Add README.md file to the epp pkg #386
Merged
Conversation
This file contains 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
k8s-ci-robot
added
do-not-merge/work-in-progress
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: ahg-g The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
k8s-ci-robot
added
approved
✅ Deploy Preview for gateway-api-inference-extension ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
k8s-ci-robot
added
size/M
ahg-g
changed the title
k8s-ci-robot
removed
the
do-not-merge/work-in-progress
ahg-g
changed the title
kfswain
reviewed
Feb 21, 2025
kfswain
reviewed
Feb 21, 2025
pkg/epp/README.md
Outdated
| @@ -0,0 +1,25 @@ | |||
| # The EndPoint Picker (EPP) | |||
| This package provides the reference implementation for the Endpoint Picker (EPP). It implements the [extension protocol](../../docs/proposals/003-endpoint-picker-protocol), enabling a proxy or gateway to request endpoint hints from an extension. As it is implemented now, an EPP instance handles a single `InferencePool` (and so for each `InferencePool`, one must create a dedicated EPP deployment). | |||
There was a problem hiding this comment.
Nit: Consider dropping 'As it is implemented now'. My thinking is: we can just update this when/if we make an EPP multitenant
pkg/epp/README.md
Outdated
|
|
||
| - Endpoint Selection | ||
| - The EPP determines the appropriate Pod endpoint for the load balancer (LB) to route requests. | ||
| - It selects from the pool of ready Pods designated by the assigned InferencePool. |
There was a problem hiding this comment.
Suggested change
| - It selects from the pool of ready Pods designated by the assigned InferencePool. | |
| - It selects from the pool of ready Pods designated by the assigned InferencePool's [Selector](https://github.com/kubernetes-sigs/gateway-api-inference-extension/blob/7e3cd457cdcd01339b65861c8e472cf27e6b6e80/api/v1alpha1/inferencepool_types.go#L53) field. |
pkg/epp/README.md
Outdated
| - It selects from the pool of ready Pods designated by the assigned InferencePool. | ||
| - Endpoint selection is contingent on the request's ModelName matching an `InferenceModel` that references the `InferencePool`. | ||
| - Requests with unmatched ModelName values trigger an error response to the proxy. | ||
| - The endpoint selection algorithm is detailed below. |
There was a problem hiding this comment.
Nit: consider removing, the headers I think are prominent enough to draw attention.
| - Requests with unmatched ModelName values trigger an error response to the proxy. | ||
| - The endpoint selection algorithm is detailed below. | ||
| - Traffic Splitting and ModelName Rewriting | ||
| - The EPP facilitates controlled rollouts of new adapter versions by implementing traffic splitting between adapters within the same `InferencePool`, as defined by the `InferenceModel`. |
There was a problem hiding this comment.
More nits: Linking to the InfModel's targetmodel field could be useful
|
This all looks great! Left some comments, but these are true nits, feel free to disregard. This is already a major improvement from what we had before. Thanks! /lgtm |
k8s-ci-robot
added
the
do-not-merge/hold
k8s-ci-robot
added
the
lgtm
k8s-ci-robot
removed
the
lgtm
ahg-g
commented
Feb 21, 2025
pkg/epp/README.md
Outdated
| @@ -0,0 +1,25 @@ | |||
| # The EndPoint Picker (EPP) | |||
| This package provides the reference implementation for the Endpoint Picker (EPP). It implements the [extension protocol](../../docs/proposals/003-endpoint-picker-protocol), enabling a proxy or gateway to request endpoint hints from an extension. As it is implemented now, an EPP instance handles a single `InferencePool` (and so for each `InferencePool`, one must create a dedicated EPP deployment). | |||
pkg/epp/README.md
Outdated
|
|
||
| - Endpoint Selection | ||
| - The EPP determines the appropriate Pod endpoint for the load balancer (LB) to route requests. | ||
| - It selects from the pool of ready Pods designated by the assigned InferencePool. |
pkg/epp/README.md
Outdated
| - It selects from the pool of ready Pods designated by the assigned InferencePool. | ||
| - Endpoint selection is contingent on the request's ModelName matching an `InferenceModel` that references the `InferencePool`. | ||
| - Requests with unmatched ModelName values trigger an error response to the proxy. | ||
| - The endpoint selection algorithm is detailed below. |
| - Requests with unmatched ModelName values trigger an error response to the proxy. | ||
| - The endpoint selection algorithm is detailed below. | ||
| - Traffic Splitting and ModelName Rewriting | ||
| - The EPP facilitates controlled rollouts of new adapter versions by implementing traffic splitting between adapters within the same `InferencePool`, as defined by the `InferenceModel`. |
k8s-ci-robot
added
size/S
|
/lgtm Thanks! |
k8s-ci-robot
added
the
lgtm
|
/hold cancel |
k8s-ci-robot
removed
the
do-not-merge/hold
kaushikmitr
pushed a commit
to kaushikmitr/llm-instance-gateway
that referenced
this pull request
Feb 27, 2025
* Polish the epp README.md file * Addressed comments
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.
No description provided.