[LEAPSECS] leap second festivities?

Miroslav Lichvar mlichvar at redhat.com
Wed Jul 1 06:48:12 EDT 2015


On Tue, Jun 30, 2015 at 09:42:50PM +0100, Rob Seaman wrote:
> Any thoughts on watching Google’s (or anybody else’s) smear in action?  Kind of like watching paint dry, but still…

It seems the Google leap smear has finished. Here is a plot of offsets
of some Google servers I included in my leap second monitoring.

https://mlichvar.fedorapeople.org/leap2015/google_smear.png

It was apparently a linear smear which is difficult to follow for NTP
clients. I'm curious why they switched from the cosine function. I
guess it allows them to mix systems which correct the clock by slewing
locally with systems that follow a leap smearing NTP server.

I was also monitoring a lot of pool.ntp.org servers. Only about 80% of
them were announcing the leap second in the last measurement that was
made before the leap second. To see how many of them handled the leap
second correctly, here is a cumulative distribution plot of their
offsets before and after the leap second.

https://mlichvar.fedorapeople.org/leap2015/pool_cdf.png

It looks like more than 10% of the servers didn't handle it. Some of
them are still off by one second.

Anyway, here is how I saw the leap second on two of my systems. The
clock on the first one was corrected by stepping, the second one by
fast slewing.

https://mlichvar.fedorapeople.org/leap2015/clock_step.gif
https://mlichvar.fedorapeople.org/leap2015/clock_slew.gif

-- 
Miroslav Lichvar


More information about the LEAPSECS mailing list