kafka-lag-lens
A tiny CLI that turns consumer-lag metrics into a capacity plan — how many pods and threads you need for a target traffic peak.
Sample project — replace with your own.
The problem
Preparing consumers for a traffic spike usually means a spreadsheet, a Grafana tab, and guesswork.
What it does
- Pulls per-pod consumption rate and peak production rate
- Computes required threads and pods with configurable headroom
- Flags topics where partitions are the bottleneck
Status & roadmap
- Prometheus source
- Markdown / CSV output
- Recommendations for
max.poll.records - GitHub Action
Contributions welcome — open an issue on GitHub.