Back to Home / #linode / 2004 / 05 / Prev Day | Next Day
#linode IRC Logs for 2004-05-19

---Logopened Wed May 19 00:00:55 2004
00:14mikegrb <- lnx bbc on screen
00:36--- <<-- sollaires [] has quit (Quit: Quit)
00:50--- Netsplit <-> quits: sadjester, david, ironie, Tenkawa, rko
00:50CULTUS \xABwSk\xB3\xBB NetSplit Detected jupiter \x95 nucleon \xABW-\xE5-\xAE-\xEB-S-k\xB3\xBB
00:51--- Netsplit over, joins: Tenkawa
01:03mikegrb dongs
01:03* caker needs to find the usb cable for his digital camera
01:03@caker I have to source code to g-mail
01:03@caker from 1988
01:06mikegrb heh
01:06@caker May 19 2004 01:09:14: Security Trap for via cookie
01:06@caker muhahahaaa
01:07mikegrb almost 1GB/month bandwidth for
01:07mikegrb sighup: dns
01:07sighup mikegrb: is located in Brazil (Comite Gestor da Internet no Brasil)
01:08@caker Starting nmap 3.48
01:08@caker Adding open port 5000/tcp
01:08@caker ^ open proxy
01:09@caker but now he's cookieified
01:09@caker <meta http-equiv="refresh" content="0;url=telnet://">
01:09@caker ^ to get real IP
01:10@caker I also have a java applet, but it doesn't work as well
01:10mikegrb hah
01:11mikegrb bofh caker++
01:11mikegrb 668 hits from yahoo search results
01:23--- ---> sadjester [] has joined #linode
01:26Efudd man
01:26Efudd bleh
01:28@caker mmm ... four port ethernet pci cards
01:29* Efudd has a buncha those
01:29@caker check this thing out
01:29Efudd i'm using this "opensource product" called macroquest2....
01:29Efudd it's an API for customizing everquest's client
01:29Efudd they reverse engineered the memory space
01:29@caker to do what?
01:29Efudd wrote an API around it
01:29Efudd using an M$ API called "detours"
01:29Efudd think of it as adding jmp's
01:30Efudd well, lots of stuff
01:30@caker ok
01:30Efudd just adding to the client
01:30Efudd anyway, there is core functionality..
01:30Efudd that changes FUCKING DAILY
01:30Efudd so i start making modifications off a codebase...
01:30Efudd and find something core broken.
01:30Efudd i get bitched at to "get a new zip"
01:30Efudd which helps very little with customizations :)
01:30Efudd opensource has serious failures :)
01:31Efudd it's an AMAZING piece of coding
01:31Efudd but holy fuck the author is an egotistic bastard :)
01:31Efudd and that's being polite.
01:31Efudd EndOfRant && /meclickxlink
01:31Efudd eh
01:31Efudd these are just 4port cards
01:31Efudd relatively new
01:32Efudd MDI/X support is kinda badass tho
01:32Efudd lots of lappies come with that now
01:33@caker the "main" product looks like a complete system
01:34Efudd ah
01:34Efudd embedded, kinda.
01:34@caker yeah .. all in one deal
01:34Efudd newer cpu's than most "standard" embeddeds
01:34@caker 266 Mhz NSC SC1100 system on a chip CPU (Pentium MMX architecture)
01:34Efudd what is cost on them?
01:35@caker looks cheap ..
01:35@caker $200-$362 for a "kit"
01:35* caker uses their calculator thing
01:36@caker anyway .. ever build/use a Linux router running bgp?
01:37@caker or suggest multi-port pci-x ethernet cards
01:38Efudd the author/maintainer of mq2 has written a decent system. but only /he/ knows it inside and out. I'm a professional code reader... by trade i bugfind.... i find his code fucking obtuse.
01:38Efudd nope
01:38Efudd never bothered to
01:38@caker it was probably meant to WorkForHim<tm>, not be well written
01:38Newsome caker: I've used the Intel e1000 dual-port cards before, and they've worked really well for me
01:38Efudd i have friends who have used the soekris
01:38Efudd basically, he refactors it daily
01:38Efudd which fucks with shit hard
01:38Efudd it's got a macro language in it that i hate.
