How the Polygon metrics are collected

FRB runs hourly probes against each listed endpoint, measuring open time, first message latency, inclusion success, and rejection reasons. Whenever a provider throttles, the harness tags the event and repeats the test from a second POP to confirm. Only stable readings roll into this page.

• Latency numbers include p50/p95 and note which POP generated them.
• Inclusion percentages map directly to the Polygon dashboard widget so you can cross-check in-app.
• Gas/tx assumes private bundle preference with refunds applied; public PGA attempts are excluded.

Need raw proof? Download the metrics JSON and filter rows where `chain="polygon"`.

How to apply Polygon metrics

• Monitor inclusion each session; if it slides below 80%, rotate endpoints and re-run the WSS test.
• Use latency targets to set alert thresholds in Ops Pulse (>150 ms p95 triggers failover).
• Quote gas/tx numbers when comparing FRB vs template bots in internal memos.

Endpoint leaderboard

Ops notes boilerplate

Add this boilerplate to your daily ops logs: list which endpoint you promoted, current inclusion delta vs baseline, and any rotation decisions. Sharing consistent summaries keeps compliance and engineering aligned when Polygon routes change.

• “Endpoint chosen” + reason (latency, success, geographic coverage).
• “Guardrails enforced” (refund guard, session cap, latency threshold).
• “Next review” timestamp so someone is accountable for checking again.

FAQ