Role fails on Fedora 43 bootc (server) due to lack of "runlevel" command

[benblasco]

Problem:

Role fails at the "Determine current NTP provider" stage:

TASK [fedora.linux_system_roles.timesync : Determine current NTP provider] ***************************************************************
fatal: [hex.lan]: FAILED! => {"changed": false, "module_stderr": "Shared connection to hex.lan closed.\r\n", "module_stdout": "/var/home/bblasco/.ansible/tmp/ansible-tmp-1769734063.0693882-189237-16933141090487/AnsiballZ_timesync_provider.sh: line 9: runlevel: command not found\r\n{\"ansible_facts\": {\"timesync_ntp_provider_current\": \"chrony\"}}\u001b]3008;end=99235c20f92644d3ad8d663269ec8595\u001b\\", "msg": "MODULE FAILURE: No end of json char found\nSee stdout/stderr for the exact error", "rc": 0}

The verbose output is attached as error.txt

error.txt

Cause:

From the above we can see the error:

runlevel: command not found

If I check the Fedora 43 bootc system manually:

[bblasco@hex ~]$ cat /etc/fedora-release 
Fedora release 43 (Forty Three)
[bblasco@hex ~]$ runlevel
-bash: runlevel: command not found

But if I run it on a Fedora 42 bootc system:

[bblasco@micro ~]$ cat /etc/fedora-release 
Fedora release 42 (Adams)
[bblasco@micro ~]$ runlevel
N 3

Code snippet with the problem

I found the code in the file timesync_provider.sh, specifically this snippet:

is_service_enabled() {
        local name runlevel prev_runlevel
        name="$1"

        # shellcheck disable=SC2034
        read -r prev_runlevel runlevel < <(runlevel)

        systemctl is-enabled "$name.service" &> /dev/null || \
                chkconfig --list "$name" 2>/dev/null | grep -q "$runlevel:on"
}

If I understand the code correctly the runlevel code only matters if the system doesn't have systemctl (ie uses systemd), so it's actually failing at the following line:

        read -r prev_runlevel runlevel < <(runlevel)

The output of that line is then used for the grep command in the following line:

        systemctl is-enabled "$name.service" &> /dev/null || \
                chkconfig --list "$name" 2>/dev/null | grep -q "$runlevel:on"

How to fix

Good question!

I tried removing anything relating to runlevel and just populating the vars manually for the purposes of my own testing:

is_service_enabled() {
        local name runlevel prev_runlevel
        name="$1"
        prev_runlevel="N"
        runlevel="3"

        # shellcheck disable=SC2034
        # read -r prev_runlevel runlevel < <(runlevel)

        systemctl is-enabled "$name.service" &> /dev/null # || \
                # chkconfig --list "$name" 2>/dev/null | grep -q "$runlevel:on"
}

But then I get this error (which I don't really understand) instead:

TASK [fedora.linux_system_roles.timesync : Determine current NTP provider] ***************************************************************
fatal: [hex.lan]: FAILED! => {"changed": false, "module_stderr": "Shared connection to hex.lan closed.\r\n", "module_stdout": "{\"ansible_facts\": {\"timesync_ntp_provider_current\": \"chrony\"}}\u001b]3008;end=58517a9662d14e8e97b8aa5b3b4060a6\u001b\\", "msg": "MODULE FAILURE: No end of json char found\nSee stdout/stderr for the exact error", "rc": 0}

Running the playbook in verbose mode I get the output in the attached troubleshoot.txt

troubleshoot.txt

Software versions

Control node:

fedoralaptop on  master [!?] via 🐍 v3.14.2 
❯ ansible --version
ansible [core 2.18.12]
  config file = /var/home/bblasco/git/fedoralaptop/ansible.cfg
  configured module search path = ['/var/home/bblasco/.ansible/plugins/modules', '/usr/share/ansible/plugins/modules']
  ansible python module location = /usr/lib/python3.14/site-packages/ansible
  ansible collection location = /var/home/bblasco/.ansible/collections:/usr/share/ansible/collections
  executable location = /usr/bin/ansible
  python version = 3.14.2 (main, Dec  5 2025, 00:00:00) [GCC 15.2.1 20251111 (Red Hat 15.2.1-4)] (/usr/bin/python3)
  jinja version = 3.1.6
  libyaml = True

fedoralaptop on  master [!?] via 🐍 v3.14.2 
❯ cat /etc/fedora-release 
Fedora release 43 (Forty Three)

fedoralaptop on  master [!?] via 🐍 v3.14.2 
❯ python --version
Python 3.14.2

fedoralaptop on  master [!?] via 🐍 v3.14.2 
❯ 

Node being automated

[bblasco@hex ~]$ cat /etc/fedora-release 
Fedora release 43 (Forty Three)
[bblasco@hex ~]$ python --version
Python 3.14.2

Please let me know what other info you need

[richm]

@mlichvar is this a bug in fedora 43 ?

[benblasco]

My work-around has simply been to comment the whole task out in timesync/main.yml:

#- name: Determine current NTP provider
  #fedora.linux_system_roles.timesync_provider:
  #when:
    #- timesync_mode != 2
    #- timesync_ntp_provider | length == 0

[benblasco]

I forgot to add that the role version I am using is:
fedora.linux_system_roles 1.119.0

[mlichvar]

It seems runlevel was removed in systemd v258. From their changelog:

 * Support for System V style system state control has been removed:
      - The /dev/initctl device node has been removed.
      - The initctl, runlevel, and telinit commands have been removed.
      - Support for system state control via the init command (e.g.
        'init 3') has been removed.
      - The units runlevel[0-6].target have been removed.
      - The concept of runlevels has been removed, so runlevel transitions
        are no longer recorded in the utmp/wtmp databases.