jason.woltje
3ae1411d5f
Addresses CPS-PRD-REV1-REREVIEW findings on PR #1491 head 13968e9a:
F1 (blocker; the CI test failure): docs/PRD.md is now the permanent shim, so
heading-fragment links into the former PRD body no longer resolve —
fleet-documentation.spec.ts failed on docs/fleet/README.md's FCM anchor. The
Part II workstream contracts live in
docs/PRDs/2026-08-31_PRD_rev1/GOV.4-workstream-contracts.md; all five anchored
links in the tree are repointed there (fleet/README.md, TASKS.md,
release-integrity/TASKS.md, ADMIN-GUIDE mos-connector-lease-operations.md,
DEVELOPER-GUIDE mos-runtime-portability-m1.md). Each anchor verified against
the GOV.4 headings with the spec's slug rule. Bundle content unchanged.
F2 (minor): the rev0 archive README's ./ROADMAP.md row now cites all three
source lines (18, 177, 249).
Fleet Configuration Management
This book documents the local roster-v2 desired-state control plane delivered under issue #758. The normative requirements are the FCM section of the repository PRD, not the older fleet-suite or observability planning pages.
Authority boundary
<MOSAIC_HOME>/fleet/roster.yaml is the sole writable desired-state authority for local fleet membership, launch policy, and persisted lifecycle. Generated environment files, systemd enablement, tmux sessions, heartbeat files, and status output are derived or observed. Rebuild projections from the roster; never edit them as desired state.
This control plane is local tmux/systemd only. Remote/SSH entries and connectors are inventory, not reconciliation targets. Arbitrary commands, channels, secret references, gateway catalog convergence, and UI configuration storage are outside this workstream. mos-comms is temporary transport glue, not permanent fleet architecture.
Choose the right workflow
- Need to inspect intent? Read the roster and use mosaic fleet get; see desired versus observed state.
- Need to inspect reality? Use
statusordoctor; useverifyfor a strict non-zero drift/ownership gate. These commands do not repair anything. - Need to change membership or persisted policy? Use generation-guarded
plan,create,update, ordelete; see safe CRUD. - Need a one-time runtime action? Use
start,stop, orrestart. These do not change persisted desired state. - Need convergence? Review apply --dry-run, resolve blockers, then use
applywith the same current generation; see reconcile and recover. - Need v1 migration evidence? Use preview only. Cutover, canary, and rollback remain held for FCM-M4-002.
- Need the gateway-backed agent catalog? That is the separate mosaic agent surface, not local fleet desired state.
Concepts
- Desired versus observed state
- Identity, class, runtime, provider, and model
- Role authority and leases
- Generated environment launch chain
Operator how-to
- Create, inspect, update, and delete
- Start, stop, restart, and reconcile
- Configure an interaction instance
- Configure a validator instance
- Customize roles
Operations and recovery
- Reconcile and recover
- Environment quarantine
- Systemd/tmux troubleshooting
- Backup and restore boundary
- Upgrade and asset-drift hold
Reference and migration
- Roster v2 fields · executable JSON Schema · validated example
- CLI and exit codes
- Role classes
- Lifecycle transitions
- Status and drift
- Generated environment boundary
- v1-to-v2 preview
- Example/profile dispositions
- Legacy class aliases
Acceptance evidence and holds
- M0/M5 IA checklist
- Legacy example/profile inventory
- M5 closure evidence
- Approved-existing deferrals and live-action holds
The canonical publishing source remains this repository. This card does not publish externally, run a migration, operate a live fleet, or close parent issue #758.