Add documentation on IdP restart

Add a note to the documentation that the IdP
needs to be restarted after increasing the number
of nodes.

README.md

@@ -141,6 +141,13 @@ To provision the infrastructure execute the playbook:
 ansible-playbook comanage_registry_training.yml
 ```
 
+**Note: After increasing the number of training nodes, you must restart
+the IdP service by doing**
+
+```
+docker service update --force idp_shibboleth-idp
+```
+
 To reconfigure only the training nodes once they have already
 been provisioned:
 
@@ -225,6 +232,13 @@ To start the stack log into the IdP node and execute
 docker stack deploy --compose-file /opt/shibboleth-idp-stack.yml idp
 ```
 
+**Note: After increasing the number of training nodes, you must restart
+the IdP service by doing**
+
+```
+docker service update --force idp_shibboleth-idp
+```
+
 Useful Docker Swarm commands for the IdP node are
 
 ```