01:39@caker Newsome: yeah, I'm on there page as I write this actually .. the hosts are all/mostly e1000s
01:39Efudd so i'm learning it to write c++ plugins for it instead
01:39@caker s/there/their/ ..
01:40Newsome caker, the Linux drivers are pretty good, and when I got the cards, they came with the 1/2 size bracket, too.
01:40--- Netsplit <-> quits: fo0bar, cout, SupaDongzu, Jayce^, Newsome, sunny, Efudd
01:40--- Netsplit <-> quits: ElectricElf, Sh8d0w, inkblot, Tenkawa, NeXTer, smohekey__, artifex, funkycow, AndyHat, thelsdj, (+9 more, use /NETSPLIT to show all of them)
01:41--- Netsplit over, joins: Sh8d0w, smokie, mesozoic, sadjester, thelsdj, AndyHat, NeXTer, Tenkawa, UML_ChanLog, fo0barf_ (+6 more)
01:41--- ---> Newsome [] has joined #linode
01:41--- Netsplit over, joins: artifex, caker, SupaDongzu, sunny, Efudd, Jayce^, cout, fo0bar
01:41Newsome Argh.
01:42caker wtf was that?
01:42--- Netsplit over, joins: unixvps
01:42caker who split?
01:42Newsome good question
01:42--- User: *** artifex is now known as Guest431
01:57Newsome caker: anyway, the Intel PRO/1000 MT dual-port worked well for me. $177 at Quads are a bit expensive (looks like $425+), but probably work just as well. Pretty decent Linux drivers.
01:59Newsome Also, I've got some Broadcom NetXtreme BCM5704's. Those have worked well, too.
02:00Newsome tg3 driver under Linux
02:02caker SuperMicro 6013P-i: 1 x 64-bit 133/100MHz Full-height, Full-length PCI-X Expansion Slot
02:02caker sweet ..
02:02caker also: 1 x 64-bit 100MHz Low-profile PCI-X eExpansion Slot
02:02caker so either/or/both ..
02:02caker have to think about building a router box with only the built-in and 1.5 slots
02:03caker (actually, would build two router boxes)
02:05Newsome 1U?
02:06caker not sure yet
02:12caker hah
02:12caker nova popper[23992]: (null) at ( -ERR Unknown command: "ls".
02:12caker nova popper[23992]: (null) at ( -ERR Unknown command: "exit".
02:12caker sucker
02:14fo0bar caker: the question begs to be asked... why do you have a POP3 server running? :)
02:15caker fo0bar: as opposed to?
02:19fo0bar imap makes the world go round, yo.
02:19fo0bar preferably imap-tls
02:19* caker wants
02:20* caker needs .. just hasn't had time
02:20fo0bar :)
02:22fo0bar I have a nice setup at work... if you send email from one office to another, and then the recipient reads, at no point is it not travelling over an SSL connection
02:24fo0bar send mail by smtps with "SMTP AUTH", the mail servers talk to each other via smtp-tls, and the reipient reads it over an imap-ssl connection
02:54--- ---> rko [] has joined #linode
04:03--- Netsplit <-> quits: unixvps, rko
04:03--- Netsplit <-> quits: ElectricElf, Sh8d0w, inkblot, Tenkawa, NeXTer, smohekey__, funkycow, CULTUS, UML_ChanLog, mesozoic, (+3 more, use /NETSPLIT to show all of them)
04:03--- Netsplit over, joins: Sh8d0w, smokie, rko, unixvps, mesozoic, NeXTer, Tenkawa, UML_ChanLog, fo0barf_, CULTUS (+5 more)
05:24--- <<-- Newsome [] has quit (Quit: Leaving)
06:56--- <<-- funkycow [] has quit (Remote host closed the connection)
07:57--- <<-- Sh8d0w [] has quit (Quit: ...)
08:28mikegrb .Downloaded 46.5MB in 16s (2746kB/s).
