Deploy app on a cluster but cannot access it successfully - docker-swarm

Question

I'm now learning to use docker follow get-started documents, but in part 4--Swarms I've met some problem. That is when deployed my app on a cluster, I cannot access it successfully.
docker#myvm1:~$ docker stack ps getstartedlab
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
gsueb9ejeur5 getstartedlab_web.1 zhugw/get-started:first myvm1 Running Preparing 11 seconds ago
ku13wfrjp9wt getstartedlab_web.2 zhugw/get-started:first myvm2 Running Preparing 11 seconds ago
vzof1ybvavj3 getstartedlab_web.3 zhugw/get-started:first myvm1 Running Preparing 11 seconds ago
lkr6rqtqbe6n getstartedlab_web.4 zhugw/get-started:first myvm2 Running Preparing 11 seconds ago
cpg91o8lmslo getstartedlab_web.5 zhugw/get-started:first myvm2 Running Preparing 11 seconds ago
docker#myvm1:~$ curl 'http://localhost'
curl: (7) Failed to connect to localhost port 80: Connection refused
➜ ~ docker-machine ls
NAME ACTIVE DRIVER STATE URL SWARM DOCKER ERRORS
myvm1 - virtualbox Running tcp://192.168.99.101:2376 v17.06.0-ce
myvm2 - virtualbox Running tcp://192.168.99.100:2376 v17.06.0-ce
➜ ~ curl 'http://192.168.99.101'
curl: (7) Failed to connect to 192.168.99.101 port 80: Connection refused
What's wrong?
In addition, very strange. After adding below content in docker-compose.yml I found above question resolved automatically
visualizer:
image: dockersamples/visualizer:stable
ports:
- "8080:8080"
volumes:
- "/var/run/docker.sock:/var/run/docker.sock"
deploy:
placement:
constraints: [node.role == manager]
networks:
- webnet
but this time the new added visualizer does not work
docker#myvm1:~$ docker stack ps getstartedlab
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
xomsv2l5nc8x getstartedlab_web.1 zhugw/get-started:first myvm1 Running Running 7 minutes ago
ncp0rljod4rc getstartedlab_visualizer.1 dockersamples/visualizer:stable myvm1 Running Preparing 7 minutes ago
hxddan48i1dt getstartedlab_web.2 zhugw/get-started:first myvm2 Running Running 7 minutes ago
dzsianc8h7oz getstartedlab_web.3 zhugw/get-started:first myvm1 Running Running 7 minutes ago
zpb6dc79anlz getstartedlab_web.4 zhugw/get-started:first myvm2 Running Running 7 minutes ago
pg96ix9hbbfs getstartedlab_web.5 zhugw/get-started:first myvm2 Running Running 7 minutes ago
from above you know it's always preparing.
My whole docker-compose.yml
version: "3"
services:
web:
# replace username/repo:tag with your name and image details
image: zhugw/get-started:first
deploy:
replicas: 5
resources:
limits:
cpus: "0.1"
memory: 50M
restart_policy:
condition: on-failure
ports:
- "80:80"
networks:
- webnet
visualizer:
image: dockersamples/visualizer:stable
ports:
- "8080:8080"
volumes:
- "/var/run/docker.sock:/var/run/docker.sock"
deploy:
placement:
constraints: [node.role == manager]
networks:
- webnet
networks:
webnet:

Answer 1

Had this problem while learning too.
It's because your none clustered image is still running from step 2 and the clustered image you just deployed uses the same port mapping (4000:80) in the docker-compose.yml file.
You have two options:
Go into your docker-compose.yml and change the port mapping to something else e.g 4010:80 and then redeploy your cluster with the update. Then try: http://localhost:4010
Remove the container you created in step 2 of the guide that's still running and using port mapping 4000:80

Answer 2

volumes:
- "/var/run/docker.sock:/var/run/docker.sock"
should be
volumes:
- /var/run/docker.sock:/var/run/docker.sock
this is an error in the dockers tutors

Answer 3

Open port 7946 TCP/UDP and port 4789 UDP between the swarm nodes. Use the ingress network. Please let me know if it works, thanks.

Answer 4

What helped for me to get the visualizer running was changing visualizer image tag from stable to latest.

Answer 5

If you are using Docker toolbox for mac, then you should check this out.

Answer 6

I had the same problem. As it says in the tutorial (see "Having connectivity trouble?") the following ports need to be open:
Port 7946 TCP/UDP for container network discovery.
Port 4789 UDP for the container ingress network.
So I executed the following before the swarm init (right after creation of myvm1 and myvm2) and then later could access the service e.g. in the browser with IP_node:4000
$ docker-machine ssh myvm1 "sudo iptables -I INPUT -p tcp --dport 7946 --syn -j ACCEPT"
$ docker-machine ssh myvm2 "sudo iptables -I INPUT -p tcp --dport 7946 --syn -j ACCEPT"
$ docker-machine ssh myvm1 "sudo iptables -I INPUT -p udp --dport 7946 -j ACCEPT"
$ docker-machine ssh myvm2 "sudo iptables -I INPUT -p udp --dport 7946 -j ACCEPT"
$ docker-machine ssh myvm1 "sudo iptables -I INPUT -p udp --dport 4789 -j ACCEPT"
$ docker-machine ssh myvm2 "sudo iptables -I INPUT -p udp --dport 4789 -j ACCEPT"
Hope it helps others.