docs(design): fleet upgrades + add-a-server onboarding (CashPilot-t5ja) - #307
Conversation
…y default, sidecar self-upgrade, no stored SSH (CashPilot-t5ja) The design contract for killing worker version skew and manual server onboarding: a per-worker upgrade_mode where 'announce' (default, GitOps- safe) never touches containers and 'managed' lets the worker upgrade itself via a short-lived sidecar (a container cannot recreate itself — the deploy path force-removes the caller before creating). Upgrade orders ride the authenticated heartbeat response; identity in /data survives byte-for-byte; never crosses a major automatically; rollback on failed health check. Add-a-server becomes a wizard around a generated bootstrap command with live enrollment status — SSH-from-the- UI is examined and rejected (a UI compromise must not escalate to root on every server), scoped enrollment tokens are phase 3.
|
Warning Review limit reached
Next review available in: 45 minutes You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (2)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
1 participant
Design doc only — no behaviour change. The contract for the two chronic fleet papercuts:
Worker upgrades: per-worker
upgrade_mode(config-table pattern, like watts/dedicated).announceis the default — the existing update-banner contract (announce, never act) extended per worker, so GitOps fleets stay sovereign.managedis per-worker opt-in: upgrade orders ride the authenticated heartbeat response, and the worker upgrades itself via a short-lived sidecar (a container cannot recreate itself: the deploy path force-removes the caller before creating the replacement, andunless-stoppedcannot resurrect a removed container). Same-major only, health-gated with rollback to the previous tag,/dataidentity byte-for-byte (the per-worker key must survive). Prerequisite_find_containerlabel guard is #306.Add a server: a wizard around a generated bootstrap command (docker run + compose, prefilled, enrollment watched live via the existing state machine) — zero stored credentials. SSH-from-the-UI is examined and rejected: the UI already commands Docker sockets; holding SSH keys would escalate a UI compromise to root on every server. Scoped single-use enrollment tokens are phase 3.
Rollout is three phases, each its own bead/PR; phase 2 (managed sidecar) is gated on phase 1's announce-mode negative-control test landing in CI.
mkdocs build --strictpasses; page added to nav under research.