fix(local run): check if manager agent is up

The changes in commit https://github.com/scylladb/scylla-cluster-tests/commit/
aea611d breaks the opportunity to run tests from local
developer PC.

The setup fails on wait_manager_agent_up function as tries to access nodes via private ip address.
Use curl from DB node when check if manager agent is up instead of send request from runner
(local machine).

Fixes: #7636

sdcm/cluster.py

@@ -1360,8 +1360,12 @@ def is_manager_agent_up(self, port=None):
         # When the agent is IP, it should answer an https request of https://NODE_IP:10001/ping with status code 204
         # Scylla Manager Agent uses Scylla listen/broadcast address - https://manager.docs.scylladb.com/stable/config/
         # scylla-manager-agent-config.html#:~:text=value.%0A%23auth_token%3A-,%23%20Bind%20REST%20API%20to,-the%20specified%20TCP
-        response = requests.get(f"https://{normalize_ipv6_url(self.scylla_listen_address)}:{port}/ping", verify=False)
-        return response.status_code == 204
+        http_address = f"https://{normalize_ipv6_url(self.scylla_listen_address)}:{port}/ping"
+        curl_output = self.remoter.run(cmd=f'''curl -k --write-out "%{{http_code}}\n" --silent --output /dev/null "{http_address}"''',
+                                       verbose=True, ignore_status=True)
+        LOGGER.debug("curl with scylla_listen_address stdout: %s", curl_output.stdout)
+        http_status_code = int(curl_output.stdout.strip())
+        return http_status_code == 204
 
     def wait_manager_agent_up(self, verbose=True, timeout=180):
         text = None