Zaino Observability Metrics discussion

[pacu]

Usage mode

zainod (running the daemon)

Problem or use case

• What do we need to track?
• What metrics do we want?
• Zaino doesn’t have any sort of metrics framework so it will be helpful to align those with whatever Zebra folks have in mind.
• Str4d put together these wallet syncing metrics https://bencher.dev/perf/zcash-speed/plots which are run on CI. can we do something similar to this on Zaino?
• What's the goal of these metrics? is it catching regressions? optimizations? just health metrics?

Proposed solution

No proposal. Just discuss this!

Alternatives considered

No response

Area

Other

Additional context

No response

[nachog00]

At first thought, i think there are 2 clear metrics categories: static/benchmark and live runtime metrics.

For daemon systems like zebra and zaino, that is, ever running, reporting on the go is key to keep awareness of the alive beast. They would serve extra metrics endpoints for this.

Static/benchmark could be ran periodically in CI, but are less of an operational feature i think.