Implement open-telemetry monitoring

[scarmuega]

Problem Statement

On long-running daemon sessions, it's important to understand the health and progress of the pipeline. The only way to do that currently is by inspecting the stderr log outputs from the process. This is a very inefficient manual task.

Feature Scope

We will add open-telemetry instrumentation, providing the following metrics:

• stage throughput (events / sec)
• channel queue size (total events)
• process resource usage (mem, cpu)
• source last checkpoint (block num)
• sink last checkpoint (block num)
• node tip delta (node tip slot - sink checkpoint slot)

A Prometheus exporter will be available as a feature flag.

[scarmuega]

closing in favor of #154