soroban-rpc: Runbook for infra operators

[paulbellamy]

What problem does your feature solve?

As an infra operator I'd like to know how to deploy/run/maintain/troubleshoot soroban-rpc.

What would you like to see?

We should write a runbook for soroban-rpc. It should include instructions how to:

• download
• build (optionally)
• deploy
• configure
  • based on soroban-rpc: Add support for configuration from a file #507
• operate
• monitor
  • Based on soroban-rpc: Add /metrics endpoint #496
  • and on soroban-rpc: Audit and standardize logging #497
• scale
  • based on soroban-rpc: Define and document a scaling strategy #506
  • and should include HW references to help provision sufficient resources based on soroban-rpc: Establish a benchmark baseline on (a few?) different hardware profiles #505
• troubleshoot their instance
  • Figure out some potential common failure modes and what to do for them
  • Worst case, tell them how to gather the data we would need to help debug and how to contact us

This content should ultimately end up in soroban-docs, but it's acceptable for now to have it be a google doc that we pass around to partners. This approach may also be easier to do initial review by the team.

The docs currently have a large section describing setup using quickstart, we do not want that to be recommended for production instances. We should update the docs and this runbook to make that very explicit.

What alternatives are there?

[sreuland]

@stellarsaur , is this intended to base or append to content from @tyvdh's RPC runbook

[stellarsaur]

This can be considered complete since we now have suitable guides for infrastructure operators in this doc and on our docs website