09:15--- Channel: mode/#linode [+o caker] by ChanServ
09:16--- ---> Redgore [] has joined #linode
09:27--- ---> wferrel [] has joined #linode
09:32--- <<-- emcnabb [] has quit (Ping timeout: 480 seconds)
09:49--- ---> Newsome [] has joined #linode
10:02--- ---> Oxnard [] has joined #linode
10:51--- ---> ashen [] has joined #linode
10:52* ashen waves
10:53rko argh
10:53ashen is there any way to tell *why* a slow linode is running so slowly?
10:53ashen by "slowly" I mean "too slowly to connect to it via ssh and issue commands to it"
10:53rko why did my max_tokens drop to 20000 since yesterday?
10:54rko you're most likely overloading it
10:54ashen hmmmmmm
10:57--- <<-- ashen [] has quit (Quit: ->Ashen)
11:18--- ---> Redgore_ [] has joined #linode
11:18--- <<-- Redgore [] has quit (Read error: Connection reset by peer)
11:19--- ---> Guest43 [] has joined #linode
11:19--- <<-- adamg [] has quit (Read error: Connection reset by peer)
11:20--- User: *** Guest43 is now known as adamg
11:23--- User: *** Redgore_ is now known as Redgore
11:32rko caker?
11:32Oxnard you are deaming...check back again after noon.
11:32Oxnard :)
11:33--- ---> ashen [] has joined #linode
11:33rko :-)
11:33Oxnard sighup: seen caker
11:33sighup caker was last seen on #linode 9 hours, 13 minutes and 7 seconds ago, saying: needs .. just hasn't had time [1084947605]
11:34rko that's more than needed for a sleep
11:35rko you can sleep twice in that time
11:39Oxnard ...but can you sober up first?
11:39ashen heh
11:39rko why bother? just start a new day with a cold, smoothy, frotsyone
11:40Oxnard always an option.
11:40rko options are can or bottle
11:54ashen mmm
12:10--- ---> sjansen [] has joined #linode
12:14--- ---> jake [~jake@] has joined #linode
12:46--- ---> emcnabb [] has joined #linode
13:19--- ---> Redgore_ [] has joined #linode
13:19--- <<-- Redgore [] has quit (Read error: Connection reset by peer)
13:19--- User: *** Redgore_ is now known as Redgore
13:37--- <<-- emcnabb [] has quit (Ping timeout: 480 seconds)
13:53rko caker?
14:06ashen sleeping I think.
14:06ashen idle 11h
14:06rko yep, not noon yet
14:06ashen what's special about noon?
14:07adamg if it is important fill in a support ticket
14:08--- <--- adamg [] has left #linode ()
14:08--- ---> adamg [] has joined #linode
14:09rko I just wanted to ask about this io_tokens thing
14:09rko maybe you know?
14:09rko yesterday I had my max_tokens as 200000, now it's 20000
14:10rko and refill only 100, I have real pain sustaining services with that low resources
14:18rko ticket #5241 opened :-)
14:20--- <<-- sjansen [] has quit (Quit: I'll show you cruel!)
14:21@caker hello
14:21rko there he is
14:21@caker rko: try now?
14:22rko I've been grinding my teeth with this io_tokens business
14:22rko :-) much better
14:22rko how does that work?
14:22@caker :)
14:23rko it dropped down to 20000 last night
14:23@caker it's the very primitive begginings of the "host-load, change io settings" thing
14:24rko it needs to bring it back to decent level after load is over
14:24@caker yup
14:25@caker ideally, it should work on specific Linodes, rather than set it across the board
14:25rko I don't think it happened this time
14:25rko * * * * * echo 200000 512 > /etc/io_tokens.risto's.linode
14:26@caker sorry bout that!
14:27rko I've been watching this thing and 20k tokens usually gets used in couple of incoming spams...
14:27rko np :-) At least my proved to be very good
14:27rko didn't drive the system to such a deadlock and was able to sustain web services
14:45Efudd Google++
14:46ashen caker: couldn't see about my 128->196 upgrade could you? :)
14:46Efudd 192 tho !
