first pass at a garbage collector

Tiltfile

@@ -66,6 +66,17 @@ k8s_resource(
   ]
 )
 
+k8s_resource(
+  workload = 'kargo-garbage-collector',
+  new_name = 'garbage-collector',
+  labels = ['kargo'],
+  objects = [
+    'kargo-garbage-collector:clusterrole',
+    'kargo-garbage-collector:clusterrolebinding',
+    'kargo-garbage-collector:serviceaccount'
+  ]
+)
+
 k8s_resource(
   workload = 'kargo-webhooks-server',
   new_name = 'webhooks-server',

charts/kargo/templates/_helpers.tpl

@@ -28,6 +28,10 @@ If release name contains chart name it will be used as a full name.
 {{ include "kargo.fullname" . | printf "%s-controller" }}
 {{- end -}}
 
+{{- define "kargo.garbageCollector.fullname" -}}
+{{ include "kargo.fullname" . | printf "%s-garbage-collector" }}
+{{- end -}}
+
 {{- define "kargo.webhooksServer.fullname" -}}
 {{ include "kargo.fullname" . | printf "%s-webhooks-server" }}
 {{- end -}}
@@ -76,6 +80,10 @@ app.kubernetes.io/component: api
 app.kubernetes.io/component: controller
 {{- end -}}
 
+{{- define "kargo.garbageCollector.labels" -}}
+app.kubernetes.io/component: garbage-collector
+{{- end -}}
+
 {{- define "kargo.webhooksServer.labels" -}}
 app.kubernetes.io/component: webhooks-server
 {{- end -}}

charts/kargo/templates/garbage-collector/cluster-role-bindings.yaml

@@ -0,0 +1,17 @@
+{{- if and .Values.garbageCollector.enabled .Values.rbac.installClusterRoleBindings }}
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  name: {{ include "kargo.garbageCollector.fullname" . }}
+  labels:
+    {{- include "kargo.labels" . | nindent 4 }}
+    {{- include "kargo.garbageCollector.labels" . | nindent 4 }}
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: {{ include "kargo.garbageCollector.fullname" . }}
+subjects:
+- kind: ServiceAccount
+  namespace: {{ .Release.Namespace }}
+  name: {{ include "kargo.garbageCollector.fullname" . }}
+{{- end }}

charts/kargo/templates/garbage-collector/cluster-roles.yaml

@@ -0,0 +1,27 @@
+{{- if and .Values.garbageCollector.enabled .Values.rbac.installClusterRoles }}
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  name: {{ include "kargo.garbageCollector.fullname" . }}
+  labels:
+    {{- include "kargo.labels" . | nindent 4 }}
+    {{- include "kargo.garbageCollector.labels" . | nindent 4 }}
+rules:
+- apiGroups:
+  - ""
+  resources:
+  - namespaces
+  verbs:
+  - get
+  - list
+  - watch
+- apiGroups:
+  - kargo.akuity.io
+  resources:
+  - promotions
+  verbs:
+  - delete
+  - get
+  - list
+  - watch
+{{- end }}

charts/kargo/templates/garbage-collector/cron-job.yaml

@@ -0,0 +1,40 @@
+{{- if .Values.garbageCollector.enabled }}
+apiVersion: batch/v1
+kind: CronJob
+metadata:
+  name: {{ include "kargo.garbageCollector.fullname" . }}
+  labels:
+    {{- include "kargo.labels" . | nindent 4 }}
+    {{- include "kargo.garbageCollector.labels" . | nindent 4 }}
+spec:
+  schedule: {{ quote .Values.garbageCollector.schedule }}
+  jobTemplate:
+    spec:
+      template:
+        spec:
+          serviceAccountName: {{ include "kargo.garbageCollector.fullname" . }}
+          containers:
+          - name: garbage-collector
+            image: {{ include "kargo.image" . }}
+            imagePullPolicy: {{ .Values.image.pullPolicy }}
+            command: ["/usr/local/bin/kargo", "garbage-collector"]
+            env:
+            - name: NUM_WORKERS
+              value: {{ quote .Values.garbageCollector.workers }}
+            - name: MAX_RETAINED_PROMOTIONS
+              value: {{ quote .Values.garbageCollector.maxRetainedPromotions }}
+            - name: LOG_LEVEL
+              value: {{ .Values.garbageCollector.logLevel }}
+            resources:
+              {{- toYaml .Values.garbageCollector.resources | nindent 14 }}
+          restartPolicy: Never
+          concurrencyPolicy: Forbid
+          {{- with .Values.garbageCollector.nodeSelector }}
+          nodeSelector:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+          {{- with .Values.garbageCollector.tolerations }}
+          tolerations:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+{{- end }}

