To achieve high availability you can place multiple instances of APM Server behind a regular HTTP load balancer, for example HAProxy or nginx.

The endpoint / always returns an HTTP 200. You can configure your load balancer to send HTTP requests to this endpoint to determine if an APM Server is running. See Server information for more information on that endpoint.

In case of temporal issues, like unavailable Elasticsearch or a sudden high workload, the data is buffered in an internal memory queue and ingestion retried.

If a given APM Server process fails, the data that has not yet been forwarded to Elasticsearch is lost.