diff --git a/pkg/service/node_static_inline.go b/pkg/service/node_static_inline.go
index cc9add1..4e4f9dc 100644
--- a/pkg/service/node_static_inline.go
+++ b/pkg/service/node_static_inline.go
@@ -25,8 +25,10 @@ func (s *Service) nodePublishVolumeStaticInlineVolume(ctx context.Context, volum
 	duration := time.Since(startedAt)
 	logger.WithContext(ctx).Infof("pulled model: %s %s", reference, duration)
 
+	mountCtx, mountCancel := context.WithTimeout(context.WithoutCancel(ctx), 30*time.Second)
+	defer mountCancel()
 	if err := mounter.Mount(
-		ctx,
+		mountCtx,
 		mounter.NewBuilder().
 			Bind().
 			From(modelDir).
diff --git a/pkg/service/worker.go b/pkg/service/worker.go
index 8628736..98dd49f 100644
--- a/pkg/service/worker.go
+++ b/pkg/service/worker.go
@@ -165,8 +165,10 @@ func (worker *Worker) pullModel(ctx context.Context, statusPath, volumeName, mou
 		}
 		defer worker.kmutex.Unlock(contextKey)
 
+		// Decouple from the kubelet RPC deadline so that large pulls aren't killed
+		// when kubelet times out and retries
 		var cancel context.CancelFunc
-		ctx, cancel = context.WithCancel(ctx)
+		ctx, cancel = context.WithCancel(context.WithoutCancel(ctx))
 		worker.contextMap.Set(contextKey, &cancel)
 		defer worker.contextMap.Set(contextKey, nil)