Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In with OpenID
Advertise on LowEndTalk.com

In this Discussion

Routing around the hurricane?

Routing around the hurricane?

pubcrawlerpubcrawler Banned
edited October 2012 in General

Anyone noticing routing weird routes in / around areas that go through New York?

Look at this from Buffalo: Buffalo to Cleveland to Chicago to Washington, DC to Chicago to New York

3: buf-b1-link.telia.net 0.682ms asymm 5 4: cle-b1-link.telia.net 6.270ms asymm 6 5: chi-bb1-link.telia.net 13.348ms asymm 7 6: xe-9-2-0.edge1.washington4.level3.net 13.155ms 7: ae-31-51.ebr1.Chicago1.Level3.net 19.985ms asymm 5 8: ae-2-2.ebr2.NewYork2.Level3.net 33.383ms asymm 7 9: ae-6-6.ebr2.NewYork1.Level3.net 21.365ms asymm 6 10: ae-62-62.csw1.NewYork1.Level3.net 22.924ms asymm 7 11: ae-4-90.edge4.NewYork1.Level3.net 21.359ms asymm 9 12: NEW-YORK-UN.edge4.NewYork1.Level3.net 21.970ms asymm 9

Take it Telia has equipment/facility down in Manhattan? This isn't the normal route we see to NYC.

Comments

  • Routing to our node in NY (Telx, DC #1)

    Equinix Chicago core1.chi1.he.net> traceroute 199.38.181.XXX

    Tracing the route to IP node (199.38.181.XXX) from 1 to 30 hops 1 11 ms 3 ms 1 ms 69.31.111.109 2 <1 ms <1 ms <1 ms 69.31.111.137 3 23 ms 24 ms 24 ms 69.22.142.33 4 43 ms 33 ms 17 ms 69.31.31.154 5 31 ms 24 ms 25 ms 69.22.142.38 6 25 ms 24 ms 24 ms 69.31.34.226 7 25 ms 25 ms 25 ms 199.38.181.XXX# Entry cached for another 58 seconds.

    Equinix Dallas core1.dal1.he.net> traceroute 199.38.181.XXX

    Tracing the route to IP node (199.38.181.XXX) from 1 to 30 hops 1 * * * ? 2 127 ms 124 ms 125 ms 69.31.63.125 3 50 ms 49 ms 49 ms 69.22.142.113 4 20 ms 24 ms 25 ms 69.22.142.118 5 34 ms 41 ms 33 ms 69.22.142.106 6 39 ms 49 ms 56 ms 69.22.142.142 7 44 ms 48 ms 49 ms 69.31.95.174 8 50 ms 49 ms 49 ms 69.31.34.226 9 49 ms 49 ms 50 ms 199.38.181.XXX# Entry cached for another 51 seconds.

    Equinix Los Angeles core1.lax1.he.net> traceroute 199.38.181.XXX

    Tracing the route to IP node (199.38.181.XXX) from 1 to 30 hops 1 26 ms 16 ms 19 ms 206.223.123.61 2 72 ms 74 ms 75 ms 69.22.142.115 3 71 ms 74 ms 75 ms 69.22.142.38 4 76 ms 73 ms 75 ms 69.31.34.222 5 80 ms 74 ms 75 ms 199.38.181.XXX# Entry cached for another 59 seconds.

  • Given how many DC's in NYC are down or having major issues i'm not surprised :(

    Francisco

    BuyVM - OpenVZ & KVM Based / TUN, PPTP, FUSE, SIT & GRE Enabled! / Stallion Control Panel
  • I saw some packet loss on he.net between USA and Europe, but this might be normal for them.

  • @Francisco said: Given how many DC's in NYC are down or having major issues i'm not surprised :(

    only the bad ones...

    Opinions/Posts are to be assumed my own/personal and not company related unless obvious
    Working @ EDIS and owning some others (and/or parts of) | Available for consulting | http://as198412.net | https://william.si

  • WebNX

    3 24 ms 24 ms 23 ms 10.245.203.225 4 34 ms 30 ms 42 ms te0-6-0-2.er11.thlon.ov.easynet.net [89.200.135.31] 5 28 ms 27 ms 36 ms 10gigabitethernet1-1.core1.lon1.he.net [195.66.224.21] 6 43 ms 35 ms 40 ms 10gigabitethernet2-4.core1.par2.he.net [72.52.92.42] 7 119 ms 118 ms 128 ms 10gigabitethernet7-1.core1.ash1.he.net [184.105.213.93] 8 134 ms 145 ms * 10gigabitethernet5-4.core1.chi1.he.net [184.105.213.245] 9 164 ms 162 ms 171 ms 10gigabitethernet3-2.core1.den1.he.net [184.105.213.86] 10 * 196 ms 189 ms 10gigabitethernet1-4.core1.lax2.he.net [72.52.92.38] 11 190 ms 190 ms 188 ms 216.218.213.250 12 190 ms 190 ms * 100-42-223-150.static.webnx.com [100.42.223.150]

    13 191 ms 189 ms 188 ms 100-42-223-75.static.webnx.com [100.42.223.75] 14 190 ms * 189 ms 142-4-49-108.static.webnx.com [142.4.49.108]

    Trace complete.

    https://nodedeploy.com | Premium VPS Solutions | Managed

  • JacobJacob Member
    edited October 2012

    The DC that our master is in just happens to be down. #SandyYouSuck

  • ZenZen Member

    @Jacob said: The DC that our master is in just happens to be down. #SandyYouSuck

    Re-re-re-re-re-re-redundancy!

