docs(operator): add safe node-pool maintenance runbook

[sklarsa]

What

Documents a complete add-before-remove node-pool maintenance workflow for replicated QuestDB clusters:

• cordon retiring Nodes before adding capacity;
• add and validate one surge replica;
• handle a primary through planned promotion;
• drain one Node at a time under the default PDB;
• wait for the original Pod/PVC identity to re-home and recover; and
• restore the steady instance count and handle the temporary PVC.

It also documents the default hostname/zone placement behavior, scheduling overrides, same-zone capacity requirements, and safe stop/recovery points.

Why

The existing instances API already supports safe node-pool rotation through scale out → drain → scale in. A proposed spec.rotation.retire API in questdb/questdb-enterprise-operator#346 had no controller behavior and duplicated this workflow, so that PR was closed in favor of documenting the supported procedure.

Verification

• yarn build
• git diff --check

[github-actions]

🚀 Build success!

Latest successful preview: https://preview-557--questdb-documentation.netlify.app/docs/

Commit SHA: a76d4ad

📦 Build generates a preview & updates the link on each commit.