Skip to main content

Production Monitoring & Health

Once your project is live in production, regular operational monitoring ensures your payment rails remain healthy, webhooks are acknowledged promptly, and the Sweeper is actively recovering orphaned payments.

[1. Track Consensus Settlement Rate] ➔ [2. Monitor Outbound Webhooks] ➔ [3. Audit Sweeper Recoveries] ➔ [4. Review Daily Reconciliation]

The 4 Core Production Health Metrics​

Monitor these four metrics in the SolydFlow Console to evaluate revenue health:

Health MetricHealthy TargetWhat an Anomaly IndicatesWhere to Check
1. Consensus Settlement Rate> 92%Sudden drop indicates bank 3DS downtime or an expired gateway secret key.Console → Dashboard Overview
2. Webhook Delivery Rate> 99.5%Spike in non-200 responses means your backend receiver is slow or timing out.Console → Outbound Webhooks
3. Sweeper Recovery VolumeConsistent baselineSpike indicates upstream telco lag (e.g., Safaricom M-Pesa or NIBSS delays).Console → Recover Logs
4. Inbound Gateway StatusGreen (200 OK)401 errors mean gateway secret keys in the Vault were rotated out of sync.Console → Inbound Gateways

Setting Up Operational Alerts​

Do not wait for customer support tickets to discover a broken webhook receiver. Configure automated alerts in Console → Project Settings → Notifications:

┌─────────────────────────────────────────────────────────────┐
│ Automated Health Alerts                                     │
├─────────────────────────────────────────────────────────────┤
│ Notify via:          (●) Slack Webhook    ( ) Email         │
│ Webhook URL:         [ https://hooks.slack.com/services/... ]│
│                                                             │
│ Trigger Alerts For:                                         │
│ [x] Webhook Delivery Failure (5 consecutive non-200s)       │
│ [x] Gateway Outage / Failover Triggered                     │
│ [x] High Volume of Orphaned Transactions (>10 in 5 mins)    │
│ [x] API Key Compromise / Emergency Revoke                   │
└─────────────────────────────────────────────────────────────┘

When an incident occurs (such as your backend server taking longer than 5 seconds to process events), your engineering team receives an instant alert with direct links to the failing transaction references.

Routine Maintenance Schedule​

Keep your revenue infrastructure operating smoothly with this maintenance cadence:

Weekly​

  • Inspect Outbound Webhook Retries: Confirm all retried events successfully resolved with HTTP 200 OK.
  • Review Orphaned Recoveries: Verify the Sweeper is successfully reconciling delayed USSD and bank transfer checkouts.

Monthly​

  • Run Financial Reconciliation: Use the Reconciliation tab to cross-check SolydFlow ledger revenue totals against raw bank payout statements before closing accounting books.
  • Audit Vault Status: Verify all connected providers (Paystack, Stripe, Flutterwave, M-Pesa, Monnify) report healthy connection pings.

Quarterly​

  • Review Team Console Access: Remove former team members and verify that all active administrators have Two-Factor Authentication (2FA) enabled.
  • Rotate Webhook Signing Secrets: If required by internal security policy, regenerate your whsec_... secret using our zero-downtime rolling grace period.

Incident Triage: When Things Go Wrong​

If you observe an unexpected drop in completed checkouts:

  1. Check Inbound Gateway Webhooks: Are incoming provider events returning green 200 statuses? If you see 401, your gateway secret key in the Vault may have expired.
  2. Check Outbound Webhooks Log: Is your server returning 504 Gateway Timeout? Your server may be doing heavy synchronous work inside the handler instead of returning 200 immediately.
  3. Inspect the Transactions Ledger: Are checkouts getting stuck in UNRESOLVED_ORPHAN? Click "Verify Now" on a transaction to force an immediate server-to-server check.