14:46@caker ashen: yeah, let me set that up
14:47ashen caker: thanks :)
14:48@caker ashen: log out and back in, you should see it
14:49ashen caker: thanks :)
15:03--- ---> emcnabb [] has joined #linode
15:07@caker ashen: btw, do NOT start the migration until after your resizes are completed
15:07ashen caker: I've already migrated, the migrations I have going now are increasing my disks back to pre-migration sizes
15:07@caker ashen: it will jump the gun and start copying your images while they're being resized (which doesn't destroy the orig images, but will give you bad copies)
15:07@caker ashen: sweet
15:08@caker ashen++
15:08ashen thanks for the advice though :)
15:08@caker it's a bug
15:08ashen heh, I tend to be careful with anything to do with resizing disks that contain data :-S
15:09@caker msn's search bot STILL stuck in the forums ...
15:09ashen heh, add a .htaccess 403 for its IP? or maybe just something that redirects it out of there?
15:10@caker robots.txt
15:10ashen since when do MS programs obey standards?
15:10@caker There are some mods for phpbb (and some apache rewrites) that make it much more search-engine friendly
15:10@caker heh
15:10@caker but, it would be good for getting better page ranks, etc
15:11ashen mm, I wouldn't know about those, I don't run any website that I count on google for
15:11ashen btw, why do you mind so much about your search engine rankings? Most of the sells you get are from word of mouth
15:12ashen e.g. I've personally referred at least 3, up to 5 people on to you who've gone on to purchase linodes.
15:12@caker not so much page rank to get sales, I just want the loads of good info in the forums indexed for others
15:12ashen and I was referred here I think by someone...... most of your sales seem to come from word of mouth.
15:13--- ---> sjansen [] has joined #linode
15:13ashen maybe getting sites which are looked at by people interested in these things to link to you, could get you a lot of hits?
15:14* ashen doesn't know, he's not a website manager or anything, just wondering if you've tried these options :)
15:14@caker I have ran ads on,, etc
15:15@caker blogs are also a good source for linkage
15:15* ashen nods
15:16ashen I have quite a list of VPS providers, I'm sure that there are sites *checks his bookmarks* that compare VPS providers that you could get a link on
15:16@caker no sites that are worth anything, imo ..
15:17ashen heh, any site that lists a lot of VPS providers, particularly if the list is not big or has a high google ranking, is worth emailing asking if they'd put up a link to you also, imho... costs you nothing, and can generate sales :)
15:18@caker they all suck, and have no page rank worth anything
15:18ashen btw, are you ever going to open a comissions program? I've referred enough people to you that I'm curious about the possibility of comissions *g*
15:18@caker maybe, if things get competitive enough
15:18* ashen shrugs, fair enough, if that's what you've found.
15:19--- ---> Redgore_ [] has joined #linode
15:19--- <<-- Redgore [] has quit (Read error: Connection reset by peer)
15:20--- User: *** Redgore_ is now known as Redgore
15:20ashen I run a free shells provider from my linode you see, and I constantly get people of all ranges from total bash newbie to the semi-experienced admins wanting accounts for everything from ircds to eggdrops to somewhere they can nmap their own home box from, to test their firewall.
15:21ashen I've found the "provider list" pages that simply list a ton of providers to be really helpful, and I thought that if you got listed on those it could lead more people to you. If you're found they don't get many hits, fair enough, it just helped me in the past and I wondered if it'd help you :)
15:28--- <<-- emcnabb [] has quit (Quit: emcnabb)
15:42--- <<-- ashen [] has quit (Quit: Lost terminal)
16:03--- ---> Sh8d0w [] has joined #linode
16:06--- ---> emcnabb [] has joined #linode
16:17wferrel now you've got me all thinking, trying to remember how I found out about linode...
16:25--- <<-- sjansen [] has quit (Quit: I'll show you cruel!)
16:34--- ---> harshy [] has joined #linode
16:44@caker wferrel: you entered "" in your signup data :)
16:44* caker is big brother
17:11sunny it was google groups for me
17:11@caker yup
17:11rko I don't remember...
17:12@caker rko: "Google search"
17:12sunny I wanted UML hosting after I found out how much FreeBSD's jail() sucked
17:12rko oh yeah
17:12rko I was looking for dedicated servers
17:12@caker probably during my strong (read: expensive) AdWords campaign
17:13sunny I thought adwords were cheap ?
