NTP Problems

English Applications
1

Hi OpenSuse forums,
I need some help from you guys regarding NTP on OpenSuse

I have a Server ( IP : 192.168.100.52) which is based on Opensuse (12.2 ) and here is my NTP conf file :


server 192.168.100.11  iburst
restrict 192.168.100.11
logfile /var/log/ntp

192.168.100.11 is another NTP Server in the network.

Now, when other linux systems have their ntp server as 192.168.100.52, the NTP service starts, but time does not sync.
in the NTP clients, on ntpdate <IP> or sntp <IP>, the error I get is : “sntp: too many bad or lost packets”

on the NTP Server (192.168.100.52), I get this error in the log file (/var/log/ntp) : 15 Jan 16:38:21 ntpd[15262]: ntpd exiting on signal 15

Also when I do ntpdate <IP> or sntp <IP> on the NTP Server, I get : “sntp: too many bad or lost packets”.

The Server is plain installation of OpenSuse along with Mysql installed.
SuseFirewall is disabled and there is no other Antivirus or firewal in place.
All ports are open,still not able to sync from Server (Also the NTP server is not able to sync with any other machine).

Please guide.

2

Which version?

Why oh why do people not start their questions with a top down aproach: I have …, I want to do …, I did …, I expected …, Instead …

:disapointed:

3

Here’s Your version of the same

I have OpenSUSE installed on a machine which I will be using as a Server (MySQL).

I want to … have this server as a NTP Server to all the clients (thinclients based on Opensuse) as I want same time across the environment.

I did the configuration as :

/etc/ntp.conf :



 	server 192.168.100.11  iburst
restrict 192.168.100.11
logfile /var/log/ntp

where 192.168.100.11 is my WIndows machine which acts as a nTP Server for my OpenSuse NTP server.

I expected it to run fine, and service starts stops witout any erorrs but …

Instead it doesnt work as expected and

I get this error in the log file (/var/log/ntp) : 15 Jan 16:38:21 ntpd[15262]: ntpd exiting on signal 15

Also when I do ntpdate <IP> or sntp <IP> on the NTP Server, I get : “sntp: too many bad or lost packets”.

So inshort, on the NTP SErver based on Opensuse (192.168.100.52) , NTP aint working as expected, I am neither able to act as client or as server.

Additional info :

OpenSuse 12.2 installed ( Cant upgrade or update with latest version as its been used in production).
Firewall is OFF.
Ethernet configured one eth0

4

On 2014-01-16 05:46, shashishirke wrote:

> /etc/ntp.conf :

>
> Code:
> --------------------
> server 192.168.100.11 iburst
> restrict 192.168.100.11
> logfile /var/log/ntp
>
> --------------------

Why only one server? An ntp server needs at least 3.

I do not find documentation for the “restrict” keyword.

> where 192.168.100.11 is my WIndows machine which acts as a nTP Server
> for my OpenSuse NTP server.

Instead, define several servers. One, that internal server, and several
outside servers. If the ntpd finds the only available clock reference is
faulty, it may quit.

> I expected it to run fine, and service starts stops witout any erorrs
> but …
>
> Instead it doesnt work as expected and
>
> I get this error in the log file (/var/log/ntp) :
> _-15_Jan_16:38:21_ntpd[15262]:ntpd_exiting_on_signal_15-

Did you look on /var/messages?

>
> Also when I do ntpdate <IP> or sntp <IP> on the NTP Server, I get :
> “sntp: too many bad or lost packets”.

You may have connectivity problems on your network. That may also cause
ntpd to quit or stop serving.

> Firewall is OFF.

Why?

–
Cheers / Saludos,

Carlos E. R.
(from 12.3 x86_64 “Dartmouth” at Telcontar)

5

My configuration:

boven:~ # grep -v '^#' /etc/ntp.conf


server 127.127.1.0 
fudge 127.127.1.0  stratum 10



driftfile /var/lib/ntp/drift/ntp.drift  

logfile /var/log/ntp             


keys /etc/ntp.keys               
trustedkey 1                     
requestkey 1                     
server ntp.xs4all.nl iburst
boven:~ #

And that fits me perfectly. I do not need three servers. Maybe that is your need for some reason, but tthat is you.

And IMHO it is not a bad idea to use a single server inside your LAN to synchronise the systems in the LAN. That one server (e.g. the one in the DMZ) can then synchronise with the outside world. It restricts the need for Internet connections. When the outside connection is broken, the LAN will still be synchrone internally.

6

On 2014-01-16 15:16, hcvv wrote:

> And that fits me perfectly. I do not need three servers. Maybe that is
> your need for some reason, but tthat is you.

It is recommended on the documentation. I saw ntpd simply quit if I had
not enough servers defined. Ntpd has no way of knowing if the only
server defined is correct because it can not do comparisons.

–
Cheers / Saludos,

Carlos E. R.
(from 12.3 x86_64 “Dartmouth” at Telcontar)

7

I never had more then one on al the Unix systems I ever supported. Never had ntpd quitting.

8

On 2014-01-16 15:56, hcvv wrote:
>
> I never had more then one on al the Unix systems I ever supported. Never
> had ntpd quitting.

Well, I did.

If that one server happens to be not stable at that instant or otherwise
incorrect, ntp servers only polling that one fail.

–
Cheers / Saludos,

Carlos E. R.
(from 12.3 x86_64 “Dartmouth” at Telcontar)

9

Carlos E. R. wrote:
> On 2014-01-16 15:56, hcvv wrote:
>> I never had more then one on al the Unix systems I ever supported. Never
>> had ntpd quitting.
>
> Well, I did.
>
> If that one server happens to be not stable at that instant or otherwise
> incorrect, ntp servers only polling that one fail.

That’s why at our office we have two NTP servers, just as two of every
other kind of network service.

10

On 2014-01-16 18:21, Dave Howorth wrote:
> Carlos E. R. wrote:

>> If that one server happens to be not stable at that instant or otherwise
>> incorrect, ntp servers only polling that one fail.
>
> That’s why at our office we have two NTP servers, just as two of every
> other kind of network service.

Exactly :slight_smile:

–
Cheers / Saludos,

Carlos E. R.
(from 12.3 x86_64 “Dartmouth” at Telcontar)