Skip to main content
Released

[os::linux::local::plugin] Adding support for timedatectl to "ntp" mode

Related products:Infra Monitoring - Connectors
  • Nov 21, 2023 9:47 AM UTC
  • 6 replies
  • 196 views
Aleksey
Forum|alt.badge.img+6

Hello,

Many modern Linux systems that are part of systemd have the systemd-timesyncd service.
If this service is used, then there are no such things as ntpd and chronyd in the system.

Please extend the "ntp" mode to work with the timedatectl utility.

# timedatectl status
               Local time: Fri 2022-10-07 18:33:28 MSK
           Universal time: Fri 2022-10-07 15:33:28 UTC
                 RTC time: Fri 2022-10-07 15:33:28
                Time zone: Europe/Moscow (MSK, +0300)
System clock synchronized: yes
              NTP service: active
          RTC in local TZ: no
 
# timedatectl timesync-status
       Server: 10.10.0.2 (srv-ntp.holding.com)
Poll interval: 34min 8s (min: 32s; max 34min 8s)
         Leap: normal
      Version: 3
      Stratum: 3
    Reference: A05A409
    Precision: 1us (-23)
Root distance: 128.654ms (max: 5s)
       Offset: +6.303ms
        Delay: 1.544ms
       Jitter: 16.867ms
 Packet count: 17
    Frequency: +1.635ppm
 

Thank you.

6 replies

rchauvel
Centreonian
Forum|alt.badge.img+18
  • rchauvel
  • Centreonian
  • Dec 1, 2023 1:26 PM UTC
New→Needs Votes
Centreonian
Forum|alt.badge.img+5
  • ProductGuru
  • Centreonian
  • Jun 12, 2024 3:59 PM UTC
Needs Votes→In Backlog
Centreonian
Forum|alt.badge.img+5
  • ProductGuru
  • Centreonian
  • Sep 5, 2025 8:47 AM UTC
In Backlog→Planned
Centreonian
Forum|alt.badge.img+5
  • ProductGuru
  • Centreonian
  • Sep 19, 2025 7:29 AM UTC
Planned→Released
ltirand
Centreonian
Forum|alt.badge.img+13
  • ltirand
  • Centreonian
  • Sep 19, 2025 7:30 AM UTC

Hello :)

This idea is released with the September 2025 release.

Kind regards

Alone we go faster, together we go further.
    Aleksey
    Forum|alt.badge.img+6
    • Aleksey
    • Author
    • Builder *
    • Jul 14, 2026 9:59 AM UTC

    Thanks for the plugin, but there is a problem with it.


    The check does not detect a problem if there is not a single peer with reach > 0 among the listed peers.
    In fact, when we have peers with reach = 0, it means that synchronization does not work.

    In this example, we're testing a system with only one time source. 
    The plugin tells us everything is fine:

    # ./centreon_plugins.pl --plugin=os::linux::local::plugin --mode=ntp 
    --critical-offset '-180000:180000' 
    --critical-peers '1:' 
    --exclude-state '.*(not used).*' 
    --filter-perfdata-adv 'not (%(label) =~ /\.(detected.count)/)' 
    --filter-uom 'null' 
    --warning-offset '-60000:60000' 
    --warning-stratum '0:6'

    OK: Number of ntp peers: 1 - Peer 'ntp.sub.holding.com' [type: Primary NTP Server] [reach: 0] [state: configured and available for use] | 
    'ntp.sub.holding.com#peer.time.offset.milliseconds'=0;;;0; 
    'ntp.sub.holding.com#peer.stratum.count'=0;0:6;;0;

    But in reality, we understand that synchronization with this single time source does not work:

    # timedatectl timesync-status

           Server: n/a (ntp.sub.holding.com)
    Poll interval: 0 (min: 32s; max 34min 8s)
     Packet count: 0

     

    Terms & ConditionsAccessibility statement