17:13rko I was running out of the database capacity at Hostway
17:13rko and was looking for alternatives
17:13@caker they can be, but if you want them to be effective, it's mega $$
17:14sunny oh, I see
17:18--- Netsplit <-> quits: ElectricElf, Sh8d0w, harshy, inkblot, Guest431, fo0bar, Tenkawa, NeXTer, smohekey__, jake, (+21 more, use /NETSPLIT to show all of them)
17:19--- ---> square [] has joined #linode
17:19--- Netsplit over, joins: Guest431, ElectricElf, AndyHat, thelsdj, sadjester, @caker, SupaDongzu, sunny, Efudd, Jayce^ (+21 more)
17:20--- ---> mark100net [~42c2cc1e@] has joined #linode
17:21--- <<-- Redgore [] has quit (Read error: Connection reset by peer)
17:21--- ---> Redgore [] has joined #linode
17:21square a quote i seen this morning, i like
17:21square Give a man an answer, he'll be back for his next question... Teach a man to Google, he'll go the fuck away
17:21square hehe :)
17:25--- Netsplit <-> quits: ElectricElf, Sh8d0w, harshy, inkblot, Guest431, fo0bar, Tenkawa, NeXTer, smohekey__, jake, (+22 more, use /NETSPLIT to show all of them)
17:28--- <<-- adamg [] has quit (Ping timeout: 480 seconds)
17:29--- <<-- mark100net [~42c2cc1e@] has quit (Ping timeout: 480 seconds)
17:29--- <<-- shakr [] has quit (Ping timeout: 480 seconds)
17:29--- <<-- emcnabb [] has quit (Ping timeout: 480 seconds)
17:30--- <<-- schweeb [] has quit (Ping timeout: 480 seconds)
17:30--- <<-- julia [~julia@] has quit (Ping timeout: 480 seconds)
17:30--- <<-- wap [] has quit (Ping timeout: 480 seconds)
17:32--- ---> mark100net [~42c2cc1e@] has joined #linode
17:32--- Netsplit over, joins: thelsdj, fo0bar
17:32--- ---> BGV [] has joined #linode
17:32--- ---> benjers [~180eea83@] has joined #linode
17:32--- ---> CompuServe [~c89ecef1@] has joined #linode
17:32--- ---> avarine [] has joined #linode
17:32--- ---> mikegrb2 [] has joined #linode
17:32--- ---> adamg [] has joined #linode
17:32--- ---> jrm [] has joined #linode
17:32--- ---> guigouz [~guigouz@] has joined #linode
17:32--- ---> ashen [] has joined #linode
17:32--- ---> VanHelsin [] has joined #linode
17:32--- ---> krag [] has joined #linode
17:32--- ---> jcap [] has joined #linode
17:32--- Netsplit over, joins: Sh8d0w, square, harshy, Oxnard, Newsome, wferrel, ElectricElf, smohekey__, chris, inkblot (+1 more)
17:32--- ---> fo0barf [blahblah@] has joined #linode
17:32--- Netsplit over, joins: UML_ChanLog, Tenkawa, NeXTer, mesozoic, unixvps, rko, cout, Jayce^, Efudd, sunny (+4 more)
17:32adamg although this one has just been picked up
17:32adamg []
17:32square oo
17:32benjers it's up now
17:32square where back up
17:32adamg were back
17:32sunny I can't ping my linode
17:33guigouz oh it's back
17:33guigouz thank god.
17:33jcap sunny, concentrate really hard
17:33ElectricElf sunny: Try again, it came back up a minute ago.
17:33--- ---> emcnabb [] has joined #linode
17:33ElectricElf sunny: ("it" being Hurricane Electric's network)
17:33* jcap applauds caker and his phone dialing skills
17:33BGV host 24 down?
17:33sunny yeah, not it works
17:33sunny s/not/now
17:33square lol jcap
17:34--- ---> smokie [] has joined #linode
17:34--- ---> duckpuppy [~d8a0ca03@] has joined #linode
17:34benjers host24 works for me
17:34guigouz \xECt's all up now.
17:34adamg looks like just a network issue this time
17:34BGV Status Failed
17:34sunny thankfully not a host11 reboot issue, heh
17:35BGV Host Message Linode already running
17:35ElectricElf A pretty huge one. Looks like all of HE was down.
