the-custodian/docs/state-hub-railiance01-cutover-phase2-2026-07-06.md

T05 Phase 2 Evidence — Data Migration to railiance01 (2026-07-06)

Workplan: CUST-WP-0054-T05
Parent runbook: docs/state-hub-railiance01-cutover-plan.md
no_secret_material_recorded: true

Checklist

Step	Action	Result	Evidence
2.1	Pre-freeze pg_dump (coulombcore)	pass	/tmp/state-hub-prefreeze-20260706T165001Z.dump (3.0 MiB)
2.2	Freeze coulombcore writer	pass	deployment/state-hub scaled to 0
2.3	Final pg_dump after freeze	pass	/tmp/state-hub-final-20260706T165019Z.dump (3.0 MiB)
2.4	Restore into railiance01 state-hub-db	pass	DROP/CREATE state_hub + pg_restore --role=state_hub
2.5	Exact-count comparison	pass	All 7 baseline tables match ±0
2.6	Scale railiance01 API + verify	pass	Forgejo main-d1a4fcf; /state/health ok

Phase 2 status: complete. Phase 3 (rewire access paths) not started.

Baseline counts (pre-freeze gate)

Captured from coulombcore state-hub-db-1 @ 2026-07-06T16:50Z:

Table	Count
workplans	640
tasks	4002
topics	14
progress_events	8723
token_events	1936
managed_repos	74
decisions	79

Post-restore counts on railiance01 matched exactly.

Production state after Phase 2

Cluster	API	DB writer
railiance01	Running 1/1 (forgejo…:main-d1a4fcf)	primary (restored data)
coulombcore	Scaled to 0 (frozen, rollback-capable)	Unchanged CNPG (source dump retained)

Workstation http://127.0.0.1:8000 still tunnels to coulombcore — unreachable until Phase 3.

API verification (port-forward 18081:8000)

{"status":"ok","db":"connected"}

GET /state/summary totals: topics=14, workstreams=640, tasks=4002, decisions=79.

Rollback

• coulombcore dump files retained under /tmp/state-hub-*-20260706T*.dump on workstation
• Scale coulombcore state-hub back to 1; do not delete railiance01 data until Phase 5 stabilizes

Next: Phase 3

• Repoint state-hub-primary tunnel → railiance01 ClusterIP 10.43.68.154:8000
• activity-core in-cluster URL; retire fleet-state-hub-coulombcore
• Smoke POST /progress/ through production path