Telemetry

[dzdidi]

It is desired to have certain amount of insight into how service operates historically and at any moment of time for both business and operational intelligence.

Prometheus metrics for different stats like:

• balances in different states and layers
• forward payments count
• forward fees earned
• number of channels in different states (public / private)
• inbound/outbound capacity
• etc

[arturgontijo]

Is anybody working on this one?

[G8XSU]

Adding offline communication:
Ideally, we would want metrics from the end-to-end LDK stack, but that will probably take a bit longer; hence, it would be best if we could get started with just ldk-server for now.

My thinking so far is to integrate with the metrics crate and metrics-exporter-prometheus.

metrics, just like the log facade, will provide us with additional flexibility, and if ldk-node and ldk do decide to integrate metrics, they can probably use the same.

prometheus-metrics, since it is standard and supported by most metrics libraries, viewers, and cloud providers.

For cloud provider support, cloud-watch-agent, open-telemetry-collector, and azure-agent, etc., can easily scrape the /metrics endpoint and publish them with in-built support.

[G8XSU]

Just checking in, @arturgontijo will you be working on this?

[arturgontijo]

Just checking in, @arturgontijo will you be working on this?

Hey, sorry about that...but yeah I can do that, will start it asap and keep you posted!

[u32luke]

LND has Lndmon. It's a docker container with Grafana & pre configured Prometheus connectors.

https://github.com/lightninglabs/lndmon