FLOW_ONLY note now names step J. Live gate on the docker rig (8 nodes rebuilt
from this branch via docker/deploy.sh, FlowLogDemo seeded by
docker/seed-flow-log-demo.sh — idempotent, add then update proven):
J step logs live+replay -> YES
live: YES root=5793ddc2-5273-4e62-86c2-ccfc8de0e080 logs=3 levels=debug,info,warn scriptRun=Failed error=CALL_FAILED
replay: YES root=5793ddc2-5273-4e62-86c2-ccfc8de0e080 logs=3 levels=debug,info,warn scriptRun=Failed
H, H2 and I stayed YES in the same run.
Browser check (Playwright, /deployment/flow-debugger?executionId=<J root>,
screenshot /tmp/flow-log-demo.png): the SCRIPT RUN node is red with subtitle
"InvalidOperationException: demo failure" and badge "! 1 ≡ 2"; the WRITE node
carries no badge; the strip reads "LOG 0 errors · 1 warnings · 2 info"; the
detail panel opens on "Log (3)" with the full System.InvalidOperationException
block and the three info/debug/warn script lines.
Rig run 2026-09-06 (image built from this branch, containers recreated from
main's compose file): full drill A-H YES (active-central kill takeover 27 s,
rejoin as standby in 3 s), then with FLOW_ONLY=1:
H live+replay -> YES
live: YES root=8e628d0a-428b-48df-b937-f4c2ac631fc3 events=8 api=Delivered db=Delivered notify=Submitted
replay: YES root=8e628d0a-428b-48df-b937-f4c2ac631fc3 events=10 api=Delivered db=Delivered notify=Delivered
I outbound live+replay -> YES
Two rig facts the drill established: (1) with the DB reachable the site's
immediate-completion path resolves a CachedWrite INSIDE the run, so the DbCall
node is already Delivered inline (Submit/Attempted/Resolve share a tick — the
check now ranks statuses instead of taking list order); (2) cached-call
telemetry reaches central on its own ~20-30 s cadence, later than the 5 s audit
push, so replay needs the bounded retry. New FLOW_ONLY=1 switch runs H + I
without the notifier build or the docker kill.
Claude-Session: https://claude.ai/code/session_01NcMG6xXdsJbmD2wYAgkK29
Both central nodes gain a root AuditLog:PerChannelRetentionDays:ScriptTrace=30 (validator
floor). verify-delmia-notifier.sh step H: finds DelmiaRecipeDownload via api-method list,
runs flow-debug run-method with the drill parameters and asserts completed=true plus the
kind set {InboundRequest, RouteCall, ScriptRun, AttributeWrite, AttributeWait} from each
event's detailsJson, then flow-debug replay on the root ExecutionId with the same check,
retried up to 40 s for the 5 s site telemetry push; prints 'H live+replay -> YES'.
VERIFICATION PENDING: the image scadabridge:latest was built from this branch
(docker/build.sh, exit 0) but the containers have not been recreated and step H has not
run yet — the next step is a recreate through the MAIN checkout's compose file (deploy.sh
refuses a worktree, register row 68: bind-mounted appsettings would rehome), a fresh
DelmiaRecipeDownload-scoped key, and the drill.
Claude-Session: https://claude.ai/code/session_01GSLkkHKz9JrAtmJjmQK3cP
First run (docker stop) found two things: SIGTERM is a graceful CoordinatedShutdown handover (takeover 0 s), so the drill now uses docker kill; and inbound API keys are a PER-NODE SQLite store (/app/data/inbound-api-keys.sqlite on each central's own volume, written by whichever node serves the management call), so the key minted through the active node did not exist on the survivor and every post-takeover call answered 401 KeyNotFound. The sync script is a rig-only workaround; the verify script's preflight refuses to start unless both nodes list the key. Second run: A/B/E/F YES, C/G NO, D 503 in the takeover window, takeover 27 s, victim rejoined as standby.
Claude-Session: https://claude.ai/code/session_013L3XdmVii9i5sY4FChU8pB
Joseph Doherty