Bug 1903777

Description of problem:
After upgrading from RHV-H 4.4.2 to RHV-H 4.4.3, chronyd is disabled.

Version-Release number of selected component (if applicable):
4.4.3

How reproducible:
100%

Steps to Reproduce:
1. Update RHV-H 4.4.2 -> 4.4.3 
2. systemctl status chronyd

Actual results:
chronyd.service is not running and disabled.

Expected results:
chronyd.service should be running, enabled

Additional info:
This caused issues in a customer environment, as even after enabling / starting chronyd, vdsmd was still confused and caused VM's to be non-responsive even though they were running fine.

Restarting vdsmd on the affected node fixes the VM state, but this would cause serious issues when performing the update/reboot as part of a clusterwide update.


- For some reason we remove /etc/systemd/system/multi-user.target.wants/chronyd.service below is from imgbased.log 

~~~
2020-11-16 18:15:19,936 [DEBUG] (migrate_ntp_to_chrony) Calling: (['systemctl', 'status', 'ntpd.service'],) {'close_fds': True, 'stderr': -2}
2020-11-16 18:15:19,955 [DEBUG] (migrate_ntp_to_chrony) Exception! b'Unit ntpd.service could not be found.\n'
2020-11-16 18:15:19,955 [DEBUG] (migrate_ntp_to_chrony) ntpd is disabled, not migrating conf to chrony
2020-12-02 18:12:04,751 [DEBUG] (remediate_etc) Planning to remove /etc/systemd/system/multi-user.target.wants/chronyd.service
2020-12-02 18:12:05,033 [DEBUG] (migrate_var) Migrating files by rpm databases: ['/var/lock', '/var/cache/ldconfig/aux-cache', '/var/cache/dnf/expired_repos.json', '/var/cache/dnf/anaconda-filenames.s
2020-12-02 18:12:05,046 [DEBUG] (migrate_var) Calling: (['rpm', '-qf', '/var/lock', '/var/cache/ldconfig/aux-cache', '/var/cache/dnf/expired_repos.json', '/var/cache/dnf/anaconda-filenames.solvx', '/v
2020-12-02 18:12:10,440 [DEBUG] (remediate_etc) os.unlink(/tmp/mnt.J7x6l////etc/systemd/system/multi-user.target.wants/chronyd.service)
2020-12-02 18:12:25,133 [DEBUG] (migrate_var) Updated file /var/lib/selinux/targeted/active/modules/100/chronyd/cil
2020-12-02 18:12:25,135 [DEBUG] (migrate_var) Updated file /var/lib/selinux/targeted/active/modules/100/chronyd/hll
2020-12-02 18:12:26,999 [DEBUG] (migrate_etc) chrony changed from 995 to 996
2020-12-02 18:12:27,000 [DEBUG] (migrate_etc) chrony changed from 992 to 993
2020-12-02 18:12:27,002 [DEBUG] (migrate_etc) chrony changed from 995 to 996
2020-12-02 18:12:27,003 [DEBUG] (migrate_etc) chrony changed from 992 to 993
~~~

- While on 4.4.2, the service is active/running/enabled:
~~~
# systemctl status chronyd
● chronyd.service - NTP client/server
   Loaded: loaded (/usr/lib/systemd/system/chronyd.service; enabled; vendor preset: disabled)
   Active: active (running) since Mon 2020-11-16 18:38:32 UTC; 2 weeks 1 days ago
     Docs: man:chronyd(8)
           man:chrony.conf(5)
 Main PID: 1311 (chronyd)
    Tasks: 1 (limit: 204376)
   Memory: 2.7M
   CGroup: /system.slice/chronyd.service
           └─1311 /usr/sbin/chronyd

Warning: Journal has been rotated since unit was started. Log output is incomplete or unavailable.
~~~


The file removed is a symlink:

# ll /etc/systemd/system/multi-user.target.wants/chronyd.service
lrwxrwxrwx. 1 root root 39 Sep 29 21:52 /etc/systemd/system/multi-user.target.wants/chronyd.service -> /usr/lib/systemd/system/chronyd.service

# ll /usr/lib/systemd/system/chronyd.service
-rw-r--r--. 1 root root 491 May 22  2019 /usr/lib/systemd/system/chronyd.service



==============================

After updating to 4.4.3, reboot


# systemctl status chronyd
● chronyd.service - NTP client/server
   Loaded: loaded (/usr/lib/systemd/system/chronyd.service; disabled; vendor preset: disabled)
   Active: inactive (dead)
     Docs: man:chronyd(8)
           man:chrony.conf(5)

# ll /etc/systemd/system/multi-user.target.wants/chronyd.service
ls: cannot access '/etc/systemd/system/multi-user.target.wants/chronyd.service': No such file or directory