check_ntp_peer result message is unclear about reason for failure (#1236)

Wil Cooley notifications at
Tue Feb 11 21:39:55 CET 2014

`check_ntp_peer` has 4 different thresholds that it checks but one cannot
tell from the result message which of the thresholds has been exceeded. This
makes it a lot harder to assess where the problem lies or how quickly the
alert needs to be responded to, particularly at 04:00 on your cell phone.

I would like the result message to make it clear which of the 4 thresholds
have been passed, either by somehow emphasizing the ones in error or only
including those in error.  (Values for the other thresholds are already
captured by  performance data, so I don't think there would be any loss from
leaving them out.) Ideally human-readable threshold ranges could also be
included in the message, but given the complexity of threshold specification
I can understand leaving this out.

Currently output looks like this:

NTP WARNING: Offset -0.021909 secs, jitter=15.843000, stratum=1,

or this:

NTP WARNING: Offset -0.002291 secs, jitter=0.659000, stratum=3,

Without consulting my perfdata or configuration, can you tell which
threshold is being exceeded?

Ideally, the output would be something like this:

NTP WARNING: Offset -0.021909 secs, expect 0.010000 > |offset| > 0.100000

and this:

NTP WARNING: Stratum 3, expect stratum < 2

Reply to this email on GitHub:
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the Devel mailing list