if machine has 'Inst id' with capitals, deploy fails

[Hans van den Bogert]

I looked into this after seeing that the calico charm does not successfully deploy and gets stuck in state "retry calico configuration"

Further investigation showed:

```
unit-calico-23: 16:47:40 INFO unit.calico/23.juju-log etcd:36: Traceback (most recent call last):
  File "/var/lib/juju/agents/unit-calico-23/charm/reactive/calico.py", line 465, in configure_node
    node = calicoctl_get('node', node_name)
  File "/var/lib/juju/agents/unit-calico-23/charm/reactive/calico.py", line 633, in calicoctl_get
    output = calicoctl(*args)
  File "/var/lib/juju/agents/unit-calico-23/charm/reactive/calico.py", line 582, in calicoctl
subprocess.CalledProcessError: Command '['/opt/calicoctl/calicoctl', 'get', '-o', 'yaml', '--export', 'node', 'NLDW1-6-26-1']' returned non-zero exit status 1

```

Manually adding the node shows the supposed faulty scenario; Adding a node on calico level fails if it has capitals.

```
./calicoctl create -f me.yaml
Failed to create 'Node' resource: [error with field Metadata.Name = 'NLDW1-6-26-1' (name must consist of lower case alphanumeric characters, '-' or '.' (regex: [a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*))]
```

Changing the node's name in MAAS to lower case solved this for me.

[George Kraft (cynerva)]

Thanks for the report. Looks like the calico charm uses gethostname() in a few places to determine the node name, where it should be using gethostname().lower().

[Peter Jose De Sousa (pjds)]

hit this one today, calico adds the node as lower, but upper will be used to fetch the node with calicoctl, failing

[Ebrar Leblebici (birru2)]

I also hit this one with calico charm from 1.28/stable channel and rev 101.