docs(upgrades): record 2026-10-05 app wave and removals

[surrealwolf]

Summary

Docs for the 2026-10-05 app wave. Follows #45 and #46.

• docs/UPGRADE_PLAN.md
  • New Executed 2026-10-05 (app wave) section: before → after, verification, PRs, impact, backup paths, learnings, open items.
  • Items docs: refresh README/ops docs and add Palworld operator #12 (ARC, removed instead of upgraded), fix(truenas-csi): pass controller/node mode flags #18 (MongoDB/OpenSearch operators) and Upgrade TrueNAS CSI v0.18 → v1.1.2 (driver + CSI sidecars) #20 (MCP images; grafana-mcp removed in chore(grafana-mcp): remove grafana-mcp server gitops-mcp#18) marked done.
  • TrueNAS key note matches docs(upgrades): record 2026-10-04 wave and cert renewal #45: don't revoke key ID 4 (shared with truenas-csi; rotate truenas-csi to a new key first); revoke keys 1 and 2.
• docs/OPS_NOTES.md: 2026-10-05 known issues (Postgres image pins, Coder restart after a switchover, orphaned graylog webhook on nprd, Vector CEF parsing bug, high-command-ui blue/green state, operators gone) and a note about terraform state rm backup files at mode 0644.
• CHANGELOG.md: Unreleased entries for the docs and the operator removal (chore(operators): remove ARC, MongoDB and OpenSearch operators #46).

PRs covered: gitops-dev#10, gitops-mcp#19, gitops-mcp#20, gitops-tools#56, #57, #58, #59, high-command-ui#51, rancher-deploy#46 (all DataKnifeAI).

---

Note

Low Risk
Markdown and changelog updates only; no runtime, Terraform, or cluster configuration changes in this diff.

Overview
Documentation-only follow-up to the live 2026-10-05 app wave (and #46 operator removal). No Terraform or application code changes in this PR.

docs/UPGRADE_PLAN.md adds Executed 2026-10-05 (app wave) with before/after versions, verification, linked gitops PRs, measured outage windows, backup paths, and post-window learnings (Coder pubsub restart, terraform state rm backup file permissions, stuck HelmChart finalizers). The prioritized table marks #12 (ARC removed), #18 (MongoDB/OpenSearch operators removed), and #20 (MCP image bumps / grafana-mcp removal) as done, and the intro now references this wave after the 2026-10-04 infra work.

docs/OPS_NOTES.md adds a Known issues (2026-10-05) section (pinned Postgres/pgbouncer images, orphaned graylog mutating webhook on nprd, Vector CEF VRL bug, high-command-ui blue/green, operators gone) plus a Secrets hygiene note to relocate or delete world-readable terraform.tfstate.*.backup files after state rm/state mv.

CHANGELOG.md extends [Unreleased] with the same wave in Changed, documents ARC/MongoDB/OpenSearch removal under Removed, and lists the graylog webhook under Known issues.

^{Reviewed by Cursor Bugbot for commit d64cfd4. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Approved. Cursor Bugbot completed successfully and reported no findings that need human review, and no approval policy requires additional review on these documentation-only changes.

  

_{Sent by Cursor Approval Agent: Pull Request Router and Approver}