Re: RH9 NTP trouble

From: Tauno Voipio (tauno.voipio_at_iki.fi.NOSPAM.invalid)
Date: 01/01/05 

Date: Sat, 01 Jan 2005 12:11:32 GMT

Lawrence DčOliveiro wrote:
> In article <c5gga2-s2d.ln1@news.sebs.org.in>,
> Dr Balwinder Singh Dheeman <bsd.sanspam@cto.homelinux.net> wrote:
>
>
>>On 12/31/04 14:40, Lawrence DčOliveiro wrote:
>>
>>>I know, I know, Red Hat 9 is obsolete, but it isn't worth the trouble of
>>>putting anything else on those machines to fix this problem.
>>>
>>>The problem is, they won't sync to NTP servers. ntpq -p reports jitters
>>>of 4000 and reach values of 0 for all servers I put in ntp.conf. Another
>>>machine running SuSE 9.1 on the same network can sync to the same
>>>servers just fine. All machines access the Internet through an ADSL
>>>router doing NAT.
>>>
>>>Any thoughts? Maybe I should just rebuild from the xntpd sources...
>>
>>For most of my customers on dialup, what I do is -- sync only one
>>machine to the real servers (running ntpdate) and this machine becomes
>>time server for rest of the host on the interanet.
>
>
> And why would that be relevant to someone on ADSL?

As already said - sync your master server to extenal NTP
and use it as the local network master clock. If the local
network in an Ethernet LAN, you could also think of using
the NTP broadcast mode for the local clients - the network
latencies are even in 10BASE-T so small that the difference
between broadcast and connected NTP is negligible.

Usually the best external NTP source is your ISP, due to
small and repeatable latencies. Try ntp.myisp (whatever
it is; mine is ntp.inet.fi, as the ISP is inet.fi).

Please note that your clock must be pretty close to the
correct time to sync with NTP. Use ntpdate to coarse
sync before starting the service. 

-- 
Tauno Voipio
tauno voipio (at) iki fi

Relevant Pages

  • Re: Simple ntp setup, but I must be doing something wrong
    ... When I get things working properly I will switch to using the pool ntp servers so that I am not relying on a single ntp server. ... Over the last several weeks I have been trying to get the NTP service running properly on my company linux file server, but am not having much success. ... My objective is to synchronize my linux server with a secondary time server or one of the ntp pools, then synchronize a handfull of desktop windows systems with the file server. ... Below is my config file, the output of ntpq -p and ntpq -pn, and the log messages from the last restart of ntp that matches the configuration file shown below. ...
    (comp.protocols.time.ntp)
  • Synchronizing RedHat to two W2k Domain Controllers (was: Newbie at NTP)
    ... Linux servers get synchronized to the PDC through NTP. ... the best time server. ... NTP (basically the same source code as you run on Linux). ... you will need to provide a real time server for the domain ...
    (comp.protocols.time.ntp)
  • Re: [SLE] 10.0 and NTP
    ... >> Yet ntp is working fine. ... If you have a server with an asterisk in front, ... source of synchronization. ... my company time server has the output ...
    (SuSE)
  • Re: Time steping regardless of "-x" slew only option
    ... I realize that I have other issues causing the need for steps. ... I only have one time server, ... But I still need to understand why NTP is stepping when the ...
    (comp.protocols.time.ntp)
  • Re: broadcast client
    ... synchronizing with the same ntp server. ... ntp displined clock. ... because the drift discipline is inversely ... The reason for the backup to long poll intervals is ...
    (comp.protocols.time.ntp)