external: make rgw call separate from cephfs and rbd

if the rge endint is not reachable don't block the creation
of rbd and cephfs sc, just print the error instead of raising exception

Signed-off-by: parth-gr <paarora@redhat.com>
(cherry picked from commit e44df84687)

.github/workflows/canary-integration-test.yml

@@ -167,8 +167,7 @@ jobs:
           timeout 15 sh -c "until kubectl -n rook-ceph exec $toolbox -- python3 /etc/ceph/create-external-cluster-resources.py --rbd-data-pool-name replicapool --rgw-endpoint $rgw_endpoint; do sleep 1 && echo 'waiting for the rgw endpoint to be validated'; done"
           # pass the invalid rgw-endpoint of different ceph cluster
           if output=$(timeout 15 sh -c "until kubectl -n rook-ceph exec $toolbox -- python3 /etc/ceph/create-external-cluster-resources.py --rbd-data-pool-name replicapool --rgw-endpoint 10.108.96.128:80; do sleep 1 && echo 'waiting for the rgw endpoint to be validated'; done"); then
-            echo "unexpectedly succeeded after passing the wrong rgw-endpoint: $output"
-            exit 1
+            echo "script run completed with stderr error after passing the wrong rgw-endpoint: $output"
           else
             echo "validation failed because wrong endpoint was provided"
           fi