fix(nemesis): disrupt_remove_node_then_add_node was missing retry=0

in ecc2c7e `long_running=True` paramter
was introduced, but it must come togther with `retry=0`, and that
was forgotten

Fixes: #9534
(cherry picked from commit e07010b)

sdcm/nemesis.py

@@ -3556,7 +3556,7 @@ def remove_node():
                           .format(rnd_node.ip_address, host_id))
             with adaptive_timeout(Operations.REMOVE_NODE, rnd_node, timeout=HOUR_IN_SEC * 48):
                 res = rnd_node.run_nodetool("removenode {}".format(
-                    host_id), ignore_status=True, verbose=True, long_running=True)
+                    host_id), ignore_status=True, verbose=True, long_running=True, retry=0)
             if res.failed and re.match(removenode_reject_msg, res.stdout + res.stderr):
                 raise Exception(f"Removenode was rejected {res.stdout}\n{res.stderr}")