I get that equipment on some hops just don't support pinging so those will always show 100% loss. Some hops, however, show intermittent loss - which I'm assuming is a problem. So by my understanding a hop that shows, for example, 75.3% loss is likely causing a problem.
However - in the latency graph below I don't fully understand why those packet losses don't show up more often as red vertical lines. Is this because tcp/ip retries are successful, or is it because I am misunderstanding the tests?

Hey Videstra,

Thanks for reaching out!

When you see all that red show up in a trace, it's only natural to think something is wrong. However, just like with those hops that always show 100% packet loss, as long as the packet loss doesn't carry through to the destination, it's not something to worry about!

This sort of packet loss indicates that (hop #11) is "de-prioritizing" or dropping ICMP Time Exceeded responses for a majority of the time however is responding to some ICMP packets. This behavior affects only the returning information and does not affect forwarded packets just like those hops that show 100% packet loss.

If you're interested, here are a couple of articles that can explain this concept a bit more:

One Poorly Responding Router
Interpreting Latency and Packet Loss


I hope that helps clear things up! If you have any other questions, please feel free to ask.

Thanks,

-TJ

Thanks - this is very helpful (as were those articles)!
I especially liked the line: