That's not congestion at the IX, that's HE.net routing traffic back to you (and me) via California. They're prefering my Vocus transit in SJC over the shorter AS path in Sydney.

Try doing a traceroute from their Sydney router on http://lg.he.net

Matt.

On 16/05/2018 10:46 AM, Edwin Groothuis wrote:
From Internode there seems to be congestion at the exchange point:

traceroute to tserv1.syd1.he.net <http://tserv1.syd1.he.net> (216.218.142.50), 64 hops max, 52 byte packets
 1  192.168.129.1 (192.168.129.1)  2.043 ms  2.373 ms  1.887 ms
 2 lo0.bras1.syd2.on.ii.net <http://lo0.bras1.syd2.on.ii.net> (150.101.32.61)  30.334 ms  31.081 ms  30.884 ms  3 be1.cr2.syd2.on.ii.net <http://be1.cr2.syd2.on.ii.net> (150.101.197.80)  54.951 ms  30.238 ms  31.129 ms  4 be51.cr2.syd7.on.ii.net <http://be51.cr2.syd7.on.ii.net> (150.101.40.242)  31.842 ms  31.422 ms  33.403 ms  5 ae6.cr2.syd4.on.ii.net <http://ae6.cr2.syd4.on.ii.net> (150.101.41.124)  31.041 ms  31.468 ms  32.887 ms  6 6939.syd.equinix.com <http://6939.syd.equinix.com> (45.127.173.24)  181.514 ms  181.298 ms  179.775 ms  7 tserv1.syd1.he.net <http://tserv1.syd1.he.net> (216.218.142.50)  180.816 ms  180.744 ms  181.266 ms

(I don't even understand why I need to investigate this as it's more than 10 years ago since I got native IPv6 on my ADSL and can't think of a reason besides that nobody providing NBN/FTTC will support it. Welcome back to early 2000's)

Edwin


On 16 May 2018 at 07:22, Cameron Murray <[email protected] <mailto:[email protected]>> wrote:

    Looks to be in Sydney for us:

    Tracing route to tserv1.syd1.he.net <http://tserv1.syd1.he.net>
    [216.218.142.50]
    over a maximum of 30 hops:

      3     1 ms     2 ms     1 ms  103.17.253.22
      4    13 ms    14 ms    13 ms as6939.nsw.ix.asn.au
    <http://as6939.nsw.ix.asn.au> [218.100.52.249]
      5    13 ms    13 ms    13 ms tserv1.syd1.he.net
    <http://tserv1.syd1.he.net> [216.218.142.50]

    Trace complete.

    On Tue, May 15, 2018 at 6:19 PM, andrew khoo
    <[email protected] <mailto:[email protected]>> wrote:

        to confirm, i am tracing from a box in a he.net
        <http://he.net> colo and it's going to australia:

        as7175-sjc>traceroute 216.218.142.50
        traceroute to 216.218.142.50 (216.218.142.50), 30 hops max, 60
        byte packets
         1 v1882.core2.fmt2.he.net <http://v1882.core2.fmt2.he.net>
        (65.19.156.125)  10.773 ms  0.234 ms  0.374 ms
         2 10ge6-16.core4.fmt2.he.net
        <http://10ge6-16.core4.fmt2.he.net> (184.105.213.189)  0.768
        ms  0.849 ms  0.319 ms
         3 100ge10-1.core1.sjc1.he.net
        <http://100ge10-1.core1.sjc1.he.net> (184.105.80.193)  0.552
        ms  0.594 ms  0.652 ms
         4 10ge2-1.core1.syd1.he.net
        <http://10ge2-1.core1.syd1.he.net> (184.105.222.86)  146.522
        ms  146.509 ms 146.462 ms
         5 tserv1.syd1.he.net <http://tserv1.syd1.he.net>
        (216.218.142.50)  146.303 ms  146.311 ms 146.285 ms




        On Tue, May 15, 2018 at 6:13 PM, Gavin Tweedie
        <[email protected]
        <mailto:[email protected]>> wrote:

            TPG buys transit from HE /in the US/ so routing via the US
            is expected, twice.
            TPG SYD -> TPG SJC -> HE SJC -> HE SYD -> HE SJC -> TPG
            SJC -> TPG SYD

            Yours looks like it's staying in Sydney on the way there,
            and coming back via the US I agree.. You don't have any
            indications in your traceroute what your source ASN is,
            but I'm going to guess that your ASN buys transit from HE,
            or someone who buys from HE would that be right?

