[RayService] Add New Status: NumServeEndpoints #1901
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
Yicheng-Lu-llll
force-pushed
the
add-num-endpoint-status
branch
from
acfbf4d to
377d18f
Compare
Yicheng-Lu-llll
changed the title
Yicheng-Lu-llll
force-pushed
the
add-num-endpoint-status
branch
from
588139b to
8110ea8
Compare
kevin85421
reviewed
Feb 7, 2024
| @@ -71,6 +71,9 @@ type RayServiceStatuses struct { | |||
| PendingServiceStatus RayServiceStatus `json:"pendingServiceStatus,omitempty"` | |||
| // ServiceStatus indicates the current RayService status. | |||
| ServiceStatus ServiceStatus `json:"serviceStatus,omitempty"` | |||
| // ActiveServingRayPods indicates the number of Ray Pods that are actively serving or have been selected by the serve service. | |||
| // Ray Pods without a proxy actor or those that are unhealthy will not be counted. | |||
| ActiveServingRayPods int32 `json:"ActiveServingRayPods,omitempty"` | |||
Yicheng-Lu-llll
force-pushed
the
add-num-endpoint-status
branch
from
d3690bc to
4680dd6
Compare
Yicheng-Lu-llll
changed the title
Signed-off-by: Yicheng-Lu-llll <[email protected]>
Yicheng-Lu-llll
force-pushed
the
add-num-endpoint-status
branch
from
f1f67ab to
0b3cce2
Compare
kevin85421
reviewed
Feb 16, 2024
| } | ||
|
|
||
| numServeEndpoints := 0 | ||
| for _, subset := range serveEndPoints.Subsets { |
There was a problem hiding this comment.
I am not familiar with endpoints. What is Subsets?
There was a problem hiding this comment.
These links explain it: https://pkg.go.dev/k8s.io/api/core/v1#Endpoints
https://pkg.go.dev/k8s.io/api/core/v1#EndpointSubset
kevin85421
reviewed
Feb 16, 2024
| @@ -224,6 +229,25 @@ func (r *RayServiceReconciler) Reconcile(ctx context.Context, request ctrl.Reque | |||
| return ctrl.Result{RequeueAfter: ServiceDefaultRequeueDuration}, nil | |||
| } | |||
|
|
|||
| func (r *RayServiceReconciler) calculateStatus(ctx context.Context, rayServiceInstance *rayv1.RayService, rayClusterInstance *rayv1.RayCluster) error { | |||
| serveSvc, err := common.BuildServeServiceForRayService(ctx, *rayServiceInstance, *rayClusterInstance) | |||
There was a problem hiding this comment.
TODO: Associate RayService and its K8s service in a smarter way.
kevin85421
approved these changes
Feb 16, 2024
Signed-off-by: Yicheng-Lu-llll <[email protected]>
2 tasks
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.
Why are these changes needed?
This PR adds the
NumServeEndpointsfield to the RayService's status. TheNumServeEndpointsfield indicates the number of active serving Ray Pods or the number of Ray Pods selected by the serve service. If a Ray Pod has no proxy actor or is unhealthy, it will not be counted.This new field can help users and CI tests determine how many Ray Pods are capable of serving, making debugging much easier. This is crucial because, with the high availability feature of RayService, even if some worker Pods become unhealthy, it is hard to tell since everything appears to work just fine on the surface.
Related issue number
Checks