17:35--- ---> artifex [] has joined #linode
17:35--- ---> schweeb [] has joined #linode
17:35--- ---> julia [] has joined #linode
17:35sunny ah well
17:35adamg there link in nyc was up and one at pyo
17:35sunny later folks
17:35adamg pao even
17:35--- <<-- benjers [~180eea83@] has quit (Quit: )
17:35ElectricElf adamg: Those were just the border routers though.
17:35--- ---> wap [] has joined #linode
17:36adamg yeah, the nyc one is located in an equinuix dc
17:36ElectricElf adamg: Nobody I checked with could get into, they just hit the borders and that was it.
17:36--- User: *** artifex is now known as Guest1
17:36BGV host24 an error
17:36* ElectricElf could only get to their Chicago exchange
17:36ElectricElf BGV: --- ping statistics ---
17:36ElectricElf 5 packets transmitted, 5 received, 0% packet loss, time 4003ms
17:36@caker interesting router loop earlier
17:36jcap caker, yeah
17:36@caker BGV: it's fixed now
17:36BGV caker thanks
17:37adamg I go through sjc to get to fmt
17:37jcap caker, I've seen similar before. usually a BGP thing, but I don't know enough to say for sure
17:37adamg something major to take down their entire network
17:37@caker no kidding
17:38--- <<-- jrm [] has quit (Ping timeout: 480 seconds)
17:38ElectricElf adamg: Almost assuredly a misconfiguration.
17:38square anyone know phpmyadmin well?
17:38@caker another notch on the "HE is questionable" side
17:38adamg a misconfig on all their routers
17:38jcap square, running gentoo?
17:38ashen caker: Pity you don't have he 100% uptime SLA with HE that you do with TP :)
17:38square jcap: slackware
17:38adamg caker: ill volunteer to be a guinea pig at the new DC :-)
17:39@caker adamg: sounds good -- was wondering .. see my network diagram?
17:39square caker: me too
17:39adamg no
17:39@caker ascii++
17:39ElectricElf caker: Yeah. I mean, shit happens, but you'd expect them to have processes in place to avoid such widespread disruptions.
17:39guigouz does this happen too much at HE ?
17:39adamg looks good
17:40adamg what linux routers are you looking at, I saw the few links you posted last night
17:40@caker router1,switchcore1, and router2,switchcore2 will be on differend power feeds
17:40ashen caker: it'd be interesting to see what uptime% they promise, considering TP promise 100% :)
17:40VanHelsin network is ok again, thanks for the promt solution guys
17:40square 100% isnt strictly possible
17:40@caker none of those, it'll be a 1U Xeon box, 1 (or 2)GB ram, dual e1000s, and a four port e1000 board
17:40adamg will all the hosts be connected to both switches
17:41BGV caker thanks, the machine power on
17:41@caker BGV: great
17:41--- <--- BGV [] has left #linode ()
17:41@caker each cabinet gets it's own switch
17:41adamg your connecting into HE
17:41@caker with uplinks into core1 and core2
17:41@caker hehe .. might not be HE, but someone
17:42jcap I setup a linux load balancer / router using multiple sdsl lines when ip2 just came out. it was very cool
17:42@caker regardless, ED will be the primary provider, and they peer everyone, so ..
17:42jcap we saved a fortune over T1/T3s
17:42ElectricElf caker: Looking to negotiate your own peering and transit?
17:42jcap no reason to mention, I just saw talk of linux routers :)
17:42@caker ElectricElf: I will be at Equinix, yes
17:42ElectricElf caker: Good good.
17:42@caker ElectricElf: save on transit, though ..
17:43jcap caker, looked at Savvis? they're quite good.
17:43@caker ElectricElf: since everything is local
17:43@caker jcap: no, I haven't .. I'll check it out
17:43adamg what sw are you going to be running on the routers?
17:43ElectricElf jcap: I've worked with Savvis before and they were pretty mediocre at the high end.
17:43--- ---> Ashen2 [] has joined #linode
17:43@caker adamg:
17:43@caker adamg: for BGP
17:43jcap ElectricElf, they were better as Exodus :D
17:43ElectricElf jcap: This was a couple of years ago, though.
