ntpd inconsistency

jonaskellens · 02-09-2019, 03:24 AM

Hello

I have several Centos 6.8 servers. Latest version of ntpd (ntp-4.2.6p5-15).

On most of my servers the time is in perfect sync.
Example :

Code:

9 Feb 10:12:37 ntpdate[10141]: adjust time server 149.210.199.182 offset -0.006567 sec

On another server however the time is never in sync, although ntpd is running :

Code:

9 Feb 07:15:12 ntpdate[11103]: step time server 185.99.49.103 offset 5.757200 sec

The only config difference I see is this :
working:

Code:

restrict default kod nomodify notrap nopeer 
restrict -6 default kod nomodify notrap nopeer noquery

not working :

Code:

#restrict default kod nomodify notrap nopeer noquery
#restrict -6 default kod nomodify notrap nopeer noquery
restrict default ignore noquery

This is an install 'out-of-the-box', no modification made to the config file. Only "yum install ntp".

michaelk · 02-09-2019, 05:19 AM

Is ntpd running?
Post the output of the command ntpq -pn

ehartman · 02-09-2019, 06:42 AM

Quote:

Originally Posted by jonaskellens

On another server however the time is never in sync, although ntpd is running :

ntpd will not update the date/time when the difference is TOO large:

Code:

-g, --panicgate
      Allow the first adjustment to be Big.  This option may appear an unlimited
      number of times.

      Normally, ntpd exits with a message to the system log if the offset exceeds
      the panic threshold, which is 1000s by default. This option allows the time
      to be set to any value without restriction; however, this can happen only
      once.

So make sure the first startup of ntpd (after reboot) has this option added

Code:

CMDLINE="/usr/sbin/ntpd -g" in whatever your system uses to start it

so that a faulty hardware clock doesn't prevent ntpd from working.

jonaskellens · 02-09-2019, 08:53 AM

Yes ntpd is running.

Code:

[root@41 admin]# ntpq -pn
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
 195.200.224.66  .INIT.          16 u    - 1024    0    0.000    0.000   0.000
 51.255.138.215  .INIT.          16 u    - 1024    0    0.000    0.000   0.000
 185.99.49.101   .INIT.          16 u    - 1024    0    0.000    0.000   0.000
 185.99.49.102   .INIT.          16 u    - 1024    0    0.000    0.000   0.000

I always stop ntpd, then run "/usr/sbin/ntpdate pool.ntp.org" and then restart ntpd. So the difference should not be "too large". But after a day or 2, it gets larger...

michaelk · 02-09-2019, 09:08 AM

Wait until reach is 177 and post the numbers again.

By default ntpd is configured to use the -g option. Is the computer a real or virtual machine and is it on continuously or does it go to sleep/hibernate?

jonaskellens · 02-18-2019, 07:19 AM

While clock is again running behind by 10 seconds, 'reach' stays at '0' :

Code:

[root@41 admin]# ntpq -pn
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
 4.53.160.75     .INIT.          16 u    - 1024    0    0.000    0.000   0.000
 208.75.89.4     .INIT.          16 u    - 1024    0    0.000    0.000   0.000
 208.67.72.50    .INIT.          16 u    - 1024    0    0.000    0.000   0.000
 216.230.228.242 .INIT.          16 u    - 1024    0    0.000    0.000   0.000

michaelk · 02-18-2019, 08:00 AM

It looks like ntp is never getting past the INIT stage.

This isn't an out of the box installation since the ntp.conf file is using "restrict default ignore noquery" and the default restrict directives have been commented out. Using that restrict blocks time polls unless specific servers are allowed.

https://support.ntp.org/bin/view/Sup...ection_6.5.1.1.