Deploying a Fleet server pod in an already set up ECK env

Elastic Stack Kibana
1

We already have a fully deployed ECK env with services, elastic pods and Kibana pods.
Our data comes from on prem forwarder hosts that run Logstash, Filebeat, and Metricbeat. The goal is to replace Metricbeat with Elastic Agent, and connect it back to our ECK Fleet server.

I tried reading the instructions here and looked over the examples. I have questions.

  • I am wondering if I am misunderstanding what this manifest is for. I assume it would just deploy Fleet servers. Is that not true? Is this more a full manifest to deploy An elastic stack where elastic has a fleet server integrated?
  • All the examples include deploying Kibana. We already have Kibana deployed. Is there a way to just only deploy the fleet server?
  • Once I have the fleet server deployed how to integrate it to Elastic + integrate our on prem hosts with elastic agent to it?

I apologize ahead of time if I am completely misunderstanding how this manifest works.

2

Those examples provide a comprehensive setup to demonstrate how each component of the stack integrates seamlessly.

If you already have Elasticsearch and Kibana running on ECK, adding Fleet servers should be straightforward. You can simply reference them using kibanaRef and elasticsearchRefs, specifying their respective names and namespaces. Ensure that Kibana is configured with the necessary agent policies for the Fleet server.

However, if the stack components are running outside of ECK, you’ll likely need to configure more settings manually—tasks that ECK typically handles automatically. It’s probably a good idea to familiarize yourself with the relevant Fleet documentation, such as the Elastic Fleet Environment Variables guide, if you haven't already done so.

The same consideration applies when connecting Elastic Agents running outside of ECK to a Fleet server hosted on ECK.