Payment Write-back Monitoring with Grafana
Why This Matters
Payment write-back failures have direct financial impact on offices. Knowing how to read the Grafana monitoring dashboard — and understanding that you're looking at attempts, not payments — is essential for on-call and incident response.
What You'll Learn
- Where to find the payment write-back Grafana dashboard
- The critical distinction: the team stores and counts attempts, not payments (one payment can have 6 retried attempts)
- What the dashboard shows: request counts, response times, failure rates per integration
- How to identify a spike in write-back failures from the dashboard
- Why the tables are designed to extract attempts rather than deduplicated payments
- How to use dashboard data to triage a payment write-back production issue
Watch Video
Lesson Summary
The integration team owns a Grafana dashboard for payment write-back monitoring. The key mental model: the team stores attempts, not payments. If a payment write-back fails and retries 6 times, the database has 6 records. The dashboard shows these attempt counts — not the number of unique payments processed. This design means failure rate analysis looks at per-attempt success/failure rather than per-payment. The dashboard panels show RPC-level metrics: each write-back RPC has its own success/failure rate graph. Spikes in failure rates are the primary signal for incidents. Use the dashboard to identify which integration and which RPC is failing before investigating the code.