Fleet Server on kubernetes document proposal

[eedugon]

Adding a Deploy Fleet Server on Kubernetes document, under Deployment models section.

Preview: https://ingest-docs_bk_1518.docs-preview.app.elstc.co/guide/en/fleet/master/add-fleet-server-kubernetes.html

[github-actions]

A documentation preview will be available soon.

• 🔨 Buildkite builds
• 📚 HTML diff
• 📙 Preview page

Request a new doc build by commenting

• Rebuild this PR: run docs-build
• Rebuild this PR and all Elastic docs: run docs-build rebuild

_{run docs-build is much faster than run docs-build rebuild. A rebuild should only be needed in rare situations.}

_{If your PR continues to fail for an unknown reason, the doc build pipeline may be broken. Elastic employees can check the pipeline status here.}

[mergify]

This pull request does not have a backport label. Could you fix it @eedugon? 🙏
To fixup this pull request, you need to add the backport labels for the needed
branches, such as:

• backport-/d./d is the label to automatically backport to the /d./d branch. /d is the digit
  NOTE: backport-skip has been added to this pull request.

[kilfoyle]

@eedugon I read through and this looks great!
Since the PR is in draft mode, should I add comments now or would you like me to wait before doing that?

[eedugon]

I still need to add the main kubernetes-related steps and also review all the content as it was written through a lot of iterations and text replacements :) Feel free to review and make comments, in the PR or in private. Now or at a later stage, whatever you prefer. A lot of sentences will definitely need improvement as they are currently more the ideas “I want to express” without thinking much on “how to express them”, so I’m sure your feedback will help a lot.

[bmorelli25]

I through in a few small grammar things, but generally liking the new flow!

[laraMorenoIgle]

As we have talked in private, I suggest:

• Add some context for quickstart and Production mode
• Add a link to the environment variables available for fleet server https://www.elastic.co/guide/en/fleet/current/agent-environment-variables.html

[eedugon]

@kilfoyle, after reviewing and applying the latest updates to the doc together I'd like devs to review the content from the technical side also.

@swiatekm : would you be able to take a look and share your thoughts?

I'd like your comments around the specific proposal to install the Fleet Server on Kubernetes. We are suggesting a single
Deployment backed up by 2 secrets, one for env vars, and another one for the CA certificate of Elasticsearch. The manifest doesn't need any RBAC and we disable the token to talk to Kubernetes API because the Fleet Server doesn't need it.

If you believe there's a quicker or better approach let us know. I haven't tried the automatic Kibana Setup directly by the Agent (using KIBANA_ env vars) because I've never been able to make it working (that functionality is not properly documented either).

cc: @nimarezainia : we would also like you to review this new document :)

[swiatekm]

@eedugon I did a first pass review and everything looks good to me. Before approving, I'm also going to actually follow the guide as well in my test environment to see if I spot any sharp edges that way.

One note for the future is that the CA setup seems like the most complex part of the guide. While we can't do much about this when it comes to ES itself, for the connection between Fleet Server and agent, we should be able to have the Fleet Server Helm Chart generate the necessary certs, or, even better, have cert-manager do it if it's available. WDYT @pkoutsovasilis ?

[swiatekm]

Found a few issues, comments below.

[eedugon]

Thanks a lot! I will work on that. Thanks for raising the issue when using the cert and key. I’ve realized that I didn’t test that part!!

[pkoutsovasilis]

sorry for the delay guys 👋 I will review this shortly 🙂

[eedugon]

@kilfoyle : Would you please perform a final review of the doc?

Due to some important findings by @swiatekm I had to apply some changes to the doc content to ensure Production mode fully works and is fully understood by users.

@swiatekm and @pkoutsovasilis : I'd also appreciate your review here also if you have some time, as it would be great to publish this document before the documentation freeze we will have soon.

I think this document, even if it's not perfect, it's going to provide a lot of help to users that have already tried in the past to run the fleet server on Kubernetes.

Of course if in the future we simplify things by running directly the fleet-server binary without the need of an agent enrolling to itself or if we create a helm chart to handle all this we will be able to update the document.

[kilfoyle]

LGTM! 🚀
Thanks a lot @eedugon! This is a really superb addition to our ingest docs.

[bmorelli25]

This is looking great! 🚀