).
While EDP offers plenty of benefits, it also comes with challenges, such as:
- Code complexity. Asynchronous code can be hard to understand and debug. In addition, layering multiple events and callbacks can lead to convoluted code.
- Testing is non-trivial. Emulating real-world event sequences for testing can be difficult.
- Ordering issues. Events may not always execute in the desired order. Race conditions can cause unpredictable results, data corruption, and system instability.
- Steep learning curve. Developers who are new to EDP may experience a steep learning curve, particularly when learning complex topics like concurrency behaviors.
Event-driven programming use cases
There are numerous real-world applications for EDP. Here are the most popular use cases:
Event-driven programming technologies
The rise of real-time applications, and the need for scalable, responsive systems has led to the emergence and adoption of a wide range of technologies optimized for event-driven programming and event-driven architectures.
Push-based protocols are generally a good fit for event-driven programming, because they ensure that as soon as an event is generated, it's immediately delivered to the consumer without waiting for them to request it. This leads to reduced latency and efficient resource utilization (there’s no need for constant polling, which is computationally expensive).
- Examples: WebSockets, Server-Sent Events (SSE), Message Queuing Telemetry Transport (MQTT), Advanced Message Queuing Protocol (AMQP).
Event streaming platforms
An event streaming platform is a system designed to handle the continuous flow of data, allowing for the production, persistent storage, and consumption of streams of events in real time. Event streaming platforms are the backbone of event-driven programming in distributed environments, bridging event producers with consumers efficiently, and allowing them to react to events as soon as they occur. Event streaming platforms can generally handle very high volumes of streaming data, allowing the system to scale up as the event load increases.
- Examples: Apache Kafka, Apache Pulsar, Amazon Kinesis, Redpanda.
Learn how some of these event streaming platforms compare:
Stream processing solutions
Stream processing complements event streaming. While event streaming platforms continuously collect, store, and deliver events from producers to consumers, stream processing is a complementary technology that allows you to analyze and transform (e.g., aggregation, windowing, deduplication, joining) streams of events in real time.
- Examples: Quix Streams, Kafka Streams, Amazon Kinesis Data Analytics, Google Dataflow, Apache Beam, Apache Spark, Apache Flink.
See how some of these stream processing solutions compare:
- Apache Beam vs. Apache Spark
- Apache Flink vs. Kafka Streams
- Apache Spark vs. Apache Flink vs. Quix Streams
You can think of an event bus as a routing middleware that receives events and delivers them to one or more destinations. Event buses are generally well-suited for routing events from many sources (producers) to many targets (consumers). Unlike event streaming platforms, event buses usually deliver events in a fire-and-forget way (data isn’t stored for long periods of time). Another difference is that event buses allow for better, more efficient event filtering and routing compared to event streaming platforms. The tradeoff is that event buses are not generally geared toward the same levels of scalability and huge data volumes as event streaming solutions.
- Examples: Amazon EventBridge, Eventarc, Azure Event Grid.
Message queuing services
Message queues help you implement asynchronous, point-to-point communication. A queue temporarily stores events generated by producers, ensuring they are processed in a specific order, often first-in-first-out (FIFO) by a consumer. As events occur, they are added to the back of the queue. Once an event is processed by a consumer, it’s deleted from the queue. Message queuing services are especially vital in environments where multiple events can occur simultaneously or nearly so, and they must be processed in a specific order.
- Examples: RabbitMQ, ActiveMQ, Amazon SQS, Azure Queue Storage.
If you’re curious to see how event queuing services compare to event streaming platforms, check out the following:
Pub/Sub messaging services
Although they both fall within the pub/sub paradigm, a pub/sub messaging service differs from an event bus. An event bus usually includes centralized routing rules based on the content of each event (message), whereas a pub/sub messaging service does not. Instead, a pub/sub service broadcasts the same event to all subscribers of a specific topic. Message routing is managed through decentralized subscriptions to topics rather than centralized rules, and there usually isn't any in-built filtering based on message content.
Note that event streaming platforms like Apache Kafka also belong to the pub/sub paradigm, but can support more advanced use cases out of the box. In contrast, pure pub/sub messaging services mainly focus on one simple task: delivering messages reliably from producers to consumers.
- Examples: Azure Service Bus, Amazon SNS, Google Pub/Sub.
