NTP time problem

UNIX AIX
1

I got an ntp time problem on AIX server.

os version is AIX7.1
OS LEVEL 7.1.0.0

i got below output,when i run the below command

bash-3.2# ntpdate -dv XXXXXXXXXXXXXXXXXXXXXXXX
 4 Dec 12:50:49 ntpdate[34209840]: 3.4y
transmit(xxxxxxxxx)
receive(xxxxxxxxx)
transmit(xxxxxxxx)
receive(xxxxxxxxx)
transmit(xxxxxxx)
receive(xxxxxxxx)
transmit(xxxxxxx)
receive(xxxxxxxx)
transmit(xxxxxxxx)
server xxxxxxxxxxxx, port 123
stratum 5, precision -6, leap 00, trust 000
refid [xxxxxxxxxxx], delay 0.04141, dispersion 0.00021
transmitted 4, in filter 4
reference time:      d649a1e9.ff99a705  Wed, Dec  4 2013 12:43:21.998
originate timestamp: d649a3aa.088f6994  Wed, Dec  4 2013 12:50:50.033
transmit timestamp:  d649a3aa.0c5de000  Wed, Dec  4 2013 12:50:50.048
filter delay:  0.04152  0.04141  0.04141  0.04141
               0.00000  0.00000  0.00000  0.00000
filter offset: -0.01428 -0.01454 -0.01475 -0.01495
               0.000000 0.000000 0.000000 0.000000
delay 0.04141, dispersion 0.00021
offset -0.014544

 4 Dec 12:50:50 ntpdate[34209840]: adjust time server xxxxxxxxxxx offset -0.014544

please can some one help me on this sync time problem with peers

2

What you think is wrong in this?

You might see some milliseconds delay.

Try the command couple of times.

3 
 Nagios monitoring tool alerting as below.may be its not syc

Service: NTP Time
State: OK -> CRITICAL (PROBLEM)
Command: check_mk-ntp.time
Output: CRIT - found 2 peers, but none is suitable
Perfdata:
4 
ntpq -np

?

5 

bash-3.2# ntpq -np
     remote           refid      st t when poll reach   delay   offset    disp
==============================================================================
*xx.xxx.xx.xxx   xx.xxx.xx.xx     5 u  124  128  377     1.01   -5.592    3.80
+xx.xxx.xx.xxx   xx.xxx.xx.xx     5 u   33  128  377     3.52    7.586    3.86
bash-3.2#
6

Two peers, both work.
Something must be wrong with your Nagios check!

7

It is usually not a good idea to use two remote ntp servers. If there is a difference, who should the host believe?

ntp: 5.3.3. Upstream Time Server Quantity.

8

In practice it works as good as one peer, plus you have a fallback.

9

What do you mean with "in practice"? Perhaps you mean you set it up like this and it worked? Maybe these servers themselves were local servers syncing to the same higher stratum ntp servers and therefore they happened to never deviate too much from one another? The recommendation is pretty clear in the link above. I have never made a setup with two.

10

Yes, have set up NTP clients with two local NTP servers configured.
The local NTP servers are NTP clients of three main NTP servers, and indeed not much drifting from each other.
Nevertheless the NTP clients prefer one-and-the-same local NTP server.
So there is least time deviation between them. If one local NTP server goes down, they all drift to the other in the same manner.