Check manual page of timesyncd

Systemd Timesyncd

Included in	All Checkmk editions
Source Code License	Open Source
Supported Agents	Linux

This check uses the output of timedatectl timesync-status as sent by the agent. It lists the information of the synchronised time server including the server address, stratum, offset and jitter. The check gets critical if the time offset exceeds a default level, if the stratum is too high or other error conditions. The stratum levels default to 10 (warning) / 16 (critical). Critical is set to 16 because that is the special NTP value for an unsynchronised clock and thus an actual error, whereas any stratum below it (1 to 15) is a valid, synchronised state. The warning at 10 flags an unusually deep chain to the reference clock: such hops are not an error in themselves, but each one adds latency and reduces accuracy, so a chain that long is bad practice and worth a heads-up. The jitter (the short-term variation of the time synchronisation) is reported as a metric but has no thresholds by default, as it is a stability signal rather than a health metric. Optional jitter levels, independent of the time offset levels, can be configured via the ruleset. It gets UNKNOWN if the time has not yet been synchronised. The check logic is shared with the ntp check. On a running instance the time is synchronised either using ntp or timesyncd. The agent therefore determines the correct service to use and subsequent check.

Note: this check is only available for versions higher than Ubuntu 18.04

Discovery

One service is created if the timedatectl timesync-status command is available.