Production Readiness

Current production-readiness status across eSHARS services. A gap here is a risk — not a blocker, but something that needs a plan.

Last reviewed: 2026-04-06

Service names below are the legacy estate

The inventory below (wa-eshars-prod, vm-sql16-np, vm-ssrs16-prod, …) predates the 2026-06 tier-based rebuild and has not been re-assessed against the rebuilt estate (app-eshars-{env}-{region}, vm-eshars-*-{tier}-cus). Treat it as the legacy-estate readiness picture.

Services

Service	Owner	Runbook	Monitoring	Rollback Plan	OS/Runtime EOL	Notes
wa-eshars-prod (HISD Web)	Leon	✅ Release Process	✅ App Insights	✅ Slot swap	✅
wa-eshars-nonprod (all slots)	Leon	✅ Release Process	✅ App Insights	✅ Slot swap	✅
eshars-mobile-api	Leon	✅ Release Process	✅ App Insights	✅ Pipeline rollback	✅
eshars-mobile-ui (Flutter)	Leon	❌ No runbook	❌ No monitoring	❌ No formal rollback	✅	Mobile app — store review cycle complicates rollback
eshars-mgmt (portal)	Leon	❌ No runbook	⚠️ Unknown	❌ No formal rollback	✅
vm-ssrs16-prod (SSRS/SQL)	Leon	⚠️ VM Troubleshooting partial	❌ No health alerts	⚠️ Manual failover only	❌ WS2012R2 + SQL2016 EOL	Critical risk — see vm-ssrs16-prod
vm-sql16-np (SSRS nonprod)	Leon	⚠️ Partial	❌ No health alerts	⚠️ Manual only	❌ WS2012R2 + SQL2016 EOL	Same risk as prod SSRS VM
vm-jobscheduler-prod	Leon	⚠️ OOM restart only	✅ OOM alert + auto-restart	❌ No general runbook	✅
aa-eshars-prod (Azure Automation)	Leon	✅ Azure Automation Runbooks	⚠️ No alerts on failed jobs	N/A	✅	~12 runbooks on deprecated RunAs auth
db-eshars-prod (Azure SQL)	Leon	✅ Database Failover	✅ Azure SQL metrics	✅ Backup & Restore	✅
App Gateway (nonprod/training)	Leon	❌ No runbook	⚠️ Unknown	❌ No documented rollback	✅	Fronts nonprodreports.myeshars.com, trainingreports.myeshars.com
kv-eshars-* (Key Vaults)	Leon	✅ EDI Credential Rotation partial	❌ No expiry alerts	N/A	✅	No secret expiry monitoring — secrets could expire silently

Legend

Symbol	Meaning
✅	In place and adequate
⚠️	Partial / needs improvement
❌	Missing

Top Gaps

These are the highest-risk gaps — prioritized by impact:

1. vm-ssrs16-prod OS/SQL EOL — Windows Server 2012 R2 and SQL Server 2016 are end-of-life. No security patches. HIPAA/compliance risk. Full VM rebuild required. See vm-ssrs16-prod.md for the remediation plan.

2. Key Vault secret expiry monitoring — No alerts configured when secrets approach expiry. A silently expired secret would cause auth failures in production (TMHP, SFTP, App Service config). Add Azure Monitor alerts on Key Vault secret expiry.

3. Azure Automation deprecated auth — ~12 runbooks still use AzureRunAsConnection (Service Principal / RunAs). Microsoft deprecated RunAs accounts in Sept 2023. Migrate to Managed Identity before the accounts stop working.

4. Mobile app rollback — No documented procedure for rolling back a bad iOS/Android release. App store review cycles mean a hotfix takes hours to days. Document the emergency process (TestFlight rollback, phased rollout revert).

5. eshars-mgmt monitoring — Unknown whether Application Insights is configured. Verify and add if missing.

6. App Gateway runbook — No documented procedure for App Gateway configuration changes or failures. Low frequency but high impact if it goes down.