feat: p2p: allow overriding bootstrap nodes with environmemnt variable

[hanabi1224]

Related Issues

ChainSafe/forest#4576
ChainSafe/forest#4580
ChainSafe/forest#4581
#12289

Proposed Changes

Allow overriding bootstrap nodes at runtime with network specific environment variables

Additional Info

Checklist

Before you mark the PR ready for review, please make sure that:

• Commits have a clear commit message.
• PR title is in the form of of <PR type>: <area>: <change being made>
  • example: fix: mempool: Introduce a cache for valid signatures
  • PR type: fix, feat, build, chore, ci, docs, perf, refactor, revert, style, test
  • area, e.g. api, chain, state, mempool, multisig, networking, paych, proving, sealing, wallet, deps
• Update CHANGELOG.md or signal that this change does not need it.
  • If the PR affects users (e.g., new feature, bug fix, system requirements change), update the CHANGELOG.md and add details to the UNRELEASED section.
  • If the change does not require a CHANGELOG.md entry, do one of the following:
    • Add [skip changelog] to the PR title
    • Add the label skip/changelog to the PR
• New features have usage guidelines and / or documentation updates in
  • Lotus Documentation
  • Discussion Tutorials
• Tests exist for new functionality or change in behavior
• CI is green

[rvagg]

Is there a good reason to make new env vars for the different network types? Wouldn't a single LOTUS_P2P_BOOTSTRAPPERS suffice for all of them?

[hanabi1224]

Is there a good reason to make new env vars for the different network types? Wouldn't a single LOTUS_P2P_BOOTSTRAPPERS suffice for all of them?

Hi @rvagg I was assuming a developer could run different networks on the same machine(not necessarily at the same time), maybe devnet with LOTUS_P2P_BOOTSTRAPPERS_DEVNET then calibnet, in this case if it's a shared variable, they will always need to change the value whenever they switch the network. Does that make sense?

[rvagg]

Not really, this seems like the kind of env var you'd want to localise to the instance itself, not set in the global environment, so it gets targeted to whatever instance you're running. We also don't have a precedent of doing this, I don't believe there's ever been a need (git grep .Getenv.*LOTUS | awk -F'"' '{print $2}' | sort | uniq). It might make sense if you could have two network types in the same lotus instance, but you can't, you have one path through the config generation, in fact, it's specific to a compiled binary.

Unless you have a specific workflow in mind how you'd consume this that would make it particularly awkward to switch for the different network types?

[hanabi1224]

@rvagg changed to using a single LOTUS_P2P_BOOTSTRAPPERS