diff --git a/kubernetes/apps/kube-system/etcd-defrag/app/helmrelease.yaml b/kubernetes/apps/kube-system/etcd-defrag/app/helmrelease.yaml
deleted file mode 100644
index bc63ab84c0..0000000000
--- a/kubernetes/apps/kube-system/etcd-defrag/app/helmrelease.yaml
+++ /dev/null
@@ -1,82 +0,0 @@
----
-# yaml-language-server: $schema=https://raw.githubusercontent.com/bjw-s-labs/helm-charts/main/charts/other/app-template/schemas/helmrelease-helm-v2.schema.json
-apiVersion: helm.toolkit.fluxcd.io/v2
-kind: HelmRelease
-metadata:
-  name: etcd-defrag
-spec:
-  interval: 30m
-  chartRef:
-    kind: OCIRepository
-    name: app-template
-
-  values:
-    defaultPodOptions:
-      # Must use host network and PID namespace to access Talos etcd on localhost:2379.
-      hostNetwork: true
-      hostPID: true
-      securityContext:
-        runAsGroup: 0
-        runAsNonRoot: false
-        runAsUser: 0
-      nodeSelector:
-        node-role.kubernetes.io/control-plane: ""
-      tolerations:
-        - key: node-role.kubernetes.io/control-plane
-          operator: Exists
-          effect: NoSchedule
-
-    controllers:
-      etcd-defrag:
-        type: cronjob
-        cronjob:
-          # The previous weekly cadence let fragmentation cross the 50% alert
-          # threshold by day 4; a full 3-member pass costs ~95s.
-          timeZone: ${TIMEZONE}
-          schedule: "0 3 * * *"
-          backoffLimit: 1
-        containers:
-          app:
-            image:
-              repository: ghcr.io/ahrtr/etcd-defrag
-              tag: v0.42.0@sha256:5138a12eda39067dabcf1d420d66f54f29f1e9297109a6b36574e010b45d58cc
-            args:
-              - --endpoints=https://127.0.0.1:2379
-              - --cluster
-              - --cacert=/certs/ca.crt
-              - --cert=/certs/server.crt
-              - --key=/certs/server.key
-              - --defrag-rule=dbQuotaUsage > 0.5 || dbSizeFree > 50*1024*1024
-              - --move-leader
-              # --cluster defragments all three control-plane etcd members.
-              - --wait-between-defrags=30s
-              # Talos uses etcd's default --quota-backend-bytes value (2 GiB).
-              - --etcd-storage-quota-bytes=2147483648
-              - --auto-disalarm
-              - --disalarm-threshold=0.8
-            resources:
-              requests:
-                cpu: 25m
-                memory: 128Mi
-              limits:
-                cpu: 100m
-                memory: 128Mi
-            securityContext:
-              # Talos etcd's client key is root-readable only on the host.
-              privileged: true
-
-    persistence:
-      certs:
-        type: hostPath
-        # Talos EtcdPKIPath.
-        hostPath: /system/secrets/etcd
-        globalMounts:
-          - path: /certs
-            readOnly: true
-      tmpfs:
-        type: emptyDir
-        advancedMounts:
-          etcd-defrag:
-            app:
-              - path: /tmp
-                subPath: tmp
diff --git a/kubernetes/apps/kube-system/etcd-defrag/app/kustomization.yaml b/kubernetes/apps/kube-system/etcd-defrag/app/kustomization.yaml
deleted file mode 100644
index 904bfa2497..0000000000
--- a/kubernetes/apps/kube-system/etcd-defrag/app/kustomization.yaml
+++ /dev/null
@@ -1,6 +0,0 @@
----
-# yaml-language-server: $schema=https://json.schemastore.org/kustomization
-apiVersion: kustomize.config.k8s.io/v1beta1
-kind: Kustomization
-resources:
-  - helmrelease.yaml
diff --git a/kubernetes/apps/kube-system/etcd-defrag/ks.yaml b/kubernetes/apps/kube-system/etcd-defrag/ks.yaml
deleted file mode 100644
index 006d07a9bb..0000000000
--- a/kubernetes/apps/kube-system/etcd-defrag/ks.yaml
+++ /dev/null
@@ -1,17 +0,0 @@
----
-# yaml-language-server: $schema=https://k8s-schemas.home-operations.com/kustomize.toolkit.fluxcd.io/kustomization_v1.json
-apiVersion: kustomize.toolkit.fluxcd.io/v1
-kind: Kustomization
-metadata:
-  name: etcd-defrag
-  namespace: &namespace kube-system
-spec:
-  interval: 1h
-  path: ./kubernetes/apps/kube-system/etcd-defrag/app
-  prune: true
-  sourceRef:
-    kind: GitRepository
-    name: flux-system
-    namespace: flux-system
-  targetNamespace: *namespace
-  wait: false
diff --git a/kubernetes/apps/kube-system/kustomization.yaml b/kubernetes/apps/kube-system/kustomization.yaml
index e85aed0a04..52e55b1913 100644
--- a/kubernetes/apps/kube-system/kustomization.yaml
+++ b/kubernetes/apps/kube-system/kustomization.yaml
@@ -18,4 +18,3 @@ resources:
   - ./reloader/ks.yaml
   - ./snapshot-controller/ks.yaml
   - ./spegel/ks.yaml
-  - ./etcd-defrag/ks.yaml
diff --git a/talos/patches/controller/cluster.yaml b/talos/patches/controller/cluster.yaml
index c2c4edc4ef..5e0d9a5ab3 100644
--- a/talos/patches/controller/cluster.yaml
+++ b/talos/patches/controller/cluster.yaml
@@ -14,6 +14,11 @@ cluster:
   etcd:
     extraArgs:
       listen-metrics-urls: http://0.0.0.0:2381
+      # Keep the etcd database lean by compacting old revisions hourly; this
+      # replaces the nightly defrag job (compaction prevents growth, defrag
+      # only reclaims space the file already holds).
+      auto-compaction-mode: periodic
+      auto-compaction-retention: 1h
     advertisedSubnets: [192.168.0.0/24]
   proxy:
     disabled: true