diff --git a/tests/integration/kubernetes/gha-run.sh b/tests/integration/kubernetes/gha-run.sh
index e30cf77f143e..89b1055aa488 100755
--- a/tests/integration/kubernetes/gha-run.sh
+++ b/tests/integration/kubernetes/gha-run.sh
@@ -436,6 +436,33 @@ function cleanup_snapshotter() {
 	echo "::endgroup::"
 }
 
+function reset_k8s_images(){
+	# we are encountering the issue (https://github.com/kata-containers/kata-containers/issues/8407)
+	# with containerd on CI is likely due to the content digest being missing from the content store,
+	# which can happen when switching between different snapshotters.
+	# To help sort it out, we now clean up or fetch related images in k8s.io namespace.
+	ctr_args=""
+	if [ "${KUBERNETES}" = "k3s" ]; then
+		ctr_args="--address /run/k3s/containerd/containerd.sock "
+	fi
+	ctr_args+="--namespace k8s.io"
+	ctr_command="sudo -E ctr ${ctr_args}"
+	pause_images=$(${ctr_command} i ls | grep 'pause' | awk '{print $1}')
+	for image in $pause_images; do
+		${ctr_command} content fetch "$image"
+	done
+	# remove related images in k8s.io namespace
+	images_to_remove=(
+		"quay.io/sjenning/nginx:1.15-alpine"
+		"quay.io/prometheus/busybox:latest"
+		"quay.io/confidential-containers/test-images:largeimage"
+	)
+
+	for image in "${images_to_remove[@]}"; do
+		${ctr_command} i remove "$image" || true
+	done
+}
+
 function deploy_nydus_snapshotter() {
     echo "::group::deploy_nydus_snapshotter"
 	ensure_yq
@@ -497,6 +524,10 @@ function deploy_nydus_snapshotter() {
 	kubectl logs "${pods_name}" -n nydus-system
 	kubectl describe pod "${pods_name}" -n nydus-system
 	echo "::endgroup::"
+	echo "::group::reset some k8s images"
+	sleep 10
+	reset_k8s_images
+	echo "::endgroup::"
 }
 
 function cleanup_nydus_snapshotter() {