An event store is a specialized database or storage system optimized for storing and querying event-driven data. Unlike traditional databases that store the current state of data entities, an event store captures the full series of state-changing events over time. Each event in the store represents a state transition, carrying information about the change, its cause, and a timestamp. This sequential log of events enables event sourcing, a pattern where the application state is reconstructed by replaying the events.
- Examples: DynamoDB, Google Datastore, Azure CosmosDB, EventStoreDB, Axon Server.
Event-oriented architecture: e-commerce example
At the beginning of this article, I mentioned that event-driven programming often goes hand-in-hand with event-driven architecture. We’ll now look at an example to see how these two concepts work together to deliver a system that reacts to events as they occur, in real time.
We’ll use an online shop as an example. When customers (event producers) submit orders, OrderPlaced events are generated. An event streaming platform then ingests these events.
The event streaming platform then sends OrderPlaced events to:
- An event store, for long-term storage. This is useful for event sourcing, analytics (e.g., analyzing historical order data to understand buying patterns), and auditing.
- A payment service, which consumes the events to process customer payments.
The payment service has an event listener (OrderListener), which listens for events and triggers an event handler (PaymentHandler) to initiate the payment process. Once a payment is processed, the service emits a PaymentProcessed event, which is sent to the event streaming platform, and from there, forwarded to the customer, as payment confirmation.
We then have another component, a warehouse service. This service consumes PaymentProcessed events from the event streaming platform. Similar to the payment service, the warehouse service has an event listener (ShipmentListener) that listens for events. The event listener triggers an event handler (ShipmentHandler) which initiates the shipping process and sends a notification to the customer, so they are aware their order will soon be underway.
In this architecture, as events flow through the system, each service can work asynchronously, improving scalability and resilience. If, for instance, the payment service goes down, orders can still be placed. Once the payment service is back, it can pick up from where it left off, ensuring no order is missed — the event streaming platform can store OrderPlaced events as long as needed for the payment service to consume them (plus, they are also stored in the event store).
Bear in mind that this is a simplified architecture — the point was to showcase how EDA and EDP work together in an easy-to-understand way. In a real-life scenario, the architecture would be more complex. We’d likely have:
- Additional services that follow the EDP paradigm (e.g., fraud detection service, inventory checker service, recommendation engine to send recommendations to shoppers).
- More types of events (for instance, FraudCheckCompleted to signal the result of a fraud detection process, or InventoryChecked events, which are fired after verifying product availability). Additionally, all types of events could be pushed to the event store for long-term storage.
- Extra components, such as a stream processing solution (it would allow you, for instance, to join OrderSubmitted and InventoryChecked events to ensure stock is available before initiating the payment process).
Simplify event-driven development with Quix
Quix enables you to develop, release, and observe scalable event-driven apps without having to worry about complex infrastructure and configuration. Under the hood, Quix combines containerized microservices with Docker, Kafka, and Git — all of this is fully managed, so there’s no infrastructure provisioning, setup, and maintenance nightmare for you to deal with. Instead, you can focus entirely on writing your apps and deploying them to production. Among other features, Quix offers:
- Open source connectors, so you can easily ingest events from different sources and stream them to various destination systems.
- Online IDE and open source code samples.
- Stream processing capabilities.
- Multi-environment support, baked in CI/CD, and Git integration.
- Built-in observability, e.g., dashboards to monitor metrics and logs, a graphical snapshot of your app’s architecture.
To learn more about what Quix can do for you, check out the official documentation.
What’s a Rich Text element?
The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.
Static and dynamic content editing
A rich text element can be used with static or dynamic content. For static content, just drop it into any page and begin editing. For dynamic content, add a rich text field to any collection and then connect a rich text element to that field in the settings panel. Voila!
How to customize formatting for each rich text
Headings, paragraphs, blockquotes, figures, images, and figure captions can all be styled after a class is added to the rich text element using the "When inside of" nested selector system.
Tomas Neubauer is Co-Founder and CTO at Quix, responsible for the direction of the company across the full technical stack, and working as a technical authority for the engineering team. He was previously Technical Lead at McLaren, where he led architecture uplift for Formula One racing real-time telemetry acquisition. He later led platform development outside motorsport, reusing the knowhow he gained from racing.