> For the complete documentation index, see [llms.txt](https://docs.ilert.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.ilert.com/developer-docs/client-libraries/ilagent/kafka.md).

# Kafka Consumer

Consuming Apache Kafka messages and forwarding them to ilert as events or heartbeats.

## Connection

Connect to your Kafka bootstrap broker(s):

```sh
ilagent daemon --kafka_brokers localhost:9092 --kafka_group_id ilagent \
    -e 'events-topic' -r 'heartbeats-topic'
```

| Flag               | Description                                       | Default      |
| ------------------ | ------------------------------------------------- | ------------ |
| `--kafka_brokers`  | Comma-separated broker list (required)            | —            |
| `--kafka_group_id` | Consumer group ID (required)                      | —            |
| `-e`               | Topic for event messages                          | —            |
| `-r`               | Topic for heartbeat messages                      | —            |
| `-p`               | HTTP server port (for health/readiness endpoints) | *(disabled)* |

{% hint style="info" %}
Kafka mode does not start the HTTP server by default. Pass `-p` explicitly if you need `/health` and `/ready` endpoints (e.g. for Kubernetes probes).
{% endhint %}

## Graceful shutdown

The Kafka consumer shuts down gracefully on SIGINT/SIGTERM — in-flight messages are finished processing and offsets are committed before exit. If the HTTP server is running (`-p`), the `/health` endpoint returns 503 during shutdown to allow load balancers or Kubernetes to drain traffic.

## At least once delivery

In Kafka mode, events are **not** buffered in SQLite. Instead, the consumer offset is used to provide at-least-once delivery. If an event cannot be delivered, the offset is not committed and ilagent will exit after 5 seconds.

We recommend running ilagent so it is automatically restarted on exit:

```sh
docker run --restart always ilert/ilagent daemon \
    --kafka_brokers localhost:9092 --kafka_group_id ilagent \
    -e 'events-topic'
```

## Payload format

All message payloads must be JSON. An example event payload using the default field names:

```json
{
  "integrationKey": "il1api123...",
  "eventType": "ALERT",
  "summary": "Hey from Kafka"
}
```

If your payloads use different field names, you can map them with `--map_key_etype`, `--map_key_summary`, `--map_key_alert_key`, and set a fixed integration key with `--event_key`. See [Event Mapping](/developer-docs/client-libraries/ilagent/event-mapping.md) for details.

{% hint style="warning" %}
The `integrationKey` (or `--event_key`) must belong to a [**Kafka** alert source](/integrations/inbound-integrations/kafka.md) in ilert. Events sent with integration keys from other alert source types will be rejected.
{% endhint %}

An example heartbeat payload:

```json
{
  "integrationKey": "il1hbt123..."
}
```

## Sample command

```sh
ilagent daemon -v -v \
    --kafka_brokers localhost:9092 \
    --kafka_group_id ilagent \
    -e 'events-topic' -r 'heartbeats-topic'
```

See [Event Mapping](/developer-docs/client-libraries/ilagent/event-mapping.md) for mapping custom message fields to ilert event properties.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.ilert.com/developer-docs/client-libraries/ilagent/kafka.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.