# RPC Reliability Readiness Checklist for Web3 Launches

Score each item 0 = missing, 1 = partial, 2 = production-ready.

## 1. Launch traffic assumptions
- Expected peak requests/sec by method and chain are estimated.
- Burst window is defined: mint, TGE, campaign, game season, exchange listing, wallet push.
- Read/write mix is known.

## 2. Endpoint resilience
- Production traffic is not dependent on a single public/free endpoint.
- Backup provider or dedicated endpoint is tested before launch.
- Failover owner and switch procedure are defined.

## 3. Data requirements
- Archive, logs, trace/debug, WebSocket and testnet needs are listed.
- Unsupported methods are known before launch.

## 4. Latency and geography
- p95/p99 target is defined by user region.
- Timeout and retry policy is explicit.

## 5. Rate limits and throttling
- Provider limit behavior is known under burst.
- Error codes are monitored separately: 429, 5xx, timeout, stale block.

## 6. Monitoring and incident response
- RPC health dashboard has owner.
- Status page, support path and escalation channel are saved.
- Rollback / traffic-shift plan exists.

## Go/no-go
0–7 high risk. 8–12 needs mitigation. 13–18 launch-ready with review.

Next step: test the riskiest chain with a GetBlock trial endpoint and compare errors/latency before launch.