[04:40] *** novae has quit IRC (Ping timeout: 272 seconds)
[07:11] *** midhun_ has joined #arpnetworks
[07:12] <midhun_> Can someone explain me how does TCP manages timed events effectively?
[07:16] *** midhun_ has quit IRC (Quit: Leaving)
[07:17] *** midhun__ has joined #arpnetworks
[07:18] *** midhun__ has quit IRC (Client Quit)
[08:02] *** tabthorpe has joined #arpnetworks
[08:02] *** tabthorpe has quit IRC (Changing host)
[08:02] *** tabthorpe has joined #arpnetworks
[12:42] <mnathani> @smokeping
[12:42] <BryceBot> https://smokeping.cobryce.com/
[12:42] <brycec> btw the Time Warner connection was out for an hour earlier
[12:43] <brycec> (hence the giant gap)
[12:43] <mnathani> brycec: you dont measure latency to each kvr host do you?
[12:44] <brycec> no
[12:44] <brycec> just from kvr07
[12:44] <brycec> I don't have IP's for each host
[12:44] <brycec> I guess I do come to think of it
[12:44] <brycec> but the value is questionable
[12:44] <mnathani> kvr[02-30].arpnetworks.com
[12:44] <mnathani> externally it would help
[12:45] <mnathani> from your slave nodes
[12:45] <brycec> How so? A connection to ARp is a connection to ARP
[12:45] <mnathani> could pinpoint DDOS against specific hosts
[12:45] <brycec> More like high load on a given host
[12:45] <brycec> a ddos would affect all of arp
[12:45] <brycec> at least all of a given transit
[16:36] <mercutio> it's better to test locally.
[17:42] <davantalus> up_the_irons: http://cdn.meme.am/instances/500x/55274185.jpg
[17:43] <up_the_irons> davantalus: LOOOOOOOOL
[17:44] <up_the_irons> mnathani: CaZe : there was a spike of traffic to kvr30 last night, yeah
[18:20] <mercutio> up_the_irons: is level3 inbound planned for this week now?
[18:21] <mercutio> i noticed that outbound seemed to be going via level3 rather than any2ix right now.
[18:25] <up_the_irons> mercutio: there hasn't been a change to outgoing.  i could possible plan level3 inbound this week...
[18:25] <mercutio> up_the_irons: curious.  it was taking that path when any2ix had issues before, but not for long.  but smokeping suggests it's been happening a couple of days or such.
[18:26] <mercutio> oh it's not all any2ix
[18:26] <mercutio> has any2ix traffic dropped in the last couple of days?
[18:26] <mercutio> facebook still seems to go over any2ix
[18:28] <brycec> facebook goes over ntt for me (from kvr07)
[18:28] <mercutio> curious
[18:28] <brycec> Also, ipv6
[18:29] <brycec> :P
[18:29] <mercutio> www.facebook.com is an alias for star.c10r.facebook.com.
[18:29] <mercutio> star.c10r.facebook.com has address 69.171.230.5
[18:29] <mercutio> oh i'm going over ipv4
[18:29] <up_the_irons> wut..  shouldn't be any difference between anyone's routing ;)
[18:29] <brycec> up_the_irons: v4 vs v6
[18:29] <up_the_irons> kvrXX all go thru same path
[18:29] <up_the_irons> ah
[18:29] <up_the_irons> mercutio: i don't see a difference in any2 traffic the last couple days.  but there does seem to have been a drop from like, a month ago
[18:30] <mercutio> curious.
[18:30] <mercutio> a looking glass would be nice :)
[18:30] <up_the_irons> set one up for me
[18:30] <brycec> Odd, v4 goes over coresite
[18:30] <up_the_irons> ;)
[18:30] <brycec> both v6 and v4 http://sprunge.us/OBbL
[18:30] <mercutio> up_the_irons: sure.
[18:30] <up_the_irons> i'll give you the session peer
[18:31] <mercutio> with bgp multihop?
[18:31] <up_the_irons> tell me if there is anything special i need to do on my end for the session (Cisco config)
[18:31] <brycec> Yay ARP's own LG
[18:31] <mercutio> can it mirror somehow
[18:31] <mercutio> will look into mirroring
[18:32] <up_the_irons> let me see... how did i do this for zeit and brycec ...  i think i put a VM under my own account and handed it off..
[18:32] <mercutio> i dunno if there's some easy way to do that so can find out what the different providers thing for routes, rather than what you think for route.
[18:32] <mercutio> but could get some idea just from having peers to all 3 of routers i suppose
[18:38] <mercutio> apparently there's something called add-path
[18:39] <mercutio> http://www.nanog.org/meetings/nanog48/presentations/Tuesday/Ward_AddPath_N48.pdf
[18:39] <mercutio> no idea if your router would support it though
[18:40] <mercutio> or what looking glass would do with it
[18:40] <mercutio> http://www.cisco.com/c/en/us/td/docs/ios-xml/ios/iproute_bgp/configuration/xe-3s/irg-xe-3s-book/irg-additional-paths.html#GUID-4EB13F76-7C14-4B74-AFE0-66BF07976BD5
[18:40] <mercutio> it's pretty cool anyway
[18:40] <brycec> I looked into what it would take to setup an LG for ARP and this was the most solid http://www.ogris.de/howtos/openbsd-looking-glass.html
[18:40] <brycec> (in the past, I did this)
[18:40] <mercutio> yeh that was my plan bryce.
[18:41] <brycec> So that's my $.02 :D
[18:41] <mercutio> but bird has bgp add-path support in beta
[18:41] <mercutio> i wonder if something else supports it
[18:41] <brycec> (I'm no expert on BGP so I wouldn't be a very good candidate at adminning it)
[18:41] <mercutio> i suppose can look into that later anyway.
