> ## Documentation Index
> Fetch the complete documentation index at: https://docs.chronosphere.io/llms.txt
> Use this file to discover all available pages before exploring further.

# HAProxy

> HAProxy connection, request, and session metrics.

The HAProxy integration requires CXDOT Collector 1.4.0 or greater.

[HAProxy](https://www.haproxy.org/) is an open source load balancer and reverse proxy
for TCP and HTTP applications. Use the HAProxy integration with the Chronosphere XDOT
Collector to collect connection, request, and session metrics from HAProxy servers
running in your environment.

The HAProxy integration supports HAProxy 3.0 or greater.

## Supported telemetry types

The HAProxy integration supports these telemetry types:

| Type    | Supported |
| ------- | --------- |
| Logs    | No        |
| Metrics | Yes       |
| Traces  | No        |
| Events  | No        |

## Prerequisites

The HAProxy integration has the following prerequisites:

* An HAProxy stats page that exposes the CSV export.
* Make the HAProxy stats page reachable from the collector.

## Configure

To configure the HAProxy integration, follow these steps:

1. Configure HAProxy to expose statistics at `/stats` on port `8404`. For example,
   add the following frontend to the HAProxy configuration:

   ```text theme={null}
   frontend stats
       bind :8404
       mode http
       stats enable
       stats uri /stats
   ```

2. For HAProxy pods that expose the stats page on port `8404`, add the
   `app.kubernetes.io/name: haproxy` label to the pod template. For instances that use
   another port or stats path, provide the stats URL through autodiscovery annotations.

   For more information, see
   [autodiscovery](https://docs.chronosphere.io/ingest/cxdot-collector/autodiscovery).

3. Optional: Configure static targets for HAProxy servers that autodiscovery doesn't
   reach. For example, add the following to the `values.yaml` for your Helm chart:

   ```yaml theme={null}
   config:
     integrations:
       haproxy:
         endpoints:
           - endpoint: haproxy.example.com:8404
   ```

### Validate

To validate the HAProxy integration, follow these steps:

1. In the Live Telemetry Analyzer, filter for `cxdot.integration.name=haproxy`.
   Confirm that the HAProxy metric names appear.

2. In Metrics Explorer, run the following query:

   ```text theme={null}
   count by ("server.address", "server.port") ({"haproxy.sessions.count"})
   ```

   Confirm that the query returns the expected time series for each target.

For more information about diagnosing a failing integration, see
[Troubleshooting](https://docs.chronosphere.io/ingest/cxdot-collector/troubleshooting).

## Configuration reference

Configure one HAProxy integration instance with the following settings. In Helm values, place
these settings under `config.integrations.haproxy`. In a Collector configuration file, place
them under `cxdot.integrations.haproxy`.

### Optional settings

* **`enabled`**
  Type: `boolean`. Optional. Default: `true`.
  Whether to enable this HAProxy integration instance. If true, the Collector collects metrics
  from HAProxy targets. If false, it doesn't run this integration instance.

* **`endpoints`**
  Type: `array of object`. Optional.
  Static HAProxy stats listeners as `host:port` values. The Collector requests the `/stats` path
  over HTTP and reads its CSV output. A nonempty list disables automatic discovery for this
  integration instance, and the Collector scrapes only the listed targets.

* **`endpoints[].endpoint`**
  Type: `string`. Required.
  HAProxy statistics listener as `host:port`. The Collector requests `/stats` over HTTP and
  reads its CSV output.

* **`collection_interval`**
  Type: `duration`. Optional. Default: `10s`.
  How often the Collector requests statistics from each HAProxy target.

* **`timeout`**
  Type: `duration`. Optional. Default: `10s`.
  Maximum time the Collector waits for an HAProxy statistics request to complete.

* **`disable_keep_alives`**
  Type: `boolean`. Optional. Default: `true`.
  Whether to close the connection after each statistics request. If true, the Collector closes
  the connection so that an idle connection doesn't increase `haproxy.sessions.count`. If false,
  the Collector keeps the connection open for reuse. Annotation-discovered endpoints instead
  take connection persistence from the annotation's `persist_connections`,
  `tls_use_host_header`, `headers`, and `extra_headers` fields.


## Related topics

- [Azure metrics available in Chronosphere Observability Platform](/ingest/xcor/integrations/server/azure/azure-metrics.md)
- [Networking](/ingest/pipeline/v2/configure/networking.md)
- [Configure Telemetry Pipeline using Helm](/ingest/pipeline/v2/configure/helm.md)
- [Coralogix destination plugin](/ingest/pipeline/plugins/destination-plugins/coralogix.md)
- [VictoriaMetrics destination plugin](/ingest/pipeline/plugins/destination-plugins/victoriametrics.md)