fix(logging): add startup and shutdown lifecycle events

[kkovaacs]

Summary

• Log structured startup configuration and readiness events for the node, validator, NTX builder, and remote prover.
• Report bootstrap start and completion, including genesis sources, commitments, configuration paths, and output directories.
• Track full-node synchronization readiness and log state transitions without duplicate events.
• Sanitize service endpoints before including them in startup logs.
• Log SIGINT/SIGTERM shutdown requests and successful completion, while returning a non-zero exit code when graceful shutdown times out.

Partially resolves part 5 of issue #2330

Changelog

[[entry]]
scope       = "general"
impact      = "added"
description = "User visible log events have been added for startup and shutdown to all services."

[sergerad]

nit: "undefined" or "custom" instead of "unknown"? Not important

[Mirko-von-Leipzig]

I'm unsure how I feel about this tbh. It achieves the goal, but I think we're wading into the territory of the telemetry and logging overshadowing the actual code that I want to see.

Or put differently, I want the majority of the LoC in a function to be actual logic. Not logging or helper functions for it.

Some of this is dictated by instrument! scoping for field recordings, but I feel this is going to be a problem no matter what given tracing API.

We can go ahead; I don't have a better solution.

[Mirko-von-Leipzig]

I wonder if it would be beneficial to create our own macro; given that we don't want the full tracing feature set.

If we re-arrange the macro order, we can put the fields at the end and remove the need for { } encapsulation?