Check manual page of ntp.time

ntp.time

Author: Mathias Kettner

License: GPL

Distribution: official part of Check_MK

Supported Agents: Linux

This check uses the output of ntpq -p as sent by the agent in order to check the quality of the NTP time synchronization of the client. If more than one peer is available, NTP chooses the "best" of them as "system peer". This check only measures the time difference to that system peer.

The check is CRIT or WARN, if the time supplied by the system peer is not good enough (see below for parameters). It is UNKNOWN if no system peer is present or the system peer is unreachable.

The check is CRIT, when there are NTP peers reachable but non of those is used by the NTP daemon. That is the case if the time provided by those peers is too bad.

Note: If you want to check the health of the peers, you might want to use ntp, which creates one individual check per NTP peer.

Check parameters

Parameter	Type	Description
critical_stratum	int	The stratum at which the check gets critical
warning	float	The offset in ms at which a warning is triggered
critical	float	the critical offset in ms

Performance data

Two variables: the time offset to the peer in milliseconds and the jitter in ms.

Inventory

If at least one peer is present, exactly one check is created, if ntp_inventory_mode is set to "summary" or "both".

Configuration variables

Variable	Type	Description
ntp_default_levels	int, float, float	The global default levels for all checks of this type. This is preset to (10, 20.0, 50.0) (which means that the check gets critical if the stratum is 10 or higher).
ntp_inventory_mode	string	This variable controls which kind of NTP checks are created. A check of this type (ntp.time) is only created, if this variable is set to "summary" or "both". Default is "summary".

Examples

# Set critical stratum to 3, global warning/critical levels to 5.0 and 10.0 ms
ntp_default_levels = (3, 5.0, 10.0)