On Oct 8, 2011, at 16:01, Tom Hek wrote: > I think the peering between my ISP, Leaseweb, and the monitoring system is > really bad, because I get every one or two times a month a mail my NTP > service is unreachable while it's perfectly available. Anyone else having > this problem? This is my system: > http://www.pool.ntp.org/scores/2001:1af8:4400:a00e:1::1 The graphs look > horrible, but from my own probes to this system from another v6 host on > another network everything looks fine, even when I'm "unreachable" according > to the pool monitoring service. > > Can something be done about this?
Yes, but we can't do anything on this list. You need to tell your provider to check their IPv6 announcements / connections / providers / peers. It looks like they only have local routes in Europe and a bit of peering. Their looking glass / route server - http://noc.leaseweb.com/lg/ - doesn't seem to support IPv6 BGP lookups, but when I traceroute to servers in California on various networks it goes nowhere (www.v6.facebook.com, a California akamai v6 IP - 2600:1406::5043:4a90 - www.pool.ntp.org, etc). >From the NTP Pool server the route I see is: flags destination gateway lpref med aspath origin I 2001:1af8::/32 2607:f238::1 120 1485 30071 16265 i A traceroute gets "stuck" somewhere in AS30071 - from the RTT times I'm guessing in Europe. [root@ntplax1 ~]# mtr --report 2001:1af8:4400:a00e:1::1 ntplax1.ntppool.net Snt: 10 Loss% Last Avg Best Wrst StDev fe2-8.hr01.lax2.phyber.com 0.0% 0.5 0.6 0.5 0.8 0.1 xe2-8.cr02.lax1.phyber.com 0.0% 0.7 1.4 0.5 7.7 2.2 2001:504:13::8 0.0% 0.6 0.7 0.6 0.7 0.1 bbr01-v441.atln01.occaid.net 10.0% 55.9 56.1 55.7 57.3 0.5 bbr01-g0-1.asbn01.occaid.net 10.0% 71.6 76.2 70.9 94.3 9.4 bbr01-p2-1.whkn01.occaid.net 10.0% 76.6 100.8 76.1 296.2 73.3 bbr01-p2-0.lndn01.occaid.net 0.0% 147.1 158.9 146.4 267.4 38.1 bbr01-g1-0.slgh01.occaid.net 10.0% 149.2 151.3 148.2 171.2 7.5 ibr01-g1-1.slgh01.occaid.net 0.0% 149.9 169.5 148.7 342.0 60.6 ??? 100.0 0.0 0.0 0.0 0.0 0.0 http://www.peeringdb.com/view.php?asn=30071 At my work network I see various routes; the "best" one is via AS25973 (our network and AS16265 are both customers or peers); the traceroute ends in AS25973 in Europe - I am guessing again at the lease web border. vyatta@edge1:~$ traceroute6 2001:1af8:4400:a00e:1::1 traceroute to 2001:1af8:4400:a00e:1::1 (2001:1af8:4400:a00e:1::1), 30 hops max, 80 byte packets 1 2001:4890:c08:2::1 (2001:4890:c08:2::1) 6.832 ms 6.822 ms 6.790 ms 2 2001:4890:4:1b::2 (2001:4890:4:1b::2) 68.401 ms 68.421 ms 68.439 ms 3 2001:4890:4:1c::1 (2001:4890:4:1c::1) 153.370 ms 153.373 ms 153.444 ms 4 2001:4890:4:1c::1 (2001:4890:4:1c::1) 153.488 ms !F-0 153.638 ms !F-0 153.476 ms !F-0 _______________________________________________ pool mailing list [email protected] http://lists.ntp.org/listinfo/pool
