Skip to content

Today's Action Items (2026-06-11)

Use this page as your operator dashboard for WC26 launch day. Every row links to the full detail; the order on this page IS the execution order. Do not skip ahead — backups first, then config changes, then SKU upgrades, then code deploy.

0. Where we are right now

  • WC26 opener: 2026-06-11 19:00 UTC (Mexico vs South Africa).
  • Prod is on the 2026-05-22 image (fea5dc41…). None of today's session fixes are on prod.
  • Staging has shipped today: timeline-include, micro-bump no-decay, per-position cooldown, OTel sampler.
  • Prod LAW daily cap = 30 GB/day already applied by operator at 06:03 UTC.

Read prod-snapshot-2026-06-11 first to see exactly what prod looks like vs what's expected.

1. Pre-flight (before any change)

# Task Owner Reference
1.1 Authenticate Azure CLI with device-code Operator azure
1.2 Verify FTL identity (az account show) Operator azure
1.3 Confirm no live match in next 90 min Operator Sportmonks schedule
1.4 Confirm release/staging CI is green Backend gh run list --branch release/staging
1.5 Audit GitHub Environments Operator github-env-audit-2026-06-11

2. Prod readiness fixes (~30–60 min)

Do in this exact order. Each row links to its full runbook section with CLI + Portal UI + risk + mitigation + rollback.

# Task Why first Reference
2.1 Backup Postgres (on-demand) Enables rollback for every step after prod-launch-runbook §3.1
2.2 Enable PG autoGrow Prevents disk-full halt during tournament prod-launch-runbook §3.2
2.3 Set Container Apps env vars (SPORTMONKS_LEAGUE_IDS=732, OTEL_TRACE_SAMPLE_RATIO=0.01, APPLICATIONINSIGHTS_CONNECTION_STRING) Picked up on next image roll; no live impact prod-launch-runbook §3.3
2.4 Upgrade PG SKU B2ms → D2ds_v5 Removes CPU credit cliff prod-launch-runbook §3.4
2.5 Upgrade Redis Basic C0 → Standard C1 SLA + memory headroom prod-launch-runbook §3.5
2.6 Enable required reviewers on production GitHub Env Stops un-gated prod deploys prod-launch-runbook §3.6
2.7 Fix VITE_APPLICATIONINSIGHTS_CONNECTION_STRING_PROD secret on ftl-frontend Or frontend ships with empty telemetry conn string github-env-audit §5.1

3. Deploy release/staging → prod (~15–25 min)

# Task Reference
3.1 Cut prod-integration-2026-06-11 off origin/release/prod prod-deployment-runbook §3
3.2 Cut staging-to-prod-release-2026-06-11 off origin/release/staging prod-deployment-runbook §3
3.3 PR 1: staging-to-prod-release → prod-integration prod-deployment-runbook §4
3.4 PR 2: prod-integration → release/prod prod-deployment-runbook §5
3.5 Approve production env gate when CI pauses prod-deployment-runbook §5
3.6 Watch migrate job — confirm dirty=false prod-deployment-runbook §7
3.7 Smoke endpoints with prod custom domains: https://prd-api.ftljeta.cloud/health + wss://prd-ws.ftljeta.cloud/health prod-deployment-runbook §8.2

4. Post-deploy validation (~15 min)

# Task Reference
4.1 Image SHAs on all 3 apps match release/prod HEAD prod-deployment-runbook §8.1
4.2 Replica state Running, restartCount=0 prod-deployment-runbook §8.3
4.3 Zero errors in container stdout last 10 min prod-deployment-runbook §8.4
4.4 Env vars present (SPORTMONKS_LEAGUE_IDS=732 + sampler + AppInsights) prod-deployment-runbook §8.5
4.5 End-to-end smoke (sign in → open → close) prod-deployment-runbook §8.6
4.6 Cost check (no spike in first hour) prod-deployment-runbook §8.7

5. Standby — during the first match

While Mexico vs South Africa kicks off (and the next 24 h of fixtures), keep the incident-debugging page open in a tab. The most likely failure modes are pre-mapped in §6 of that doc.

6. Things explicitly skipped (per operator decision)

Item Why skipped today
OTel Collector sidecar (ADR-0031) Operator chose to launch with no distributed traces. Sampler in code is dead-armed; activates when sidecar lands later.
KEDA scale-to-zero on non-game hours Saves ~₹1 500/tournament; cold-start UX hit not worth it.
1-year / 3-year reserved PG 39-day commitment too short to break even.
Adding ACS / Email Service to prod RG Welcome / milestone emails fail silently until added — accepted post-launch.

7. Rollback shortcuts (if needed)

Scenario Fast rollback
App crashloops after deploy az containerapp revision activate -g ftl-prd-rg-cin -n <app> --revision <prev> (Section 9 of prod-deployment-runbook)
PG SKU resize stuck >30 min File Azure support, do NOT cancel mid-resize
Redis failover broken Revert to Basic C0 with same az redis update command flipped
Migration dirty=true Manual flip after operator review (prod-deployment-runbook §7)
Bad data introduced PITR restore to pre-deploy backup (prod-deployment-runbook §9.3)

8. Read these in this order if you're new

  1. azure — CLI scoping basics. Five minutes.
  2. This page (today-2026-06-11) — what to do today.
  3. prod-snapshot-2026-06-11 — measured prod state.
  4. prod-launch-runbook — one-time prod readiness fixes.
  5. prod-deployment-runbook — staging → prod promotion.
  6. incident-debugging — when something breaks.
  7. usage-analytics — measure how it's going.
  8. cost-estimate-audit-2026-06-11 — full cost model + verification.
  9. github-env-audit-2026-06-11 — GH Envs + secrets gaps.
  10. deployment — overall deploy architecture reference.