oxidecomputer / oxidecomputer/dendrite

NAT encapsulated flows have one 5-tuple per sled

Open
#324 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Rust
Stars
20
Forks
3
Avg merge
3d 11h
Merged PRs (30d)
4

Description

https://github.com/oxidecomputer/dendrite/blob/f9fba02fd5351f468b15bda21d83d6dedae90818/dpd/p4/sidecar.p4#L515-L516
https://github.com/oxidecomputer/dendrite/blob/f9fba02fd5351f468b15bda21d83d6dedae90818/dpd/p4/sidecar.p4#L527
https://github.com/oxidecomputer/dendrite/blob/f9fba02fd5351f468b15bda21d83d6dedae90818/dpd/p4/sidecar.p4#L529-L530

I think as a consequence every packet from the external network to a guest will have the same flowhash when seen by the NIC, and thus land on the same ring and softring. Packets will at least be hashed among viona rings properly, since we're using the inner 5-tuple at that point.

We probably need to hash (some of?) the inner 5-tuple and place it in the encap's UDP source port field.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the linked sections of dpd/p4/sidecar.p4, especially lines 515-516, 527, and 529-530, and trace how encapsulated packets are hashed before reaching the NIC. Determine how the inner 5-tuple can influence the encapsulation UDP source port, then verify that external-to-guest packets distribute across rings while viona ring hashing remains intact.

Written by the indexing model from the issue text.

Assessment

Domain
networking
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.