    I work for Nodisto.

  • We're currently getting pings that are about 100ms longer than normal in Europe. We're colocating in San Jose, California, but I assume that the longer pings are because of Hurricane Sandy, right?

    http://www.maxshosting.com/
    Unofficial community troll. AMA!

  • JarJar Member

    Can I blame the hurricane for 30mbit speed from Denver to Dallas? Because I'd like to.

  • Appears to be back to normal from Buffalo to NYC. Wasn't around here this afternoon to keep an eye on things:

    3: buf-b1-link.telia.net 3.296ms asymm 5 4: nyk-bb1-link.telia.net 9.961ms asymm 6 5: nyk-b5-link.telia.net 10.052ms asymm 6 6: level3-ic-120265-nyk-b5.c.telia.net 10.451ms asymm 9 7: ae-3-80.edge4.NewYork1.Level3.net 12.438ms asymm 9 8: NEW-YORK-UN.edge4.NewYork1.Level3.net 14.354ms asymm 9 9: NYUGWA-VL10.NET.NYU.EDU 15.836ms asymm 12 10: WSQDCGWA-VL901.NET.NYU.EDU 22.483ms asymm 13

    Appears Telia was probably fine but that Level3's handoff in NYC was down earlier.

  • @Zen said: Re-re-re-re-re-re-redundancy

    The majority of the DCs that are down had redundancy...

    Power went of course went out and they had to go to generator power. Problem is that the fuel storage was flooded out and fuel could not be pumped up to the generators. This resulted in about 6 hours of fuel for the generators to run, before they ran out and had to shutdown the DC totally. Either that or their generators were flooded out, because they located in the basement...

    So, they do have redundancy. Although pretty stupid how they put everything for it in the basement or ground level, so the equipement won't be useful in a high water situation.

  • @NateN34 said: . Although pretty stupid how they put everything for it in the basement or ground level, so the equipement won't be useful in a high water situation.

    That is State Code

    Thanked by 1fly
  • flyfly Member

    @Spencer said: That is State Code

    the truth. everyone and their grandmother's generators were flooded out

  • pubcrawlerpubcrawler Banned
    edited October 2012

    State codes are pretty moronic (about generators being subterranean) and need changed in NY after this.

    Glad I didn't colo gear in NYC. Always deemed it a very big risky place.

    Right now NYC is having a once every decade issue that impacts infrastructure.

  • @pubcrawler said: Right now NYC is having a once every decade issue that impacts infrastructure.

    What was the last one? 9/11?

  • Yeppers @Nick :)

    Bound to see some issues in NYC infrastructure from this flooding far into the future.

    Media already bemoaning the outdated infrastructure and lack of maintaining Manhattan.

  • Updates from our DC in NY (Telx NYC1)

    NYC1 - 14:25 10/30 NOW ON GENERATOR POWER - THE BUILDING LOST MUNICIPAL POWER AT 19:15 10/29, AT CURRENT FUEL CONSUMPTION RATE WE HAVE AMPLE FUEL UNTIL 10/31, A NEW FUEL DELIVERY ARRIVED AT APPROX 14:00, FUEL RESERVES ARE STAGED ON THE ISLAND. WE HAVE COORDINATED DAILY FUEL DELIVERIES.

    So as you can see not all DCs went down

  • ZenZen Member

    @NateN34 said: The majority of the DCs that are down had redundancy...

    Power went of course went out and they had to go to generator power. Problem is that the fuel storage was flooded out and fuel could not be pumped up to the generators. This resulted in about 6 hours of fuel for the generators to run, before they ran out and had to shutdown the DC totally. Either that or their generators were flooded out, because they located in the basement...

    So, they do have redundancy. Although pretty stupid how they put everything for it in the basement or ground level, so the equipement won't be useful in a high water situation.

    I'm talking about user-level redundancy, not provider-level. Jacob should have a redundant master set up and not a single point of failure. You shouldn't always be relying on one PoP.

    I work for Nodisto.

  • I'd actually like to see the amount of providers that have a secondary master setup, prior to me posting this... not many.

    @Zen said: Jacob should have a redundant master set up

  • ZenZen Member

    @Jacob said: I'd actually like to see the amount of providers that have a secondary master setup, prior to me posting this... not many.

    Not picking on you, and I agree with that.. just saying that it isn't all that hard, everyone seems to be lazy though when it comes to the only thing that makes them $

    I work for Nodisto.

Sign In or Register to comment.