I read this traceroute to indicate two things (return-path would be nice but gotta work with what I got): The routers at hop #11 (I see it varies due to load balancing per packet) claim that the destination, 38.101.32.14, is unreachable (i.e. you are getting back ICMP host-unreach from those routers). That could mean either they don't have a route for whatever 38.101.32.14 falls under (network-wise), or that the routers are aware that 38.101.32.14 isn't reachable/alive (god forbid they have a /32 for it) You might wonder "what's up with hop #11 then 'doing it again' at hop #12?", and the answer is probably that the routers at hop #11 are supposed to have knowledge/reachability of/to 38.101.32.14, but don't (for whatever reason), and are configured by default to punt packets back to another router to try and find a different path (i.e. bouncing between 154.54.28.82 and 154.54.5.161, hoping one can find the thing the other cannot). TL;DR -- You should be talking to both Cogent, as well as whoever maintains this video.primelink1.com service/server/network, to inform them of what's going on. It's obvious something somewhere is either broken or administratively set to behave this way. -- | Jeremy Chadwick jdc@koitsu.org | | UNIX Systems Administrator http://jdc.koitsu.org/ | | Making life hard for others since 1977. PGP 4BD6C0CB | On Fri, May 09, 2014 at 02:09:22PM -0400, Kyle Farmer wrote:
$ traceroute video.primelink1.com traceroute to video.primelink1.com (38.101.32.14), 64 hops max, 40 byte packets 1 199.48.129.177 (199.48.129.177) 0.451 ms 0.619 ms 0.361 ms 2 208.79.82.101 (208.79.82.101) 0.428 ms 0.376 ms 0.468 ms 3 ge-6-8.car2.Raleigh1.Level3.net (4.71.162.13) 3.732 ms 1.847 ms 2.320 ms 4 ae-9-9.ebr1.Washington1.Level3.net (4.69.132.178) 7.817 ms 7.624 ms 7.572 ms 5 ae-91-91.csw4.Washington1.Level3.net (4.69.134.142) 7.578 ms ae-61-61.csw1.Washington1.Level3.net (4.69.134.130) 7.565 ms ae-81-81.csw3.Washington1.Level3.net (4.69.134.138) 7.814 ms 6 ae-2-70.edge3.Washington4.Level3.net (4.69.149.82) 7.635 ms ae-4-90.edge3.Washington4.Level3.net (4.69.149.210) 7.400 ms 7.593 ms 7 COGENT-COMM.edge3.Washington4.Level3.net (4.68.63.174) 8.157 ms 8.163 ms 8.661 ms 8 be2177.ccr22.dca01.atlas.cogentco.com (154.54.41.206) 9.385 ms be2113.mpd21.dca01.atlas.cogentco.com (154.54.6.170) 9.258 ms be2177.ccr22.dca01.atlas.cogentco.com (154.54.41.206) 9.237 ms 9 be2148.ccr21.jfk02.atlas.cogentco.com (154.54.31.118) 15.374 ms 15.926 ms be2149.ccr22.jfk02.atlas.cogentco.com (154.54.31.126) 15.705 ms 10 be2109.ccr21.alb02.atlas.cogentco.com (154.54.3.138) 27.944 ms be2107.ccr21.alb02.atlas.cogentco.com (154.54.3.94) 28.124 ms be2108.ccr21.alb02.atlas.cogentco.com (154.54.3.134) 27.568 ms 11 te0-0-2-2.agr11.alb02.atlas.cogentco.com (154.54.28.82) 28.007 ms 28.353 ms te0-0-2-0.agr11.alb02.atlas.cogentco.com (154.54.5.161) 27.332 ms 12 te0-0-2-2.agr11.alb02.atlas.cogentco.com (154.54.28.82) 4653.993 ms !H te0-0-2-0.agr11.alb02.atlas.cogentco.com (154.54.5.161) 4608.391 ms !H 4553.602 ms !H $
Above is the full traceroute. What I find is pinging the end route you'll see it jump up to 10000MS and down to 9000MS. Almost like a bad BGP upstream.
On Fri, May 9, 2014 at 2:07 PM, Joel Maslak <jmaslak@antelope.net> wrote:
The !H indicates host unreachable. I'm guessing that some router near the host has a 4.5 second ARP timeout. It's not a latency issue, but probably an end host issue.
On Fri, May 9, 2014 at 11:12 AM, Kyle Farmer <bannereddivpool@gmail.com>wrote:
Cogent has had a issue for a few days now, causing extreme latency across the network;
be2437.ccr21.alb02.atlas.cogentco.com (66.28.4.197) 38.219 ms 10 te0-0-2-0.agr11.alb02.atlas.cogentco.com (154.54.5.161) 37.838 ms te0-0-2-2.agr11.alb02.atlas.cogentco.com (154.54.28.82) 37.968 ms te0-0-2-0.agr11.alb02.atlas.cogentco.com (154.54.5.161) 38.100 ms 11 te0-0-2-0.agr11.alb02.atlas.cogentco.com (154.54.5.161) 4645.684 ms !H 4503.251 ms !H te0-0-2-2.agr11.alb02.atlas.cogentco.com (154.54.28.82) 4581.944 ms !H
Can anyone verify what is going on? I called them a few weeks ago so I'm hoping this might hit some admins who work there.
--Kyle
_______________________________________________ Outages mailing list Outages@outages.org https://puck.nether.net/mailman/listinfo/outages
_______________________________________________ Outages mailing list Outages@outages.org https://puck.nether.net/mailman/listinfo/outages