chore: Cleanup helm chart

.github/workflows/e2e.yml

@@ -113,7 +113,7 @@ jobs:
 
       - name: Run e2e test
         run: |
-          OUTPUT_TYPE=type=registry make e2etests
+         make e2etests
 
       - name: Cleanup e2e resources
         if: ${{ always() }}

charts/gpu-provisioner/values.yaml

@@ -98,18 +98,10 @@ tolerations:
     operator: Exists
 # -- Additional volumes for the pod.
 extraVolumes: []
-# - name: aws-iam-token
-#   projected:
-#     defaultMode: 420
-#     sources:
-#     - serviceAccountToken:
-#         audience: sts.amazonaws.com
-#         expirationSeconds: 86400
-#         path: token
 controller:
   image:
     # -- Repository path to the controller image.
-    repository: aimodelsregistry.azurecr.io/gpu-provisioner
+    repository: ghcr.io/azure/gpu-provisioner
     # -- Tag of the controller image.
     tag: 0.2.0
     # -- SHA256 digest of the controller image.
@@ -156,10 +148,6 @@ controller:
   logEncoding: ""
   # -- Additional volumeMounts for the controller pod.
   extraVolumeMounts: []
-  # - name: aws-iam-token
-  #   mountPath: /var/run/secrets/eks.amazonaws.com/serviceaccount
-  #   readOnly: true
-  # -- Additional sidecarContainer config
   sidecarContainer: []
   # -- Additional volumeMounts for the sidecar - this will be added to the volume mounts on top of extraVolumeMounts
   sidecarVolumeMounts: []
@@ -175,33 +163,13 @@ logLevel: debug
 logEncoding: console
 # -- Global Settings to configure gpu-provisioner
 settings:
-  # -- The maximum length of a batch window. The longer this is, the more pods we can consider for provisioning at one
-  # time which usually results in fewer but larger nodes.
-  batchMaxDuration: 10s
-  # -- The maximum amount of time with no new ending pods that if exceeded ends the current batching window. If pods arrive
-  # faster than this time, the batching window will be extended up to the maxDuration. If they arrive slower, the pods
-  # will be batched separately.
-  batchIdleDuration: 1s
   # -- Azure-specific configuration values
   azure:
     # -- Cluster name.
     clusterName: new_demo
-    # -- Cluster endpoint.
-    clusterEndpoint: https://newdemo-llm-test-ff05f5-vs68qyru.hcp.eastus.azmk8s.io:443
-    # -- The VM memory overhead as a percent that will be subtracted from the total memory for all instance types
-    # TODO: not used yet ...
-    vmMemoryOverheadPercent: 0.075
-    # TODO: autogenerate
-    sshPublicKey:
-    networkPlugin: "kubenet" # TODO: get this from the cluster
-    networkPolicy: ""
     # -- The global tags to use on all Azure infrastructure resources (VMs, etc.)
     # TODO: not propagated yet ...
     tags:
   # -- Feature Gate configuration values. Feature Gates will follow the same graduation process and requirements as feature gates
   # in Kubernetes. More information here https://kubernetes.io/docs/reference/command-line-tools-reference/feature-gates/#feature-gates-for-alpha-or-beta-features
   featureGates:
-    # -- driftEnabled is in ALPHA and is disabled by default.
-    # Setting driftEnabled to true enables the drift deprovisioner to watch for drift between currently deployed nodes
-    # and the desired state of nodes set in provisioners and node templates
-    driftEnabled: false

pkg/apis/settings/settings.go

@@ -18,9 +18,6 @@ import (
 	"context"
 	"encoding/json"
 	"fmt"
-	"hash/fnv"
-	"math/rand"
-	"net/url"
 
 	"github.com/go-playground/validator/v10"
 	"go.uber.org/multierr"
@@ -33,32 +30,14 @@ type settingsKeyType struct{}
 var ContextKey = settingsKeyType{}
 
 var defaultSettings = Settings{
-	ClusterName:             "",
-	ClusterEndpoint:         "",
-	VMMemoryOverheadPercent: 0.075,
-	Tags:                    map[string]string{},
-	ClusterID:               "",
-	SSHPublicKey:            "",
-	NetworkPlugin:           "",
-	NetworkPolicy:           "",
+	ClusterName: "",
+	Tags:        map[string]string{},
 }
 
