Full List of NTP Peers, Leap Second 2008

Mix of IPv6 clients - no large offsets, so it looks like it all went well.

Local clock - nothing to see here.

Our GPS refclock - looks like we latched on to the wrong second pulse after the leap second, and this fixed itself up when the GPS receiver next lost sync. However, it was identified as a falseticker and so didn't cause problems.

Rackety was well behaved, as expected.

Some wobbles in swisstime, but no leap second misbehaviour.

We lost a few packets talking to this TCD peer, but there are no big jumps around the leap second.

This TCD client looks like it got it wrong, and probably steped the clock to fix things.

This TCD client looks like it got it wrong and then slewed the clock to fix things.

This TCD peer seems to have done the right thing.

This is a peer in Galway that does the right thing, though there are some strange blips in the jitter. It has a GPS refclock, so this may be the signal comming and goint there.

This machine looks like it does the wrong thing and then steps to fix.

The step in this peer looks like it is due to an exceptionally delayed packet rather than the leap second.

Peer in UCD - seems well behaved.

Some spikes in the delay here, but nothing much suspicious in the offsets.

This peer looks like it gets it wrong and then oversteps on the way back.

This peer probably got it wrong and stepped, but we aren't seeing enough samples to really be sure.

Some oddities here after the leap, but they seem to be RTT related.

This peer has some odd behaviour, and seems to wind up off by a bit over half a second for a while.

This peer seems to decide to sit at about -100ms (maybe it's behind a link with asymetric load with an internal source of time) and goes back to that after getting the leap wrong by slewing the clock.

This peer is so badly behaved it is hard to know what's actually going on.

Looks like a well behaved peer with a RTT spike.

Usually seems well behaved, bar some RTT spikes.

Looks generally badly behaved.

Looks like leap second was missed and then stepped back.

Very well behaved looking peer.

Well behaved peer with interesting diurnal behaviour.


No problems here.

This actually looks like routing problems rather than leap second.

Another routing spike, but no obvious leap second problems.

Spike, no leap.

Probably a leap second related blip?

A jump followed by decay to the previous value. Interesting long term behaviour too.

Pretty well behaved peer.

Another well behaved peer.

Probably does something bad for the leap second, but it pretty erratically behaved at the best of times.

Another pretty confused machine.

Pretty well behaved, certainly no leap blips.

Looks like a missing leap followed by a step to correct.

Probably actually a routing blip?

Very weird - no leap blip, but the host seems to be getting further and further away!

The delay for this one looks weird, but there is a jump of 500ms near the leap - could be related to delay asymetry somehow though.

Looks like a reasonably well behaved peer, bar the big asymetric rtt about a day before the leap.

This one looks like it began to wander off in the wrong direction!

Seems well behaved.

This host shows odd delay spurts, but there may be a leap realted jump too.

Looks like it was well synced and then missed the leap.

A leap followed by a damped oscilation. Interestingly, the delay jumps after the leap.

Looks like a step half a second in the wrong direction!

Just looks like a rtt bump for this one.

I think this one missed the leap, but usually has a offset anyway.

No problems with the leap here, but clear problems with asymetric round trips.

There's rather a lot of points on this, maybe multiple machines? Looks like one missed the leap though.

Similar to the last one.

Looks like a little leap-related purturbation.

Lots of delay spikes here, but nothing leap realted.

Slowly wandering around with a big jump for the leap.

Another machine getting further and further away...

Not enough points to say what's going on.