Are traceroute hop latency measurements relative to the previous hop or to the origin?

Question

With traceroute output of this:

traceroute to google.com (216.58.223.14), 64 hops max, 52 byte packets
 1  192.168.119.1 (192.168.119.1)  1.321 ms  0.980 ms  0.936 ms
 2  rtr3-c16-dc1.macrolan.co.za (41.222.225.255)  1.815 ms  1.578 ms  1.788 ms
 3  ae0.0.rtr1-ca12-tc1.macrolan.co.za (154.70.222.7)  1.861 ms  2.603 ms  1.989 ms
 4  xe-0/0/3.4000.rtr1-c3h12-tc2.macrolan.co.za (129.205.134.27)  38.447 ms  46.682 ms  21.668 ms
 5  google.ixp.joburg (196.60.8.166)  17.787 ms  18.059 ms  17.713 ms
 6  72.14.237.239 (72.14.237.239)  17.668 ms  17.882 ms  17.497 ms
 7  jnb01s07-in-f14.1e100.net (216.58.223.14)  17.528 ms  18.004 ms  17.642 ms

How do I get the latency to a particular hop destination? i.e. hop No. 3 has a latency of 1.861ms. I assume this is relative to the origin (i.e. my PC). Is this assumption correct?

Answer 1

The return times in traceroute are always relative to the sending source. Due to different ICMP handling in the hops, some returns may be slower than subsequent ones that are actually more distant.

Even though traceroute doesn't normally use ICMP echo requests (most often it uses dummy UDP datagrams), the TTL expired message from the hop in question does use ICMP and some hops may require some (more) time to generate and send out that messages.