 // +k8s:deepcopy-gen=true
 type Settings struct {
-	ClusterName             string  `validate:"required"`
-	ClusterEndpoint         string  `validate:"required"` // => APIServerName in bootstrap, except needs to be w/o https/port
-	VMMemoryOverheadPercent float64 `validate:"min=0"`
-	Tags                    map[string]string
-
-	// Cluster-level settings required for nodebootstrap (category "x")
-	// (Candidates for exposure/accessibility via API)
-	// TODO: consider making these AKS-specific (e.g. subkey?)
-
-	ClusterID string
-
-	SSHPublicKey  string // ssh.publicKeys.keyData => VM SSH public key // TODO: move to node template?
-	NetworkPlugin string `validate:"required"` // => NetworkPlugin in bootstrap
-	NetworkPolicy string // => NetworkPolicy in bootstrap
+	ClusterName string `validate:"required"`
+	Tags        map[string]string
 }
 
 func (*Settings) ConfigMap() string {
@@ -71,25 +50,14 @@ func (*Settings) Inject(ctx context.Context, cm *v1.ConfigMap) (context.Context,
 
 	if err := configmap.Parse(cm.Data,
 		configmap.AsString("azure.clusterName", &s.ClusterName),
-		configmap.AsString("azure.clusterEndpoint", &s.ClusterEndpoint),
-		configmap.AsFloat64("azure.vmMemoryOverheadPercent", &s.VMMemoryOverheadPercent),
 		AsStringMap("azure.tags", &s.Tags),
-		configmap.AsString("azure.clusterID", &s.ClusterID),
-		configmap.AsString("azure.sshPublicKey", &s.SSHPublicKey),
-		configmap.AsString("azure.networkPlugin", &s.NetworkPlugin),
-		configmap.AsString("azure.networkPolicy", &s.NetworkPolicy),
 	); err != nil {
 		return ctx, fmt.Errorf("parsing settings, %w", err)
 	}
 	if err := s.Validate(); err != nil {
 		return ctx, fmt.Errorf("validating settings, %w", err)
 	}
 
-	// if clusterID is not set, generate it from cluster endpoint
-	if s.ClusterID == "" {
-		s.ClusterID = getAKSClusterID(s.GetAPIServerName())
-	}
-
 	return ToContext(ctx, s), nil
 }
 
@@ -115,26 +83,10 @@ func (s Settings) Data() (map[string]string, error) {
 func (s Settings) Validate() error {
 	validate := validator.New()
 	return multierr.Combine(
-		s.validateEndpoint(),
 		validate.Struct(s),
 	)
 }
 
-func (s Settings) validateEndpoint() error {
-	endpoint, err := url.Parse(s.ClusterEndpoint)
-	// url.Parse() will accept a lot of input without error; make
-	// sure it's a real URL
-	if err != nil || !endpoint.IsAbs() || endpoint.Hostname() == "" {
-		return fmt.Errorf("\"%s\" not a valid clusterEndpoint URL", s.ClusterEndpoint)
-	}
-	return nil
-}
-
-func (s Settings) GetAPIServerName() string {
-	endpoint, _ := url.Parse(s.ClusterEndpoint) // already validated
-	return endpoint.Hostname()
-}
-
 func ToContext(ctx context.Context, s *Settings) context.Context {
 	return context.WithValue(ctx, ContextKey, s)
 }
@@ -148,16 +100,6 @@ func FromContext(ctx context.Context) *Settings {
 	return data.(*Settings)
 }
 
-// AsTypedString passes the value at key through into the target, if it exists.
-func AsTypedString[T ~string](key string, target *T) configmap.ParseFunc {
-	return func(data map[string]string) error {
-		if raw, ok := data[key]; ok {
-			*target = T(raw)
-		}
-		return nil
-	}
-}
-
 // AsStringMap parses a value as a JSON map of map[string]string.
 func AsStringMap(key string, target *map[string]string) configmap.ParseFunc {
 	return func(data map[string]string) error {
@@ -171,14 +113,3 @@ func AsStringMap(key string, target *map[string]string) configmap.ParseFunc {
 		return nil
 	}
 }
-
-// getAKSClusterID returns cluster ID based on the DNS prefix of the cluster.
-// The logic comes from AgentBaker and other places, originally from aks-engine
-// with the additional assumption of DNS prefix being the first 33 chars of FQDN
-func getAKSClusterID(apiServerFQDN string) string {
-	dnsPrefix := apiServerFQDN[:33]
-	h := fnv.New64a()
-	h.Write([]byte(dnsPrefix))
-	r := rand.New(rand.NewSource(int64(h.Sum64()))) //nolint:gosec
-	return fmt.Sprintf("%08d", r.Uint32())[:8]
-}