[backport 17.06] Handle cleanup DNS for attachable container

[fcrisciani]

Attachable containers they are tasks with no service associated
their cleanup was not done properly so it was possible to have
a leak of their name resolution if that was the last container
on the network.
Cleanupservicebindings was not able to do the cleanup because there
is no service, while also the notification of the delete arrives
after that the network is already being cleaned

Signed-off-by: Flavio Crisciani flavio.crisciani@docker.com
(cherry picked from commit 1c04e19)

[mavenugo]

LGTM

[sampipe]

Hi,

I've just updated to 17.06.2-ee-5 from 17.06.2-ee-2

We run a swarm cluster.

Specific use case we need to route some requests direct to a container rather than via the ingress mesh network and therefore we utilised the DNS query of tasks.[service_name] to acquire a list of IP addresses of the running containers.

On 17.06.2-ee-2 this DNS query returned correctly with the list of tasks (container) running across the swarm cluster on all swarm nodes:

e.g.

;; ANSWER SECTION:
tasks.service_name. 600 IN A 10.0.0.13
tasks.service_name. 600 IN A 10.0.0.14

Now on 17.06.2-ee-5 I am seeing that the DNS is only returning the IP address of the task (container) running on the server that node..

e.g.

DNS query on Server 1 (swarm node 1)

;; ANSWER SECTION:
tasks.service_name. 600 IN A 10.0.0.13

DNS query on Server 2 (swarm node 2)

;; ANSWER SECTION:
tasks.service_name. 600 IN A 10.0.0.14

Could this have been broken in the latest update?

Sam

[sampipe]

Ok.. ignore the above.. seems that the DNS got itself into a odd state during the update process.. I updated each server one at a time to maintain service.. for some reason this must have impacted the DNS entries..

I did a full restart of docker on both servers and the tasks.service_name DNS is returning correctly now.

Sam