Re: NTP PEERUNREACH

pwheelehan · ‎05-20-2004

Hello, i have a 7500 running 12.3. It is the NTP Master for my remote routers. I have the remote routers configured with the ntp peer x.x.x.x source serial 0/0 prefer command. Just this past week i have started getting NTP PEERUNREACH Messages and the ntp status shows that they are no longer sync'd.

Any ideas?

Richard Burts · ‎05-20-2004

Where is the 7500 getting time from?

Has anything changed on the 7500?

If for some reason the 7500 is no longer considered authoritative, the other routers will not sync with it.

You did not say anything about it so I will assume that there are no IP connectivity problems. If there were that might disrupt sync (but a number of other thingsl would also probably be disrupted).

Are you doing any NTP authentication or any NTP access-group on either the 7500 or the remotes? This might also disrupt sync.

HTH

Rick

pwheelehan · ‎05-20-2004

Hello, nothing changed on the 7500. I do not have it getting time from any ntp source I just set the time and made it master. The remote routers are connected via satelite t-1s and the circuits do bounce quite often, but the disruption in sync occurs even on routers whose circuits have not bounced. We have recently swaped out the routers from 1750's to 1760's and upgraded to ios 12.3.4t4 and 12.3.6a

Thanks

Richard Burts · ‎05-21-2004

I have a number of 1721s running 12.3(4)T1 which report loss of NTP sync every 12 hours. I have an open case with TAC trying to establish that there is a bug with NTP in the T train of 12.3.

My routers report loss of NTP sync and immediately regain NTP sync. Do your routers lose/regain or does the loss of NTP sync last for some time?

HTH

Rick

pwheelehan · ‎05-21-2004

When The routers loss NTP sync some of them do not recover immediately. Some were not sync'd in over a day.

jeff.bankston · ‎05-21-2004

I'm running 12.2(15) Ip Plus code on our 1760 and 2600 sites back to a 3745 same codeset whom is exhbiting the same problem. The 3745 host was config'd as ntp master 1 for the field sites.

when we started sync'ing the 3745 to a public ntp stratum 2 clock, that problem went away.

FWIW

-Jeff