public-address is wrong

[Aaron Bentley]

In my current "juju status", the "public address" is actually a private address, and the public address is listed as dns-name.
This is with juju 1.13.3-raring-amd64

I achieved this configuration by
1. configuring use-floating-ip: False
2. deploying jenkins
3. using "nova" to assign a floating IP to the jenkins instance.
I believe this will be a common pattern in clouds where floating IP addresses are scarce, such as private clouds.

Look at machine 4 and jenkins/0 below:
$ juju status
environment: canonistack
machines:
  "0":
    agent-state: started
    agent-version: 1.13.3
    dns-name: 10.55.32.53
    instance-id: 20e41b17-f5e2-40ea-bb2a-cb554a93cff3
    instance-state: ACTIVE
    series: precise
    hardware: arch=amd64 cpu-cores=1 mem=512M
  "4":
    agent-state: started
    agent-version: 1.13.3
    dns-name: 162.213.35.31
    instance-id: cf2e8ec5-dd15-41cd-b616-25a76595d9a6
    instance-state: ACTIVE
    series: precise
    hardware: arch=amd64 cpu-cores=1 mem=2048M root-disk=10240M
services:
  jenkins:
    charm: local:precise/jenkins-31
    exposed: true
    units:
      jenkins/0:
        agent-state: started
        agent-version: 1.13.3
        machine: "4"
        open-ports:
        - 8080/tcp
        public-address: 10.55.32.104

[José Antonio Rey (jose)]

This bug is still affecting. See my status:

joseeantonior@dragonborn:~$ juju status
environment: amazon
machines:
  "0":
    agent-state: started
    agent-version: 1.16.6
    dns-name: ec2-54-85-147-201.compute-1.amazonaws.com
    instance-id: i-bca5439d
    instance-state: running
    series: precise
    hardware: arch=amd64 cpu-cores=1 cpu-power=20 mem=613M root-disk=8192M
  "1":
    agent-state: started
    agent-version: 1.16.6
    dns-name: ec2-54-85-138-22.compute-1.amazonaws.com
    instance-id: i-ddae48fc
    instance-state: running
    series: precise
    hardware: arch=amd64 cpu-cores=1 cpu-power=20 mem=613M root-disk=8192M
  "2":
    agent-state: started
    agent-version: 1.16.6
    dns-name: ec2-54-84-166-48.compute-1.amazonaws.com
    instance-id: i-a4af4985
    instance-state: running
    series: precise
    hardware: arch=amd64 cpu-cores=1 cpu-power=20 mem=613M root-disk=8192M
services:
  mysql:
    charm: cs:precise/mysql-38
    exposed: false
    relations:
      cluster:
      - mysql
    units:
      mysql/0:
        agent-state: started
        agent-version: 1.16.6
        machine: "1"
        public-address: 172.31.18.7
  phpbb:
    charm: local:precise/phpbb-1
    exposed: true
    units:
      phpbb/0:
        agent-state: started
        agent-version: 1.16.6
        machine: "2"
        open-ports:
        - 80/tcp
        public-address: 172.31.18.223

[Junaid Ali (junaidali)]

This bug is still here. I'm facing this issue in one of my deployments.

juju version:
1.25.5

Output of
$ juju stat ubuntu
is http://paste.ubuntu.com/17701137/

[Cheryl Jennings (cherylj)]

Junaid - if you still have the environment up, can you attach the state server log? /var/log/juju/machine-0.log

If this is something you can recreate easily, can you set the log level to DEBUG and attach the above log after a recreate? You can set the log level with `juju set-env logging-config="<root>=DEBUG"`.

[Junaid Ali (junaidali)]

Hi Cheryl, sorry for the delayed response. I just got back that setup again. After setting the logging-config as per your suggestion, tested deploying machines (with FQDN test.maas and compute3.maas) but the issue is still there. Instead of FQDN of machines, juju is setting the node's IP as public address.
Attached are the logs of state server.