neutrino: add distinct Neutrino sub-server

[Roasbeef]

As is, when running a neutrino node, the only way a user can really keep track of what's going on is to check out the logs. Other than that, they aren't able to add new peers, fetch filters/headers manually, re-trigger a rescan, etc. In order to enhance the external facing capabilities of neutrino, we shoul dadd a new sub-server that's online all the time, but only "available" when the neutrino client mode is actually activated.

Steps To Completion

• [x] Create a new neutrino-specific sub-server that is always online and has a method that allows (or maybe we add it to the existing state RPC?) users to query to see if neutrino backend is active or not.

• A non-exhaustive list of relevant calls to add includes the following:

  • [x] adding/disconnecting peers
  • [x] checking the current block height
  • [x] fetching filters, blocks, and block headers
  • [ ] triggering a rescan for a specific address/script
  • [ ] query into internal rescan/filter/notification state
  • [ ] start/stop automated broadcast of certain transactions

[ErikEk]

If no one else wants it, I will gladly do it.

[ErikEk]

Update: I have made some progress already. Done with:

• adding/disconnecting/listing connected peers
• checking the current block height, if synced and if active
• fetching filters, blocks, and block headers
• if peer is banned.
• triggering a rescan for a specific address

[Roasbeef]

@ErikEk any updates/comments/concerns on the PR?

[ErikEk]

@Roasbeef I can release a work-in-progress right now. Still not happy with the handling of the re scanning mechanism I've built. Edit: Draft on the way...

[arshbot]

that's online all the time, but only "available" when the neutrino client mode is actually activated.

What is the reasoning behind keeping a subserver online but not available? This would make sense if one could activate neutrino client mode without restarting lnd

[Roasbeef]

What is the reasoning behind keeping a subserver online but not available? This would make sense if one could activate neutrino client mode without restarting lnd

If neutrino isn't active, then the sub-server wouldn't be able to do anything. We don't explicitly support switching from a full node backend to a neutrino one. It may work in most cases, but we haven't explicitly verified no internal state is incorrectly mutated during the process.

[Roasbeef]

Everything here but rescanning is implemented in: https://github.com/lightningnetwork/lnd/pull/5652