Drag the scrubber or click a tick — GET api/tminus/{index} fetches that milestone; the team cards and flow node highlight update.
▶ Play auto-advances through all nine milestones with the packet moving along the flow.
⏭ Jump to T-0 opens the 48-hour cutover detail with the supplied times (Fri 8 PM → Sun 12 PM).
🚀 Advance fleet to this point — moves servers in the backend state to match the milestone (e.g., T-6m moves 50 to validated, T-4w moves 1,000 to replicating) via POST api/servers/bulk, so other pages reflect it.
T-12 months — Discovery
Flow
T-12mDiscovery
➜
T-6mLanding zone
➜
T-4wReplication
➜
T-048-h cutover
➜
T+14dHypercare
Click a node to learn what it does.
Milestone JSON from the API appears here.
Deep dive
How
Work backwards from the lease end: fix T-0 per wave, then lay out discovery, landing zone, replication and rehearsal.
Keep a 2-week hypercare buffer before the next wave.
Why
A T-minus plan makes dependencies visible: no replication without a landing zone, no landing zone without a network design.
When
Publish at T-9 months; re-baseline after the pilot at T-6.
Pitfalls
Starting replication before Direct Connect/ExpressRoute is live (lead time is weeks).
Skipping the dress rehearsal.
Ending hypercare too early — most firewall misses appear at month-end batch runs.
Key ideas
The 48-hour cutover window
Friday 8 PM stop traffic · Saturday 2 AM final DB delta · Saturday 10 AM DNS cutover + smoke tests · Sunday 12 PM business sign-off. Everything else in the programme exists to make these 40 hours boring.