Disclaimer: I'm not nearly as experienced with IPv6 as I am with IPv4, so if I'm leaving out any obvious info, let me know.
Has anyone noticed any increased latency (recently) connecting to google services (google.com/gmail/youtube/etc) over IPv6 (like a +20-25ms increase) versus connecting over IPv4?
I'm running dual stack (comcast native ipv6, and ipv4... asus rt-n66u router, all my devices are using SLAAC to get their IPv6 address, most devices prefer IPv6 over IPv4, so thats why I've noticed this slight latency issue).
basically, if (while wired, not wireless) I do a ping to google services, i get average of 16-21ms response time against their IPv4 addresses. If I do the same using ping6, the response time is 40-45ms (this is while my LAN is idle, btw). It's not all IPv6 destinations.. Facebook is "fine" (no difference in response times, v4 vs v6), a site like berkeley.edu, all good.
So, my thinking is there is something in transit between me and google services that is introducing that extra 20ms of latency.. sure enough, looking at some traceroutes, you can see where the extra latency comes in:
ryan@skynet:~$ traceroute google.comtraceroute to google.com (74.125.239.39), 30 hops max, 60 byte packets 1 bradyrouter.bradynet.local (192.168.1.1) 0.243 ms 0.500 ms 0.600 ms 2 c-67-187-136-1.hsd1.ca.comcast.net (67.187.136.1) 21.574 ms 21.584 ms 32.555 ms 3 te-8-3-ur01.toyon.ca.ccal.comcast.net (68.87.203.149) 12.323 ms 12.532 ms 12.653 ms 4 te-9-3-ar01.fresno.ca.fresno.comcast.net (68.87.200.5) 13.835 ms 13.846 ms 13.869 ms 5 pos-0-4-0-0-cr01.sacramento.ca.ibone.comcast.net (68.86.90.21) 16.979 ms pos-0-2-0-0-cr01.sacramento.ca.ibone.comcast.net (68.86.92.181) 15.130 ms pos-0-4-0-0-cr01.sacramento.ca.ibone.comcast.net (68.86.90.21) 16.973 ms 6 he-0-1-0-0-cr01.sanjose.ca.ibone.comcast.net (68.86.87.178) 17.875 ms 17.513 ms 17.491 ms 7 pos-0-1-0-0-pe01.529bryant.ca.ibone.comcast.net (68.86.87.2) 19.082 ms 19.022 ms 18.799 ms 8 66.208.228.226 (66.208.228.226) 16.524 ms 16.024 ms 16.021 ms 9 72.14.232.136 (72.14.232.136) 16.557 ms 14.816 ms 15.415 ms10 72.14.232.35 (72.14.232.35) 19.959 ms 19.064 ms 18.587 ms11 nuq04s19-in-f7.1e100.net (74.125.239.39) 18.137 ms 16.447 ms 16.887 ms ryan@skynet:~$ traceroute -6 google.comtraceroute to google.com (2607:f8b0:4010:801::1006), 30 hops max, 80 byte packets 1 2601:c:9c80:1d6:62a4:4cff:fe29:6cb8 (2601:c:9c80:1d6:62a4:4cff:fe29:6cb8) 0.314 ms 0.546 ms 0.591 ms 2 * * * 3 te-9-3-ur01.toyon.ca.ccal.comcast.net (2001:558:1a2:5d::1) 11.189 ms 11.181 ms 11.166 ms 4 2001:558:210:a8::1 (2001:558:210:a8::1) 12.399 ms 12.390 ms 12.919 ms 5 pos-0-4-0-0-cr01.sacramento.ca.ibone.comcast.net (2001:558:0:f685::1) 17.241 ms pos-0-1-0-0-cr01.sacramento.ca.ibone.comcast.net (2001:558:0:f6a1::1) 15.276 ms pos-0-4-0-0-cr01.sacramento.ca.ibone.comcast.net (2001:558:0:f685::1) 17.198 ms 6 he-0-6-0-0-cr01.sanjose.ca.ibone.comcast.net (2001:558:0:f5b2::2) 17.266 ms 18.997 ms 18.921 ms 7 pos-0-2-0-0-pe01.529bryant.ca.ibone.comcast.net (2001:558:0:f5c1::2) 20.681 ms 20.653 ms 20.713 ms 8 2001:559::386 (2001:559::386) 45.158 ms 45.071 ms 45.135 ms 9 2001:4860::1:0:21 (2001:4860::1:0:21) 43.994 ms 44.498 ms 43.952 ms10 2001:4860:0:1::3d5 (2001:4860:0:1::3d5) 44.428 ms 44.494 ms 41.626 ms11 2607:f8b0:8000:24::4 (2607:f8b0:8000:24::4) 41.974 ms 2607:f8b0:8000:24::6 (2607:f8b0:8000:24::6) 44.079 ms 2607:f8b0:8000:24::4 (2607:f8b0:8000:24::4) 42.126 ms ryan@skynet:~$ ping6 google.comPING google.com(nuq05s02-in-x00.1e100.net) 56 data bytes64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=1 ttl=49 time=41.0 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=2 ttl=49 time=42.7 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=3 ttl=49 time=41.5 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=4 ttl=49 time=43.1 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=5 ttl=49 time=43.2 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=6 ttl=49 time=41.8 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=7 ttl=49 time=53.0 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=8 ttl=49 time=39.1 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=9 ttl=49 time=45.9 ms64 bytes from nuq05s02-in-x00.1e100.net: icmp_seq=10 ttl=49 time=41.1 ms ryan@skynet:~$ ping google.comPING google.com (74.125.239.142) 56(84) bytes of data.64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=1 ttl=54 time=17.7 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=2 ttl=54 time=18.6 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=3 ttl=54 time=16.3 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=4 ttl=54 time=18.4 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=5 ttl=54 time=17.0 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=6 ttl=54 time=18.9 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=7 ttl=54 time=17.7 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=8 ttl=54 time=16.5 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=9 ttl=54 time=18.3 ms64 bytes from nuq05s02-in-f14.1e100.net (74.125.239.142): icmp_req=10 ttl=54 time=18.0 ms
anyways, if I'm leaving out anything obvious, let me know. again, this issue does not appear to be specific to just one of my computers. I also believe this is a "new" issue, because I distinctly remember doing various latency and "speed" checks when I lit up IPv6 (couple weeks ago) and never noticed this increase of latency on one stack versus the other.
Finally, I'm located in Northern California. Sonora to be exact. I'm on blast (50/10), my signals are all good, connection quality is fine (no hiccups, no packet loss, no signal blips).
Comcast does have an outstanding networking issue on my CMTS that is related to some overloaded downstream ports (one in particular, mine, due to combining...6 nodes on one port), and they are working with Cisco on another issue regarding firing up 4 additional QAMs (going 8 channel downstream), but these issues have been occuring for months and is separate from this specific IPv6 issue.
thanks!
-ryan
↧