fix: commitments package refactoring/pipeline config (#601)

## Changes
- commitment uuid format enforced
- reservation deletion considers allocations and targetHost
- commitments use configurable pipeline for scheduling
- set validation time
- refactor package and controller
- refactor logging

Author: mblos

cmd/main.go

@@ -51,7 +51,6 @@ import (
 	"github.com/cobaltcore-dev/cortex/internal/scheduling/pods"
 	"github.com/cobaltcore-dev/cortex/internal/scheduling/reservations"
 	"github.com/cobaltcore-dev/cortex/internal/scheduling/reservations/commitments"
-	reservationscontroller "github.com/cobaltcore-dev/cortex/internal/scheduling/reservations/controller"
 	"github.com/cobaltcore-dev/cortex/internal/scheduling/reservations/failover"
 	"github.com/cobaltcore-dev/cortex/pkg/conf"
 	"github.com/cobaltcore-dev/cortex/pkg/monitoring"
@@ -487,18 +486,37 @@ func main() {
 			os.Exit(1)
 		}
 	}
-	if slices.Contains(mainConfig.EnabledControllers, "reservations-controller") {
-		setupLog.Info("enabling controller", "controller", "reservations-controller")
-		monitor := reservationscontroller.NewControllerMonitor(multiclusterClient)
+	if slices.Contains(mainConfig.EnabledControllers, "committed-resource-reservations-controller") {
+		setupLog.Info("enabling controller", "controller", "committed-resource-reservations-controller")
+		monitor := reservations.NewMonitor(multiclusterClient)
 		metrics.Registry.MustRegister(&monitor)
-		reservationsControllerConfig := conf.GetConfigOrDie[reservationscontroller.Config]()
+		commitmentsConfig := conf.GetConfigOrDie[commitments.Config]()
+		commitmentsDefaults := commitments.DefaultConfig()
+		if commitmentsConfig.RequeueIntervalActive == 0 {
+			commitmentsConfig.RequeueIntervalActive = commitmentsDefaults.RequeueIntervalActive
+		}
+		if commitmentsConfig.RequeueIntervalRetry == 0 {
+			commitmentsConfig.RequeueIntervalRetry = commitmentsDefaults.RequeueIntervalRetry
+		}
+		if commitmentsConfig.PipelineDefault == "" {
+			commitmentsConfig.PipelineDefault = commitmentsDefaults.PipelineDefault
+		}
+		if commitmentsConfig.SchedulerURL == "" {
+			commitmentsConfig.SchedulerURL = commitmentsDefaults.SchedulerURL
+		}
+		if commitmentsConfig.ChangeAPIWatchReservationsTimeout == 0 {
+			commitmentsConfig.ChangeAPIWatchReservationsTimeout = commitmentsDefaults.ChangeAPIWatchReservationsTimeout
+		}
+		if commitmentsConfig.ChangeAPIWatchReservationsPollInterval == 0 {
+			commitmentsConfig.ChangeAPIWatchReservationsPollInterval = commitmentsDefaults.ChangeAPIWatchReservationsPollInterval
+		}
 
-		if err := (&reservationscontroller.ReservationReconciler{
+		if err := (&commitments.CommitmentReservationController{
 			Client: multiclusterClient,
 			Scheme: mgr.GetScheme(),
-			Conf:   reservationsControllerConfig,
+			Conf:   commitmentsConfig,
 		}).SetupWithManager(mgr, multiclusterClient); err != nil {
-			setupLog.Error(err, "unable to create controller", "controller", "Reservation")
+			setupLog.Error(err, "unable to create controller", "controller", "CommitmentReservation")
 			os.Exit(1)
 		}
 	}
@@ -677,9 +695,13 @@ func main() {
 		setupLog.Info("starting commitments syncer")
 		syncer := commitments.NewSyncer(multiclusterClient)
 		syncerConfig := conf.GetConfigOrDie[commitments.SyncerConfig]()
+		syncerDefaults := commitments.DefaultSyncerConfig()
+		if syncerConfig.SyncInterval == 0 {
+			syncerConfig.SyncInterval = syncerDefaults.SyncInterval
+		}
 		if err := (&task.Runner{
 			Client:   multiclusterClient,
-			Interval: time.Hour,
+			Interval: syncerConfig.SyncInterval,
 			Name:     "commitments-sync-task",
 			Run:      func(ctx context.Context) error { return syncer.SyncReservations(ctx) },
 			Init:     func(ctx context.Context) error { return syncer.Init(ctx, syncerConfig) },

helm/bundles/cortex-nova/templates/pipelines_kvm.yaml

@@ -236,6 +236,7 @@ spec:
 
     This is the pipeline used for KVM hypervisors (qemu and cloud-hypervisor).
     Specifically, this pipeline is used for general purpose workloads.
+    It is also used for (CR/HA) reservation requests.
   type: filter-weigher
   createDecisions: false
   # Fetch all placement candidates, ignoring nova's preselection.

helm/bundles/cortex-nova/values.yaml

@@ -135,13 +135,33 @@ cortex-scheduling-controllers:
       - nova-deschedulings-executor
       - hypervisor-overcommit-controller
       - explanation-controller
-      - reservations-controller
+      - committed-resource-reservations-controller
       - failover-reservations-controller
     enabledTasks:
       - nova-decisions-cleanup-task
-    # Endpoints configuration for reservations controller
-    endpoints:
-      novaExternalScheduler: "http://localhost:8080/scheduler/nova/external"
+    # CommittedResourceFlavorGroupPipelines maps flavor group IDs to pipeline names for CR reservations
+    # This allows different scheduling strategies per flavor group (e.g., HANA vs GP)
+    committedResourceFlavorGroupPipelines:
+      "2152": "kvm-hana-bin-packing-all-filters-enabled"  # HANA flavor group
+      "2101": "kvm-general-purpose-load-balancing-all-filters-enabled"  # General Purpose flavor group
+      "*": "kvm-general-purpose-load-balancing-all-filters-enabled"  # Catch-all fallback
+    # Default pipeline for CR reservations when no CommittedResourceFlavorGroupPipelines entry matches
+    committedResourcePipelineDefault: "kvm-general-purpose-load-balancing-all-filters-enabled"
+    # How often to re-verify active reservations
+    # 5m = 300000000000 nanoseconds
+    committedResourceRequeueIntervalActive: 300000000000
+    # How often to retry when knowledge is not ready
+    # 1m = 60000000000 nanoseconds
+    committedResourceRequeueIntervalRetry: 60000000000
+    # Timeout for watching reservations to become ready before rolling back
+    # 10s = 10000000000 nanoseconds
+    committedResourceChangeAPIWatchReservationsTimeout: 10000000000
+    # How often to poll reservation status during watch
+    # 500ms = 500000000 nanoseconds
+    committedResourceChangeAPIWatchReservationsPollInterval: 500000000
+    # Whether the change-commitments API endpoint is active
+    # When false, the endpoint returns HTTP 503. The info endpoint remains available.
+    committedResourceEnableChangeCommitmentsAPI: true
     # OvercommitMappings is a list of mappings that map hypervisor traits to
     # overcommit ratios. Note that this list is applied in order, so if there
     # are multiple mappings applying to the same hypervisors, the last mapping
@@ -189,6 +209,9 @@ cortex-knowledge-controllers:
       - datasource-controllers
       - knowledge-controllers
       - kpis-controller
+    # How often the commitments syncer reconciles Limes commitments to Reservation CRDs
+    # 1h = 3600000000000 nanoseconds
+    committedResourceSyncInterval: 3600000000000
     enabledTasks:
       - commitments-sync-task