diff --git a/overlay.d/05core/usr/lib/dracut/modules.d/40ignition-ostree/ignition-ostree-transposefs.sh b/overlay.d/05core/usr/lib/dracut/modules.d/40ignition-ostree/ignition-ostree-transposefs.sh
index 6f89c4007e..3ba12a6602 100755
--- a/overlay.d/05core/usr/lib/dracut/modules.d/40ignition-ostree/ignition-ostree-transposefs.sh
+++ b/overlay.d/05core/usr/lib/dracut/modules.d/40ignition-ostree/ignition-ostree-transposefs.sh
@@ -139,13 +139,20 @@ should_autosave_rootfs() {
         return
     fi
     local agcount
+    # We want to run xfs_info on the unmounted filesystem, because actually just
+    # even mounting an XFS filesystem that has grown an excessive number of allocation groups
+    # can be very slow.
     eval $(xfs_info "${root_part}" | grep -o 'agcount=[0-9]*')
-    # Semi-arbitrarily chosen: this is roughly ~64G currently (based on initial
-    # ag sizing at build time) which seems like a good rootfs size at which to
+    # This is roughly ~700GiB currently (based on initial ag sizing at build time) which
+    # seems like a good rootfs size at which to
     # discriminate between "throwaway/short-lived systems" and "long-running
-    # workload systems". It's not like XFS performance is way worse at 128.
-    if [ "$agcount" -lt 128 ]; then
-        echo "Filesystem agcount is $agcount; skipping" >&2
+    # workload systems".  Specifically for e.g. OpenShift, it's below the default node
+    # sizes.
+    local threshold
+    threshold=400
+    echo "${root_part} agcount=$agcount" >&2
+    if [ "$agcount" -lt "${threshold}" ]; then
+        echo "agcount=$agcount is lower than threshold=${threshold}"
         echo 0
         return
     fi