clean up

Author: YuryHrytsuk

services/rabbit/Makefile

@@ -81,14 +81,11 @@ TIMEOUT = 2m
 delete-volumes: ${CLEAN_VOLUMES_TEMP_COMPOSE} guard-optional-TIMEOUT guard-optional-bool-FORCE_CONFIRM
 delete-volumes:  ## Delete rabbit data volumes
 	@$(call confirm_action,Forever delete all rabbit data?,deleteforever,${FORCE_CONFIRM})
-
 	## avoid accumulating tasks https://github.com/moby/moby/issues/45443
 	@docker stack rm --detach=false ${CLEAN_VOLUMES_STACK_NAME}
-
 	## use --detach=false to wait until all jobs complete successfully
 	@timeout ${TIMEOUT}	\
 	docker stack deploy --detach=false --with-registry-auth --prune --compose-file ${CLEAN_VOLUMES_TEMP_COMPOSE} ${CLEAN_VOLUMES_STACK_NAME}
-
 	## cleaning up
 	@docker stack rm ${CLEAN_VOLUMES_STACK_NAME}
 

services/rabbit/README.md

@@ -27,9 +27,3 @@ Not supported at the moment.
 2. (Inside container) execute `rabbitmq-upgrade drain`
 
 Source: https://www.rabbitmq.com/docs/upgrade#maintenance-mode
-
-## Rotating erlang cookie (zero-downtime)
-
-In our case, a full stop of rabbit cluster is required (it means downtime). Nodes shall be started from scratch with a new erlang cookie. If zero-downtime update is an absolute must, blue-green deployments is a way to go according to core developers (see sources)
-
-Source: https://github.com/rabbitmq/rabbitmq-server/discussions/14391

services/rabbit/docker-compose.yml.j2

@@ -149,5 +149,6 @@ configs:
 secrets:
   rabbit_erlang_cookie:
     # no rolling update since it requires full cluster restart
+    # see https://github.com/rabbitmq/rabbitmq-server/discussions/14391
     name: rabbit_erlang_cookie
     file: ./configs/erlang.cookie.secret