17:43adamg so you are going to be doing your own bgp then
17:43@caker yup
17:44ElectricElf adamg: Required for multi-homing.
17:44ElectricElf Well, you probably already know that.
17:44jcap mmm bgp
17:44* ElectricElf shuts up :)
17:44adamg knew that but at one point equinix were going to do that and caker only use ED
17:44@caker I just have to learn ARIN's forums and policies like (I/we) did back in the days of domain registration
17:44jcap setup some auto email routines :P
17:45* jcap still thinks back to the glory days of email template domain registration
17:45adamg so are you going for your own ASn then
17:45@caker jcap: big time :)
17:45adamg jcap: RIPE still do that
17:45@caker jcap: I registered so many domains via little script I made, and never paid for them ..
17:45--- <<-- ashen [] has quit (Quit: ->Ashen)
17:45jcap ha
17:45@caker jcap: netsol was so dumb
17:45--- User: *** Ashen2 is now known as Ashen
17:45jcap no kidding. I didn't know RIPE still did
17:46adamg not ripe nominet
17:46--- <<-- CompuServe [~c89ecef1@] has quit (Quit: CGI:IRC (Ping timeout))
17:46@caker ... was interesting how slowed down .. vmstat wasn't going crazy, and server-status showed many children free .. strange
17:47--- <<-- mark100net [~42c2cc1e@] has quit (Quit: CGI:IRC (Ping timeout))
17:50--- <<-- emcnabb [] has quit (Ping timeout: 480 seconds)
17:50--- <<-- VanHelsin [] has quit (Quit: ( :: NoNameScript 3.81 :: ))
17:51--- <<-- Efudd [] has quit (Remote host closed the connection)
17:51--- ---> Efudd [] has joined #linode
17:53--- ---> emcnabb [] has joined #linode
17:58square is it me, or has it gone again?
17:58duckpuppy Looks like network problems again.
17:58square i thought it was the batterys on my kb :P
17:58adamg ok here
17:58guigouz look like
17:58guigouz it's slow
17:58mikegrb2 yup it is back
17:58guigouz didn't stop, but it's missing packets.
17:58square yeh
17:58duckpuppy Ick.
17:58--- <<-- krag [] has quit (Quit: nos da)
17:59* guigouz afraid
17:59guigouz :)
17:59square Ping statistics for
17:59square Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
17:59mikegrb2 odd
17:59mikegrb2 my traces stopped for a bit
17:59guigouz it's missing some packets, not much, but that's how it started the other time.
18:00square yeh
18:00guigouz 37 packets transmitted, 20 packets received, 45% packet loss
18:00adamg i am not seeing any packet loss
18:00square Packets: Sent = 11, Received = 7, Lost = 4 (36% loss),
18:00square Approximate round trip times in milli-seconds:
18:00square yep
18:00square gettng worse
18:01adamg still none here
18:01@caker HE says "another issue cropped up, hoping to have it resolve momentarily"
18:01square jeez
18:01square caker, gonna need a longer list :P
18:01--- ---> greggy [] has joined #linode
18:01@caker some routes are fine (TP --> HE is fine for me), but comcast->HE is the pits
18:01duckpuppy Sent = 4, received = 1 75% loss
18:01greggy to me it is the level3<=>he link that everything stops at
18:01@caker better now?
18:02greggy unfortunately all my accounts directly on he's backbone have been deleted : |
18:02guigouz better
18:02greggy for the moment
18:02duckpuppy Much better.
18:02avarine Looks okay here:
18:02avarine ing statistics for
18:02avarine Packets: Sent = 91, Received = 91, Lost = 0 (0% loss),
18:02avarine pproximate round trip times in milli-seconds:
18:02avarine Minimum = 170ms, Maximum = 180ms, Average = 174ms
18:02mikegrb caker: use this to negotiate better bandwidth contracts!
18:02guigouz 0% packet loss
18:02@caker I think my sales rep is going to be getting a phone call from me
18:02@caker hope she's prepared
18:03adamg you are probably not the only that is going to be calling
18:03greggy is their network flakey or are they just under seige all the time?
18:03@caker it's called Redundancy, folks ...
