Thanks for sending the data - this helps us understand the issue a lot better.

There are a few participating factors here.

First, hop 14, 15 and 16 in your route have stopped responding completely, and have not responded for a significant period of time.

Second, your "Samples to include" is set to only a period in time where hops 14-16 have not responded.

So, when you look at the upper graph, you're seeing hops 1-13 just fine, and a "phantom" hop 14 (which, honestly, should be showing "Destination Unreachable", but is instead showing a phantom hop 14).

If you change your "Samples to include" to 0 (ie: ALL), you'll once again see hops 14-16, but with some amount of packet loss (you have to go back about 8 hours to see real data here).

Now, when you export with your "Samples to include" set to 10 (as you did), the export logic shows the last full route it saw, which is 16 hops long, but since the last 3 hops have not responded in the time window you're examining, those hops get removed from the display. The export logic doesn't properly handle this, and it lists the full route and then doesn't show the timeouts on the final 3 hops - it just shows samples up to hop 13.

This is certainly a problem in the export logic, which we'll get logged. For now, you can know that this situation means the final 3 hops have not responded during the sample period in question. In addition, if you want to see the history for those hops, you can increase your "Samples to Include" to a big enough number to include a time when those hops *did* respond. Those hops haven't participated in routing for a while, so the valid way to report is "Unreachable" and to hide those hops, but you can always get to that information by increasing the time window you're examining.

Make sense? It's a bit complex, feel free to redirect with any questions / misunderstanding.

- Pete