dohertj2/ScadaBridge
1
0
Fork 0
Code Issues 2 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

docs(grpc): Phase 5 live gate — PASS (8/8), migration complete

Browse Source 
Full eight-check gate on the Phase 4 deletion build (main @ 7fd5cb2b),
rig rebuilt from main. All 8 checks PASS: PSK negatives; site->central
matrix (notif no-loss/dupe, both audit paths, reconcile self-heal);
central->site matrix (Query/Parked/Lifecycle); active-central kill ->
sticky flip 1s + central-b active 26s; mid-drain total==distinct;
297,574-byte gRPC reply (128KB frame-class retired); cluster membership
pair-only (no cross-boundary Akka association); full-rig restart 0
receptionist/ClusterClient lines on any of 8 nodes.

Check-1 clarification recorded: site-side interceptor uses the node's
single GrpcPsk and ignores x-scadabridge-site (a central-side routing
hint) -- per-site isolation proven by wrong-key reject. Instance-
dependent central->site RPCs (OpcUa/Route/standby-parked/TriggerFailover)
carried forward, unit-proven.
This commit is contained in:
Joseph Doherty
2026-07-23 13:35:16 -04:00
parent 7fd5cb2b56
commit 9f91e84d83
3 changed files with 150 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/plans/2026-07-22-clusterclient-to-grpc-plan.md
+3 -2
View File
@@ -331,8 +331,9 @@ Critical path ≈ 1B: **~4–6 weeks total**, matching the design estimate.
- [x] Grep-gates pass (`CentralContactPoints` → only "replaces the former" doc refs + plan trackers; deleted symbols → 0 live refs; remaining `clusterclient` in src = the misleadingly-named `ClusterClientSiteAuditClient` [transport-agnostic, works unchanged] + stale inline doc-comments, noted as follow-up)
- [x] Docs updated: `Component-Communication.md`, `components/Communication.md`, `Component-Host.md`, `Component-StoreAndForward.md`, `topology-guide.md`, `grpc_streams.md`, known-issues frame-size amendment, **CLAUDE.md** transport decisions
**Phase 5 — live gate** (record in `2026-07-22-clusterclient-to-grpc-live-gate.md`)
- [ ] 1 PSK negatives · [ ] 2 site→central matrix · [ ] 3 central→site matrix · [ ] 4 failover/failback both directions · [ ] 5 mid-drain kill · [ ] 6 >128 KB frame-class proof · [ ] 7 no cross-boundary Akka association · [ ] 8 full-rig restart clean
**Phase 5 — live gate** (record in `2026-07-22-clusterclient-to-grpc-live-gate.md`) — **PASS 2026-07-23** (deletion build, `main` @ `7fd5cb2b`)
- [x] 1 PSK negatives · [x] 2 site→central matrix · [x] 3 central→site matrix · [x] 4 failover/failback both directions · [x] 5 mid-drain kill · [x] 6 >128 KB frame-class proof · [x] 7 no cross-boundary Akka association · [x] 8 full-rig restart clean
- All 8 PASS. Notables: check 1 clarified the site-side gate uses the node's single `GrpcPsk` and ignores `x-scadabridge-site` (central-side routing hint) — per-site isolation proven by wrong-key reject, not the header; check 6 = a **297,574-byte** single gRPC reply (a payload Akka's 128 KB frame dropped); check 7 = Akka cluster membership strictly pair-only, 0 cross-boundary association; check 8 = full-rig restart, 0 receptionist/ClusterClient lines on any of the 8 nodes. Instance-dependent central→site RPCs (`ExecuteOpcUa`/`ExecuteRoute`/standby parked retry/`TriggerFailover`) carried forward unit-proven — bare rig has no OPC-bound instance / inbound method / parked op / CLI failover verb.
## Gotchas for the executor (will bite; read twice)