Monitoring And Metrics

How to track and debug bundler performance in ERC-4337 systems.

Running a reliable bundler requires robust observability. From mempool activity to simulation results and gas refunds, operators need insight into every layer of the ERC-4337 stack.

📊 What to Monitor

UserOp Lifecycle

  • Ingress: Count of UserOps received per second

  • Validation results: pass/fail reasons

  • Bundle size and frequency

  • Inclusion latency (time-to-handleOps)

Simulation

  • Calls to simulateValidation

  • Revert rates by failure class (signature, paymaster, initCode)

  • Gas estimation accuracy

EntryPoint Interaction

  • Successful vs failed handleOps() calls

  • Average gas used per UserOp

  • Revenue tracking: priority fee, refund breakdown

🧰 Tools and Dashboards

Custom Logs and Metrics

  • Export Prometheus or OpenTelemetry metrics

  • Log UserOp hashes, error types, and bundling cycles

Ready-made Dashboards

  • Pimlico: Offers a bundler dashboard with real-time metrics and mempool stats

  • Silius: Exposes simulation logs and operation state via API

🔄 Integration with Alerting

  • Alert on sudden spikes in rejected UserOps

  • Monitor bundler uptime (especially if self-hosted)

  • Track deviation between simulated and actual gas usage

✅ Summary

A performant and trustworthy bundler is a transparent one. Whether you're running infrastructure at scale or testing locally, observability is essential to understanding mempool dynamics, simulation accuracy, and end-to-end UserOp throughput.

PreviousSimulation RequirementsNextPaymasters

Last updated