            Hurricane do have a LG @ https://lg.he.net/ which might
            help you work out the reverse path.

            Gav

            On Tue, 15 May 2018 at 16:07, Christopher Hawker
            <[email protected] <mailto:[email protected]>> wrote:

                Hi All,


                I've seen that HE now supposedly run a IPv6 Tunnel
                Server in Sydney, although I have doubts as to it's
                location after a few traceroutes. This is from SY4 via
                NSW-IX:


                default.master@cr1:~$ traceroute 216.218.142.50
                traceroute to 216.218.142.50 (216.218.142.50), 30 hops
                max, 38 byte packets
                 1  upstream.carrier (192.0.2.254)  0.183 ms 0.123 ms 
                0.108 ms
                 2 as6939.nsw.ix.asn.au <http://as6939.nsw.ix.asn.au>
                (218.100.52.249)  150.979 ms  152.368 ms  150.857 ms
                 3 tserv1.syd1.he.net <http://tserv1.syd1.he.net>
                (216.218.142.50)  150.751 ms  150.839 ms  150.755 ms

                However, this trace is from an AAPT NBN HFC service:


                C:\Users\default.master>tracert 216.218.142.50
                Tracing route to tserv1.syd1.he.net
                <http://tserv1.syd1.he.net> [216.218.142.50]
                over a maximum of 30 hops:
                1     2 ms     1 ms  1 ms  172.16.10.254
                  2     *        *   *     Request timed out.
                  3    12 ms    10 ms    9 ms
                po41.sglebbrdr11.aapt.net.au
                <http://po41.sglebbrdr11.aapt.net.au> [202.10.14.198]
                  4     8 ms     9 ms    9 ms
                syd-gls-har-wgw1-be-30.tpgi.com.au
                <http://syd-gls-har-wgw1-be-30.tpgi.com.au>
                [203.219.107.197]
                  5    10 ms    15 ms   14 ms
                203.29.134-67.tpgi.com.au
                <http://203.29.134-67.tpgi.com.au> [203.29.134.67]
                  6   173 ms   170 ms  162 ms
                100ge13-1.core1.sjc1.he.net
                <http://100ge13-1.core1.sjc1.he.net> [216.218.139.233]
                  7   308 ms   311 ms  308 ms
                10ge2-1.core1.syd1.he.net
                <http://10ge2-1.core1.syd1.he.net> [184.105.222.86]
                  8   309 ms   309 ms  309 ms tserv1.syd1.he.net
                <http://tserv1.syd1.he.net> [216.218.142.50]
                Trace complete.

                The first trace looks like it is going over to the US
                (based on the latency), however the second looks like
                it is going to the US and back. Anyone else seen
                anything like this? If there are any HE techs on list,
                I'd appreciate it if they could please contact me
                off-list.


                CH.

                _______________________________________________
                AusNOG mailing list
                [email protected] <mailto:[email protected]>
                http://lists.ausnog.net/mailman/listinfo/ausnog
                <http://lists.ausnog.net/mailman/listinfo/ausnog>



-- Global Interconnection Director
            Megaport <https://www.megaport.com>
            +61 498 498 458

            _______________________________________________
            AusNOG mailing list
            [email protected] <mailto:[email protected]>
            http://lists.ausnog.net/mailman/listinfo/ausnog
            <http://lists.ausnog.net/mailman/listinfo/ausnog>



        _______________________________________________
        AusNOG mailing list
        [email protected] <mailto:[email protected]>
        http://lists.ausnog.net/mailman/listinfo/ausnog
        <http://lists.ausnog.net/mailman/listinfo/ausnog>



    _______________________________________________
    AusNOG mailing list
    [email protected] <mailto:[email protected]>
    http://lists.ausnog.net/mailman/listinfo/ausnog
    <http://lists.ausnog.net/mailman/listinfo/ausnog>




_______________________________________________
AusNOG mailing list
[email protected]
http://lists.ausnog.net/mailman/listinfo/ausnog


_______________________________________________
AusNOG mailing list
[email protected]
http://lists.ausnog.net/mailman/listinfo/ausnog

Reply via email to