[ntp:questions] "pool" directive and 4.2.8p8

Brian Inglis Brian.Inglis at SystematicSw.ab.ca
Mon Nov 14 14:15:32 UTC 2016

On 2016-11-14 04:57, Gabor Kiss wrote:
> Recently I reconfigured two dozens of NTP 4.2.6p5 clients
> in order to use a single "pool" config line instead of four "server"s.
> Everything works well.
> Now I met a 4.2.8p8 daemon on a virtual machine with IPv6.
> I set the config as I used to:
> 	tos minclock 3 maxclock 5
> 	pool 2.hu.pool.ntp.org iburst
> (2.hu.pool instead of simple hu.pool is intentional)
> But after restart I can see this:
> # ntpq -p
>      remote           refid      st t when poll reach   delay   offset  jitter
> ==============================================================================
>  2.hu.pool.ntp.o .POOL.          16 p    -   64    0    0.000    0.000   0.000
> #
> That is quite strange. There are no servers, no sync.
> I had to put back the more conventional config:
> 	server 0.hu.pool.ntp.org iburst
> 	server 1.hu.pool.ntp.org iburst
> 	server 2.hu.pool.ntp.org iburst
> 	server 3.hu.pool.ntp.org iburst
> What did I wrong?

Try running:
	$ ntpq -p -c rv $server
to see what associations are mobilised.

Check your pool server(s) with:
	$ host 2.hu.pool.ntp.org
2.hu.pool.ntp.org has address
2.hu.pool.ntp.org has address
2.hu.pool.ntp.org has address
2.hu.pool.ntp.org has address
2.hu.pool.ntp.org has IPv6 address 2001:738:0:811:230:48ff:fef2:8240
2.hu.pool.ntp.org has IPv6 address 2001:738:0:830:1::106
2.hu.pool.ntp.org has IPv6 address 2001:4c48:1::1
2.hu.pool.ntp.org has IPv6 address 2001:738:0:800:1::251

Try removing the tos statement and add preempt to the pool statement:
	pool 2.hu.pool.ntp.org iburst preempt
to allow ntpd to drop unavailable or poor candidates in recent releases.

Then rerun ntpq as above after the service has been up 15-20 minutes.
Post the results if you need more help.

Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

More information about the questions mailing list