[ntp:questions] Curious change in traffic
mayer at ntp.isc.org
Sat Mar 17 03:47:23 UTC 2007
David Malone wrote:
> Over the last couple of weeks I have seen a slightly strange effect
> over the last few weeks. Here are some NTP stats from a former NTP
> server of ours, plotting the iostats shown by ntpdc:
> This machine is still used by a lot of "legacy" clients. For the
> last few yeas, its stats show that it usually sends and receives
> the same number of NTP packets. Since sometime about late Feb/start
> of March, it has been sending more than it receives, which is a
> little strange.
> I've no real ideas what might be causing this, other than maybe
> some software upgrade that's been performed as a consequence of the
> recent TZ change. I'm curious if anyone has any other ideas as to
> a possible cause?
Well apart from the fact that only the US changed time zone this week,
the only way you get amplification like this is by using queries (ntpdc
and ntpq). If you restrict queries to only your own internal network
does it subside? Also what does ntpdc -c monlist tell you about the
machines contacting it?
More information about the questions