What
The indexer currently relies on HTTP polling (eth_getLogs and eth_getBlockByNumber) which quickly exhausts rate limits/credits on hosted RPC providers like Infura/Alchemy for testnet deployments.
How to Reproduce
- Run the indexer against a hosted HTTP RPC endpoint (Infura/Alchemy) on testnet.
- Observe frequent polling requests from the indexer.
- Monitor provider credit usage and rate-limit behavior.
Expectation
Research and implement a WebSocket-based RPC ingestion path (where supported by Ponder and provider infrastructure), or document a robust alternative that materially reduces polling load and credit burn.
Additional Context (optional)
Increasing polling intervals is only a partial mitigation and can hurt freshness. We need a production-friendly default for non-Anvil environments.
Estimation
Dependencies
What
The indexer currently relies on HTTP polling (
eth_getLogsandeth_getBlockByNumber) which quickly exhausts rate limits/credits on hosted RPC providers like Infura/Alchemy for testnet deployments.How to Reproduce
Expectation
Research and implement a WebSocket-based RPC ingestion path (where supported by Ponder and provider infrastructure), or document a robust alternative that materially reduces polling load and credit burn.
Additional Context (optional)
Increasing polling intervals is only a partial mitigation and can hurt freshness. We need a production-friendly default for non-Anvil environments.
Estimation
Dependencies