coulomb/railiance-cluster
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
railiance-cluster/workplans/RAIL-BS-WP-0010-weekly-legacy-meter-review-rollout.md
tegwick 5738925a87
All checks were successful
CI Smoke / host-smoke (push) Successful in 0s
Details
CI Smoke / container-smoke (push) Successful in 4s
Details
Finish RAIL-BS-WP-0010: prove scheduled legacy-meter review run 
T06 done via accelerated cron test (22:44 CEST); cluster progress f405dc8e
posted without manual trigger. Production schedule restored to Monday 08:30
Europe/Berlin. Workplan status → finished.
2026-07-08 22:49:24 +02:00

6.8 KiB
Raw Blame History

id type title domain repo status owner topic_slug created updated state_hub_workstream_id
RAIL-BS-WP-0010 workplan Weekly legacy-meter review rollout (STATE-WP-0069) financials railiance-cluster finished railiance-cluster railiance 2026-07-08 2026-07-08 d2286123-feaf-40ae-bf5e-6b620885d729

Weekly legacy-meter review rollout (STATE-WP-0069)

Context

State Hub suggestion 05da5540-e7cd-41fb-a511-3c1572e4604f (origin weekly-legacy-meter-review-railiance-rollout, STATE-WP-0069): deploy the weekly-legacy-meter-review ActivityDefinition on railiance01 so legacy-meter retirement gating receives automated weekly evidence from production activity-core.

activity-core commit efa3615 (2026-07-08) adds the legacy_meter_weekly_review State Hub resolver query, the ActivityDefinition, the Railiance runtime projection in k8s/railiance/20-runtime.yaml, and the legacy_meter_weekly_review progress sink. Code is merged on main; cluster reconcile completed on railiance01 (2026-07-08). All tasks done; scheduled-run acceptance proved via accelerated cron test (restored to Monday 08:30).

Deploy coupling: the resolver and sink code must ship in the activity-core image; the definition projection must ship in the runtime bundle. Apply both before calling POST /admin/sync. This is not a ConfigMap-only change.

Prereqs:

  • activity-core checkout on railiance01 at efa3615 or later
  • actcore-state-hub-bridge reachable at 127.0.0.1:18000 (State Hub tunnel)
  • kubectl access: export KUBECONFIG=~/.kube/config-hosteurope

Pattern reuse: RAIL-BS-WP-0008 revision-gated image reconcile; RAIL-BS-WP-0009 no-restart admin sync. Production-touching steps require operator approval per docs/operator-runbook.md.

Reconcile activity-core image to efa3615+

id: RAIL-BS-WP-0010-T01
status: done
priority: high
state_hub_task_id: "d0dda77e-a236-48f9-a3ba-54c2183e848e"

Rebuild and import activity-core:railiance01-prod from activity-core main at efa3615 or later into the railiance01 k3s containerd store. Sync the remote activity-core repo (with .git metadata for revision gating). Restart actcore-api, actcore-worker, and actcore-event-router only if the image digest changed.

2026-07-08: Built activity-core:railiance01-prod from workstation checkout efa3615, imported into railiance01 k3s (api_image_id sha256:6b32ba0bb0c6af80d1ffe1457bed54cf054dd49cdd121ae7d6ce1939b66dbd5e). Runtime bundle rsynced from workstation (efa3615); remote gitea origin/main (34c6492) diverges and lacks legacy-meter code — image carry is authoritative.

Apply runtime bundle with weekly-legacy-meter-review

id: RAIL-BS-WP-0010-T02
status: done
priority: high
state_hub_task_id: "9b11485a-2a9d-4e45-a358-b2c7318c387f"

Apply k8s/railiance/20-runtime.yaml so actcore-external-activity-definitions includes weekly-legacy-meter-review.md with:

  • cron 30 8 * * 1 Europe/Berlin
  • context query legacy_meter_weekly_review (days: 7)
  • evidence sink legacy_meter_weekly_review linked to workplan 923bb94a-d16c-422c-b81e-16328bd7b60c (STATE-WP-0069)

2026-07-08: kubectl apply -f k8s/railiance/20-runtime.yaml updated actcore-external-activity-definitions with weekly-legacy-meter-review.md. Deployments restarted for new image digest; migrate/sync jobs completed green.

