Add script to cleanup podman on a single cluster node

scripts/podman-cleanup

+#!/bin/bash
+#
+# This script removes unused resources for podman on cluster nodes in the bcf slurm cluster.
+# It assumes that the podman run directory is located at /tmp/podman-run-${UID}. It will delete
+# unused images, stopped containers, unused volumes and the podman run directory.
+#
+set -eo pipefail
+PODMAN_DIR=/tmp/podman-run-${UID}
+
+if [[ -e $PODMAN_DIR ]]
+then
+  echo "Podman found on $(hostname)"
+  RUNNING_COUNT=$(podman ps --noheading | wc -l)
+  podman system prune -a -f
+  if [[ $RUNNING_COUNT -gt 0 ]]
+  then
+    echo "You still have ${RUNNING_COUNT} jobs running on $(hostname)"
+  else
+    rm -rf $PODMAN_DIR
+  fi
+fi