When using Ansible 2.8.1 with SLES 12 SP5, the nmcli module appears to have a
bug:

Example:

TASK [nic_nmcli : Set NICs configuration]
******************************************************************************
Saturday 15 January 2022  11:53:45 +0000 (0:00:00.505)       0:00:06.085 ******
failed: [wolf20] (item={u'prefix': 16, u'network': u'ice1-1', u'ip4':
u'10.11.0.20', u'interface': u'eth1', u'ifname': u'eth1', u'type': u'ethernet',
u'mac': u'52:54:00:aa:aa:aa'}) => changed=false
  ansible_loop_var: item
  item:
    ifname: eth1
    interface: eth1
    ip4: 10.11.0.20
    mac: 52:54:00:aa:aa:aa
    network: ice1-1
    prefix: 16
    type: ethernet
  msg: |-
    Error: invalid <setting>.<property> 'autoconnect'.
  name: eth1
  rc: 2

This seems to be have been fixed by the Ansible maintainers via:
https://github.com/ansible-collections/community.general/pull/478/files

Using the version of nmcli.py from this pull request appears to resolve the
issue, e.g.

cp /usr/lib/python2.7/site-packages/ansible/modules/net_tools/nmcli.py
/usr/lib/python2.7/site-packages/ansible/modules/net_tools/nmcli.py.orig
wget
https://raw.githubusercontent.com/Akasurde/community.general/8c3301dc3657bbd9994009e5ce2215f22f4a404b/plugins/modules/net_tools/nmcli.py
mv nmcli.py /usr/lib/python2.7/site-packages/ansible/modules/net_tools/nmcli.py