Repository navigation
GEP-1619: Backend as attachment point for Session Persistence - #4876
kubernetes-prow[bot] merged 4 commits into
Conversation
|
Skipping CI for Draft Pull Request. |
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: gcs278 The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
Backend is already experimental :) |
|
Btw, the GEP link for the backend resource in the PR description seems to point to a missing GEP https://gateway-api.sigs.k8s.io/geps/gep-4488/, if that could be updated to the correct GEP and/or other information on the Backend resource that would be helpful, thanks! |
|
Oh yeah the GEP got renamed; it's https://gateway-api.sigs.k8s.io/geps/gep-4894/ |
Done, thanks for pointing that out. I'll update the PR Description to be more complete when I update this PR. |
15a928b to
a9af888
Compare
21e0518 to
6bc1c97
Compare
Move session persistence from BackendTrafficPolicy and route-inline to the Backend resource (GEP-4488). Remove BTP and route-inline API sections, update API Attachment Points to only Backend, and deprecate route-inline.
74b55a8 to
d05444f
Compare
Prior ART shows that most, if not all, implementations and data planes apply session persistence after routing (backend selection). Update traffic splitting guidance so that session persistence does not override backend selection. Also remove redundant mixed-backends edge case.
d05444f to
a55ee65
Compare
|
@mikemorris @youngnick @rikatz ready for another review. Sorry for the delays. One commit remove BackendTrafficPolicy and Route-inline for backend on the GEP + adding a deprecation notice for them on the API objects, and another commit follows up on the traffic splitting suggestion that @mikemorris made. |
|
@mikemorris @youngnick @rikatz PTAL when you get a chance. I'm hoping to get this in sometime this week so I can create a follow up PR that adds |
- Move SessionPersistence to be within EndpointSelector in Backend - Specify relationship with backend protocol - Clarify that the cookie/header MAY not be sent to the backend
|
This basically LGTM once @rikatz's questions are sorted. |
|
@youngnick @rikatz thanks ready for review again |
Update from MAY to SHOULD for whether the backend behavior of receiving the gateway generated cookie based on prior art.
0a6b0ef to
23dcf0a
Compare
|
LGTM now, thanks @gcs278. /approve |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: gcs278, youngnick The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
/lgtm thanks! we may still need to validate that the protocol used is not some that does not support session persistence based on cookie, I would like to see it on conformance and once moving to experimental for real, but for now this is fine |
|
/hold cancel |
|
/test pull-gateway-api-verify |
Implement Session Persistence API as described in kubernetes-sigs#4876 in Backend spec.endpointSelector. Update GEP-4894 to reflect Session Persistence is no longer future item.
Implement Session Persistence API as described in kubernetes-sigs#4876 in Backend spec.endpointSelector. Update GEP-4894 to reflect Session Persistence is no longer future item.
Implement Session Persistence API as described in kubernetes-sigs#4876 in Backend spec.endpointSelector. Update GEP-4894 to reflect Session Persistence is no longer future item.
* GEP-1619: Add Session Persistence API to Backend Implement Session Persistence API as described in #4876 in Backend spec.endpointSelector. Update GEP-4894 to reflect Session Persistence is no longer future item. * GEP-1619: Move SessionPersistence to Top Level of Backend Refactor SessionPersistence API to move to the top level of backend so that it can potentially be re-used with other object references of backend in the future. Use CEL to restrict it to EndpointSelector for now. * GEP-4894: Make inlined labelselector optional The spec.endpointSelector.labelSelector field was inlined and has no JSON property name which causes all objects to fail with invalid required: [""] error. The matchLabel field within labelSelector is already required, while the inlined struct should be optional. * GEP-1619: Code Review Responses - Update CEL to use ternary and CEL message - Fix double negative of API doc - Remove session persistence from GEP-4894
What type of PR is this?
/kind gep
What this PR does / why we need it:
Champions the Backend resource (GEP-4894) as the primary attachment point for session persistence, deprecating route-inline and BackendTrafficPolicy session persistence. Updates the GEP API sections, edge cases, and examples to reflect Backend. Adds deprecation notices to the
sessionPersistencefields on HTTPRouteRule, GRPCRouteRule, and BackendTrafficPolicy.This PR is a declaration of direction, not an implementation. It does not actually add
sessionPersistenceto the Backend API - that is blocked on Backend supporting Service or EndpointSelector bindings #4731.Which issue(s) this PR fixes:
Fixes # N/A
Related: #4462 #4258
Does this PR introduce a user-facing change?: