Add log aggregation

[josibake]

For a running warnet, ship debug logs back to a central server. This should include:

• Debug logs
• bonus: USDT tracepoints per node, all aggregated back to the logging server
  • this gives us data on p2p messages

[willcl-ark]

Going to try promtail

[willcl-ark]

bonus: USDT tracepoints per node, all aggregated back to the logging server

0xB10C has been working on an RPC-based (as opposed to tracing-based) addrman-observer over the last few days. Once bitcoin/bitcoin/pull/28523 is merged, you'd be able visualize the addrman of each warnet tank.

[0xB10C]

0xB10C has been working on an RPC-based (as opposed to tracing-based) addrman-observer over the last few days. Once bitcoin/bitcoin/pull/28523 is merged, you'd be able visualize the addrman of each warnet tank.

You probably need to run a proxy like 0xB10C/addrman-observer#4 as I don't think accessing the getrawaddrman RPC from the browser is a good idea (or even possible). This needs another container but will allow accessing n warnet nodes. It will be similar in terms of configuration like fork-observer.

I'll be developing such a proxy and can also take care of packaging it for docker. I'll need the proxy anyways for my monitoring.

[pinheadmz]

closed by #133