comit-network
ASB observability
Hi!
I'm currently working on an automated setup for ASBs, and while it's great that
it has some neat trace-based logging, to derive dashboards and set up alerts,
one needs to go around parsing those logs (which yeah, are structured when
using --json, very nice), which I believe is not very optimal.
Have you folks thought about other ways of providing observability around the state that ASB is dealing with?
Personally, I'm a big fan of prometheus for exposing simple counters that one can then set alerts on (+ how lightweight it is to integrate), but given that the project already has trace-based logging, one could even go one step further and extend that to use a standardized sdk like opentelemetry which would then allow one to hook up ASB's traces with whatever platform they use / self-host.
The goal here would be able to:
- from metrics, derive an "overview" dashboard + alert rules
- from tracing, be able to dig into how a form of error / something unexpected occurred (given that lack of large number of traces, one can even go without sampling)
Wdyt?
Thank you! (and again, kudos on the traction the project is getting!)
addendum: see https://github.com/open-telemetry/opentelemetry-rust
We just merged a PR #780 that stores the swaps state and other info in a sqlite database. This should make it easy for developers to build apps on top of xmr-btc-swap. Would that work for you?
