[c-nsp] Odd ntp problem...

Jeff Kell jeff-kell at utc.edu
Fri Aug 30 14:41:34 EDT 2013


Ran across a strange one with NTP [again, had seen it happen once
before, but thought it was a fluke...]

Have a 6500 core running 12.2(33)SXI that is setup to sync to an
external NTP source, and in turn provide NTP for our networked devices.

Basic NTP configuration, with ntp logging, ntp update-calendar, and ntp
server statements pointing to the external sources.

Earlier this week, noticed an odd NTP log event, "%NTP-6-PEERREACH: Peer
a.b.c.d is reachable" when in fact a.b.c.d was not a configured peer. 
It was a Windows server configured to use the 6500 as it's NTP source. 
Worse still, the 6500 adjusted it's clock to match the erroneous peer,
and there was resulting hilarity and chaos :(

I'm at a loss to explain why the 6500 would accept an unsolicited peer
(assuming the Windows host was mistakenly trying to setup a peer rather
than client/server relationship), let alone let it override the
established external trust.

Jeff



More information about the cisco-nsp mailing list