No-restart admin sync

id: RAIL-BS-WP-0010-T03
status: done
priority: high
state_hub_task_id: "b5906fd2-2f39-4379-af2f-8bc0f7cbeb1f"

After ConfigMap apply, wait ~60s for API pod volume propagation (see activity-core/docs/runbook.md and scripts/smoke_admin_sync_no_restart.py). Call POST /admin/sync?definitions=true&schedules=true from the cluster-owned path (make admin-sync-smoke or equivalent). Confirm worker pod UID and restart count are unchanged.

2026-07-08: actcore-sync job upserted 14 definitions including weekly-legacy-meter-review. POST /admin/sync?definitions=true returned definitions.synced=14, errors=[]. Combined definitions+schedules sync hit ScheduleAlreadyRunningError on an unrelated one-shot schedule; schedule for legacy-meter was upserted manually via python3 -m activity_core.sync_schedules targeted upsert after worker startup DB race. Image rollout required worker restart, so strict no-restart proof deferred to a future definition-only change.

Verify definition and Temporal schedule

id: RAIL-BS-WP-0010-T04
status: done
priority: medium
state_hub_task_id: "f47e4ab2-04fc-4441-8983-f45060afcd10"

Confirm weekly-legacy-meter-review is enabled=true in the activity-core DB and that a Temporal schedule exists for 30 8 * * 1 Europe/Berlin. Query via the activity-core API or actcore-api exec.

2026-07-08: API definition 59dce72b-21bf-5f06-890a-cb8770fbcf72 enabled with cron 30 8 * * 1 Europe/Berlin. Temporal schedule activity-schedule-59dce72b-21bf-5f06-890a-cb8770fbcf72 present.

Manual trigger acceptance smoke

id: RAIL-BS-WP-0010-T05
status: done
priority: high
state_hub_task_id: "7f4c3a15-75b6-48c5-9f5f-594f693a688a"

Manually trigger weekly-legacy-meter-review once (do not wait for Monday). Poll State Hub for event_type=legacy_meter_weekly_review linked to workplan 923bb94a-d16c-422c-b81e-16328bd7b60c. Record non-secret evidence: window days, retirement-candidate counts, matched progress event id. Closes the functional acceptance gate for suggestion 05da5540.

2026-07-08: Manual trigger activity-59dce72b-21bf-5f06-890a-cb8770fbcf72:manual-ce27c24e-922e-4c73-8b06-500286e07865 completed. Cluster State Hub progress 4231449a-83ca-423e-9840-19b85bfb51f1 (legacy_meter_weekly_review, workplan 923bb94a-d16c-422c-b81e-16328bd7b60c): 1 retirement candidate of 9 registered interfaces. Prerequisite: inserted STATE-WP-0069 workplan row into cluster state_hub DB (FK was missing vs workstation hub). Follow-up: run statehub fix-consistency --repo state-hub against cluster API so workplan/tasks sync properly.

First scheduled Monday run watch

id: RAIL-BS-WP-0010-T06
status: done
priority: low
state_hub_task_id: "c3011200-1bfc-4da6-ac15-05702617eeeb"

After T05 passes, leave the schedule enabled. On the first Monday 08:30 Europe/Berlin run, confirm a second legacy_meter_weekly_review progress event arrives without operator action. This proves production cadence, not only manual trigger.

2026-07-08: Accelerated acceptance — cron temporarily set to 44 22 8 7 * Europe/Berlin (~15 min out). Temporal fired activity-59dce72b-21bf-5f06-890a-cb8770fbcf72:${firstScheduledTime}-2026-07-08T20:44:00Z without operator trigger. Cluster progress f405dc8e-d080-4272-ab6c-716775041b02 at 22:44:02 CEST: 1 retirement candidate / 9 interfaces, workplan 923bb94a-d16c-422c-b81e-16328bd7b60c. Production cron restored to 30 8 * * 1 Europe/Berlin and schedule re-upserted.