charts/kargo/templates/garbage-collector/service-account.yaml

@@ -0,0 +1,9 @@
+{{- if .Values.garbageCollector.enabled }}
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ include "kargo.garbageCollector.fullname" . }}
+  labels:
+    {{- include "kargo.labels" . | nindent 4 }}
+    {{- include "kargo.garbageCollector.labels" . | nindent 4 }}
+{{- end }}

charts/kargo/values.yaml

@@ -156,3 +156,37 @@ webhooksServer:
   nodeSelector: {}
 
   tolerations: []
+
+garbageCollector:
+
+  ## Whether the garbage collector is enabled.
+  enabled: true
+
+  ## When to run the garbage collector.
+  schedule: "0 * * * *"
+  ## The number of concurrent workers to run. Tuning this too low will result in
+  ## slow garbage collection. Tuning this too high will result in too many API
+  ## calls and may result in throttling.
+  workers: 3
+  ## The maximum number of Promotions in terminal phases PER PROJECT that may be
+  ## spared by the garbage collector.
+  maxRetainedPromotions: 20
+
+  logLevel: INFO
+
+  resources: {}
+    # We usually recommend not to specify default resources and to leave this as
+    # a conscious choice for the user. This also increases chances charts run on
+    # environments with little resources, such as KinD. If you do want to
+    # specify resources, uncomment the following lines, adjust them as
+    # necessary, and remove the curly braces after 'resources:'.
+    # limits:
+    #   cpu: 100m
+    #   memory: 128Mi
+    # requests:
+    #   cpu: 100m
+    #   memory: 128Mi
+
+  nodeSelector: {}
+
+  tolerations: []

cmd/controlplane/garbage_collector.go

@@ -0,0 +1,60 @@
+package main
+
+import (
+	"github.com/pkg/errors"
+	log "github.com/sirupsen/logrus"
+	"github.com/spf13/cobra"
+	corev1 "k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/runtime"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+
+	api "github.com/akuity/kargo/api/v1alpha1"
+	"github.com/akuity/kargo/internal/garbage"
+	"github.com/akuity/kargo/internal/os"
+	versionpkg "github.com/akuity/kargo/internal/version"
+)
+
+func newGarbageCollectorCommand() *cobra.Command {
+	return &cobra.Command{
+		Use:               "garbage-collector",
+		DisableAutoGenTag: true,
+		SilenceErrors:     true,
+		SilenceUsage:      true,
+		RunE: func(cmd *cobra.Command, args []string) error {
+			ctx := cmd.Context()
+
+			version := versionpkg.GetVersion()
+			log.WithFields(log.Fields{
+				"version": version.Version,
+				"commit":  version.GitCommit,
+			}).Info("Starting Kargo Garbage Collector")
+
+			cfg := garbage.CollectorConfigFromEnv()
+
+			var kubeClient client.Client
+			{
+				restCfg, err := getRestConfig(ctx, os.GetEnv("KUBECONFIG", ""))
+				if err != nil {
+					return errors.Wrap(err, "error loading REST config")
+				}
+				scheme := runtime.NewScheme()
+				if err = corev1.AddToScheme(scheme); err != nil {
+					return errors.Wrap(err, "error adding Kubernetes core API to scheme")
+				}
+				if err = api.AddToScheme(scheme); err != nil {
+					return errors.Wrap(err, "error adding Kargo API to scheme")
+				}
+				if kubeClient, err = client.New(
+					restCfg,
+					client.Options{
+						Scheme: scheme,
+					},
+				); err != nil {
+					return errors.Wrap(err, "error initializing Kubernetes client")
+				}
+			}
+
+			return garbage.NewCollector(kubeClient, cfg).Run(ctx)
+		},
+	}
+}

cmd/controlplane/root.go

@@ -21,6 +21,7 @@ var (
 func Execute(ctx context.Context) error {
 	rootCmd.AddCommand(newAPICommand())
 	rootCmd.AddCommand(newControllerCommand())
+	rootCmd.AddCommand(newGarbageCollectorCommand())
 	rootCmd.AddCommand(newVersionCommand())
 	rootCmd.AddCommand(newWebhooksServerCommand())
 	return rootCmd.ExecuteContext(ctx)