[18:41] <mercutio> basically add-path means if there are routes from level3 and ntt on the same router, they can both be forwarded along, so the route server in theory can show both paths and the as lengths.
[18:42] <mercutio> err the looking glass can show.
[18:42] <brycec> ah
[18:42] <mercutio> but a simple looking glass would still be way better than none
[18:44] <mercutio> apparently bird has a beta looking glass and bgp add-path in beta.
[18:45] <mercutio> not sure if the bird looking glass works with add-path though
[19:05] <mercutio> it seems there aren't over 512k routes yet.
[19:06] <mercutio> http://bgp.potaroo.net/as2.0/bgp-active.html
[19:06] <mercutio> getting close though
[19:06] <up_the_irons> we are getting very close to 512k routes
[19:07] <mercutio> yeah, there was that incident a while back when it went over temporarily.
[19:07] <up_the_irons> s7.lax>sh ip cef sum
[19:07] <up_the_irons> IPv4 CEF is enabled for distributed and running
[19:07] <up_the_irons> VRF Default
[19:07] <up_the_irons>  506853 prefixes (506853/0 fwd/non-fwd)
[19:07] <up_the_irons>  Table id 0x0
[19:07] <up_the_irons>  Database epoch:        3 (506853 entries at this epoch)
[19:07] <up_the_irons> 506K for me :)
[19:07] <up_the_irons> yes, we all blame verizon
[19:07] <mercutio> i was looking at how much ram openbgpd uses.
[19:08] <mercutio> and atm looking glass is probably fine on 512mb of ram, but i dunno how long that'd last.
[19:08] <mercutio> taht's with 1.25 million routes.
[19:09] <mercutio> well two full tables plus one partial.
[19:10] <mercutio> i imagine there's some separte ram usage for used routes vs known routes et.c
[19:48] *** mike-bur1 has joined #arpnetworks
[19:48] *** ChanServ sets mode: +o mike-bur1
[19:49] *** pjs_ has joined #arpnetworks
[19:50] *** m0unds_ has joined #arpnetworks
[19:50] *** acf__ has joined #arpnetworks
[19:50] *** notion has joined #arpnetworks
[19:50] *** SpaceDum1 has joined #arpnetworks
[19:54] *** BryceBot has quit IRC (*.net *.split)
[19:54] *** up_the_irons has quit IRC (*.net *.split)
[19:54] *** atmark has quit IRC (*.net *.split)
[19:54] *** dne has quit IRC (*.net *.split)
[19:54] *** vissborg has quit IRC (*.net *.split)
[19:54] *** pjs has quit IRC (*.net *.split)
[19:54] *** mike-burns has quit IRC (*.net *.split)
[19:54] *** m0unds__ has quit IRC (*.net *.split)
[19:54] *** SpaceDump has quit IRC (*.net *.split)
[19:54] *** notion_ has quit IRC (*.net *.split)
[19:54] *** phlux has quit IRC (*.net *.split)
[19:54] *** acf_ has quit IRC (*.net *.split)
[19:56] *** BryceBot has joined #arpnetworks
[19:56] *** BryceBot is now known as Guest56255
[20:00] *** atmark has joined #arpnetworks
[20:01] *** atmark is now known as Guest86251
[20:02] *** dne has joined #arpnetworks
[20:03] *** up_the_irons has joined #arpnetworks
[20:03] *** ChanServ sets mode: +o up_the_irons
[20:04] *** phlux has joined #arpnetworks
[20:06] *** vissborg has joined #arpnetworks
[20:30] *** reardencode has quit IRC (Ping timeout: 260 seconds)
[20:30] *** reardencode has joined #arpnetworks
[20:35] *** reardencode has quit IRC (Ping timeout: 260 seconds)
[20:35] <mnathani> +1 for the ARP Looking Glass. Look forward to it. mercutio & up_the_irons
[20:36] *** reardencode has joined #arpnetworks
[20:36] *** reardencode has quit IRC (Remote host closed the connection)
[20:46] *** mike-bur1 is now known as mike-burns
[20:57] *** Guest56255 is now known as BryceBot
[20:57] *** BryceBot is now known as Guest17668
[21:17] *** toeshred has quit IRC (Ping timeout: 250 seconds)
[21:27] *** acf__ is now known as acf_
[21:46] *** toeshred has joined #arpnetworks
[21:59] *** Guest17668 is now known as BryceBot
[21:59] *** BryceBot is now known as Guest1914
[22:00] <brycec> If the host were ram-starved, swapping wouldn't be the end of the world... An LG doesn't need stellar performance I imagine.
[22:06] *** Guest1914 has quit IRC (Changing host)
[22:06] *** Guest1914 has joined #arpnetworks
[22:07] *** Guest1914 is now known as BryceBot
[22:40] *** Guest86251 has quit IRC (*.net *.split)
[22:40] *** awyeah has quit IRC (*.net *.split)
[22:40] *** dj_goku has quit IRC (*.net *.split)
[22:40] *** sga0_ has quit IRC (*.net *.split)
[22:40] *** mhoran has quit IRC (*.net *.split)
[22:40] *** tooth has quit IRC (*.net *.split)
[22:46] *** Guest86251 has joined #arpnetworks
[22:46] *** awyeah has joined #arpnetworks
[22:46] *** dj_goku has joined #arpnetworks
[22:46] *** sga0_ has joined #arpnetworks
[22:46] *** mhoran has joined #arpnetworks
[22:46] *** tooth has joined #arpnetworks
[22:46] *** wilhelm.freenode.net sets mode: +o mhoran
[23:47] <mercutio> hmm openbsd 5.6 is out in a couple of weeks