[WIP] feat: use quic multipath #3381
Draft
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Documentation for this PR has been generated and is available at: https://n0-computer.github.io/iroh/pr/3381/docs/iroh/ Last updated: 2025-08-01T09:29:38Z |
dignifiedquire
changed the title
dignifiedquire
force-pushed
the
feat-multipath
branch
3 times, most recently
from
72cb071 to
db712c0
Compare
dignifiedquire
force-pushed
the
feat-multipath
branch
2 times, most recently
from
4827e62 to
946f71c
Compare
can't use task::yield_now in a sync function
There's a silly bug in how discovery is hooked up that if you don't do this check you fail hard. Rather than clean up that discovery logic I would prefer to just get on with the NodeStateActor refactor.
While that's currently possible, once we do QUIC-NAT-TRAVERSAL we can only do this once there's a connection anyway. So let's just delay this by another RTT.
- When a path is selected make sure it is open on all connections. - Do not open a path on a connection when it is already opened.
This no longer tries to figure out if discovery is needed based on how recently we may have communicated with a remote endpoint. Maybe that's too simple and we'll have to add back some extra check for discovery. But most active connections would have had a response by that time and not trigger the discovery.
The need for these was removed after the latest upstream quinn changes.
MagicStack is just an Endpoint. mesh_stacks is a StaticProvider with a task watching the node addresses. This simplifies things a lot. And means I can delete some more unused code that this stuff was using.
I'm not sure this test does anything useful anymore. Maybe there's a new test that needs to exist now but I don't yet know.
Our AsyncUdpSocket is always an IPv6 socket, because we need to sent to our IPv6 ULAs for the mapped addresses. This means Quinn will convert any IPv4 destination into an IPv4-mapped IPv6 address (::ffff:a.b.c.d). But our sockets are bound to a specific family, so even if the OS supports and IPv6 socket that can handle IPv4 it will not accept those addresses. So when we are using the destinations we need to convert back to the canonical addresses in the IPSender.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Work on integrating n0-computer/quinn#28 into the iroh magic