chore: Restart Network Manager on OCP

[e0ne]

We need to restart NetworkManager on OpenShift only to ensure that the network configuration is applied correctly included bonding interfaces.

[schmaustech]

Okay so I tried to first create a bond0 environment and I used the following nmstate yaml

apiVersion: nmstate.io/v1
kind: NodeNetworkConfigurationPolicy
metadata:
name: bond0-enp220s0np0-enp206s0np0-policy
spec:
desiredState:
interfaces:
- name: bond0
type: bond
state: up
ipv4:
address:
- ip: 192.168.10.205
prefix-length: 24
dhcp: false
enabled: true
link-aggregation:
mode: 802.3ad
options:
miimon: '140'
port:
- enp220s0np0
- enp206s0np0

This setup a bond0 on the two Mellanox interfaces.

Then I first tried to deploy doca-driver 25.01-0.6.0.0-0 without the fix. Driver deployed fine and the bond0 interface was still up and functional after driver deployment. This was not the behavior we had seen before - however the bond0 interface had br-ex on top of it and was used for primary cluster traffic. So reproducer is not quite the same.

I did try the new patched image 25.04-0.6.1.0-200 however when it went to restart NetworkManager it was unable to from the container - which leads me to think maybe in the past troubleshooting I did on this I did the restart by being on the node via a console since networking was down .

Error when trying to restart NetworkManager:

• timestamp_print Executing command: systemctl restart NetworkManager
  ++ date +%d-%b-%y_%H:%M:%S
• date_time_stamp=16-May-25_15:35:03
• msg='[16-May-25_15:35:03] Executing command: systemctl restart NetworkManager'
• echo '[16-May-25_15:35:03] Executing command: systemctl restart NetworkManager'
• tee -a /tmp/entrypoint_debug_cmds.log
  [16-May-25_15:35:03] Executing command: systemctl restart NetworkManager
  ++ eval 'systemctl restart NetworkManager'
  +++ systemctl restart NetworkManager
  System has not been booted with systemd as init system (PID 1). Can't operate.
  Failed to connect to bus: Host is down
• output=
• exit_code=1
• echo ''