ntpdate[1364]: no server suitable for synchronization found

[SM_3]

i keep getting this error message, although I have configured the ntp files

can anyone help
in this case both ntp server and client are hp-ux

[Paul Sperry]

usr/sbin/ntpq is very useful in diagnosing time server problems
will show the associations between the various servers and their states

also


If the time difference between your servers is 1000 seconds or greater, xntpd will ignore the servers time. If a client cannot find a server to which to synchronize in about 320 seconds, the xntpd daemon will die. Therefore, check the time differences when you startup and manually adjust them if necessary, first.

[Helen French]

This message is considered "normal" if you have everything set properly. Only thing you need to check is the ntp server it it's sync'd to it's internal clock first (ntpq -p). A bette explanation can be found from TKB #KBRC00001040:

http://support2.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000063204535

[Bill Hassell]

Checki /etc/ntp.conf for the servers by typing:

ntpq -p

It will report back the server(s) that your machine us using. To see if any of them are reachable and running as an NTP server, use:

ntpq -p server_name1
ntpq -p server_name2

and so on. If you have edited /etc/rc.config.d/netdaemons file and specified the preferred source for a reboot NTP source, then a reboot will automatically set the time (jump to the right value) before anything time-sensitive has been started by using ntpdate. But if ntpdate can't find a working server, you'll get the same message.

[SM_3]

from the client ntpq -p returns with:

ntpq -p pilot
remote refid st t when poll reach delay offset disp
==============================================================================
pilot 0.0.0.0 16 - - 64 0 0.00 0.000 16000.0
#

[Helen French]

Try refering to more ntp servers on the config file. In this way if the network is busy on one side, the system will sync to the best available server. You can also check the network traffic. The ntpq output shows it's not in sync now. If you just restarted the ntp daemons, then you may need to wait for some time.

[SM_3]

can you shed any more light on the ntqp -p output?

many thanks

[Michael Knaup · Accepted solution]

Hi,

with this ntpq -p you haven't synced:

remote - the ntp server to your client
refid - the box your ntp server get's it's time from
st - stratum, how much steps is your client away from the real clock (16 is maximum afaik)
t - type (typically u for unicast or b for broadcast)
when - in x seconds the next sync will happen
poll - the sync interval in seconds
reach - a 8 bit shift register that shows you the last 8 syncs, a healthy system should sho 377 after a while
delay - Round trip time (smaller is better)
offset - difference between client and server clock (smaller is better)
disp - don't know exactly, but (smaller is better :-)

Cheers,

Michael

[Helen French]

The output of ntpq shows it's not sync'd yet. Once it's sync'd the '0's will be replaced with respective values.

You need to make sure that an ntp server service is running at the other end. You could check the communication to the server too.

[Indira Aramandla]

I have a similar problem. Our HP Server synchronises with NT server. ntpdate was woring fine, but one day when the NT server was restarted then the ntpdate cron job puts the error "ntpdate[27982]: no server suitable for synchronization found".

And when I saw this forum, I was wondering how can I find out that an ntp server service is running on the NT server.
Because when I give ntpq -p it gives "ntpq: read: Connection refused"
Any info on this??????

Thanks in advance.

[SM_3]

thanks for the info

now i will not use the NT as a time server

i have a connection from the hp-ux box to the internet

thanks to everyone!

[Rod White]

I'll add this comment so that anyone in the future can read this although it is way too late for the current thread.

I used this thread and others to configure three client HP-UX 11.11 boxes to one HP-UX 11.11 box ntp server.

Two of the three ntp clients showed correct connections and one did not, giving a "no suitable ...." message when xntpd started.

ntpdate -d replied that synchronisation was possible. The offset was only two minutes or so.

I tracked the problem down to a syntax error in the server line in /etc/ntp.conf.

sam had put one too many spaces between the version number and the "prefer" word, that is,

"server version 3 prefer" should have been
"server version 3 prefer"

I shut down and restarted xntpd countless times. I check connectivity, ip resolution etc, etc, etc and it wasn't until I removed the extra space that the client connected correctly.

Rod White