diff --git a/contrib/cirrus/setup_environment.sh b/contrib/cirrus/setup_environment.sh
index 436024ef69e0..2b9fe34e2eec 100755
--- a/contrib/cirrus/setup_environment.sh
+++ b/contrib/cirrus/setup_environment.sh
@@ -93,6 +93,13 @@ case "$CG_FS_TYPE" in
     *) die_unknown CG_FS_TYPE
 esac
 
+# Force the requested database backend without having to use command-line args
+cat <<EOF >>/etc/containers/containers.conf
+
+[engine]
+database_backend = "$CI_DESIRED_DATABASE"
+EOF
+
 if ((CONTAINER==0)); then  # Not yet running inside a container
     # Discovered reemergence of BFQ scheduler bug in kernel 5.8.12-200
     # which causes a kernel panic when system is under heavy I/O load.
diff --git a/test/system/005-info.bats b/test/system/005-info.bats
index 42ea58a1d634..f97ea8653192 100644
--- a/test/system/005-info.bats
+++ b/test/system/005-info.bats
@@ -94,7 +94,6 @@ host.slirp4netns.executable | $expr_path
 }
 
 @test "podman info - confirm desired database" {
-    skip "FIXME: no way yet (2023-03-16) to override DB in system tests"
     if [[ -z "$CI_DESIRED_DATABASE" ]]; then
         # When running in Cirrus, CI_DESIRED_DATABASE *must* be defined
         # in .cirrus.yml so we can double-check that all CI VMs are