18:03mikegrb right now they are all in a room
18:03mikegrb being told what to say
18:03guigouz they should have redundancy
18:03mikegrb they say they have redundancy, no?
18:03@caker of course
18:03greggy hahaha they make a big deal about redundancy on their website
18:03greggy but all of their backup links are running near saturation, as near as i can tell
18:04adamg far from what to expect in a Level 1 dc
18:04@caker agreed
18:05adamg I would still love to know how their UPS maintance work took down the entire dc
18:05guigouz adamg: me too!
18:05adamg one would hope they have more than one UPS bank
18:05mikegrb dongs
18:06Sh8d0w well, at least they didn't take out the entire east coast
18:06mikegrb they probably try
18:06square heh
18:07duckpuppy In a completely out-of-the-blue question... has anyone tried the Mandrake 9.1-9.2 update method of "add main and contrib sources for URPMI, then run urpmi --auto-select"? Scary, but just curious...
18:07Sh8d0w wtf, I got a partial email just now, nothing but header
18:07Sh8d0w could that be related to the network problem?
18:08@caker duckpuppy: I did it on a freshly installed 9.1 image
18:08@caker duckpuppy: if you have the space, use the LPM to duplicate your disk image first
18:10square cool, its back
18:11square or did it come back ages ago, and i'm just slow?
18:11guigouz square: ages ago.
18:12square damn
18:12guigouz square: not ages, but about 10 minutes.
18:12guigouz it was actually down for less than 5 minutes.
18:12square yeh, i was reading somthing
18:12square forgot all about it
18:13guigouz "Hurricane Electric" Our self-healing network architecture virtually eliminates a single point of failure.
18:13square heh
18:14guigouz virtually.
18:17--- <<-- duckpuppy [~d8a0ca03@] has quit (Quit: CGI:IRC 0.5.3 (2003/10/29))
18:18--- ---> duckpuppy [~d8a0ca03@] has joined #linode
18:19guigouz well, thanks a lot for everything.
18:19--- <<-- guigouz [~guigouz@] has quit (Quit: There's no place like
18:20square hmm
18:21square anyone know what the deal is once a domain expires?
18:21square oe
18:21square one
18:22Ashen most registrars will give you a short time to reclaim it
18:22--- <<-- Sh8d0w [] has quit (Quit: ...)
18:23Ashen otherwise, it drops
18:23square i want to register it again
18:23square but with someone else
18:23square without paying any transfer out fees
18:23Ashen let it drop
18:23Ashen and register it the same day, with the other company
18:23square yeh, but how long does it take to drop?
18:24square after expire date
18:25Ashen I think if you've set it up with the company that you're not going to renew it, it drops on the exact date
18:26adamg uk domains can take 12 months+ to drop
18:26square :o
18:26adamg it is not automated
18:27mikegrb I'd try doing the transfer stuff a week before
18:27mikegrb transfer fee is same price as one year right?
18:27mikegrb and gets you another year?
18:27adamg when you do not renew it, it gets de-tagged by the company that registered it
18:27adamg and then it is up to nominet when they remove it
18:27adamg mikegrb: uk domains are a bit different
18:28mikegrb heh yeah
18:28adamg the registration period is 2 years and is not as automatted as the .com etc
18:28mikegrb that is why I added the question marks
18:28square want 12.99 per .uk domain, and the company i want to transfer too, want 5.99 per domain
18:28square !
18:28square day light robbery
18:29adamg well are a shite company
18:29square yeh i realise that now :P
18:29adamg and I would not go with a company which offered them at 5.99 either
18:29square 123reg?
18:29adamg sighup 5 * 1.175
18:29sighup 5.875
18:29adamg they are only making 12p profit on that
18:30square heh its 4.99 :p
18:30adamg and you can not really call that profit either
18:30adamg it costs them \xA35 + vat to register them
18:38--- <<-- avarine [] has quit (Quit: Too much, too soon, too far to go, too late to play, the game is over)
18:48--- <<-- greggy [] has quit (Quit: Leaving)
19:26--- ---> shakr [] has joined #linode
19:29--- ---> gkire [~ducky@] has joined #linode
19:41--- ---> david [] has joined #linode
19:41david hello
19:41@caker hello david
19:43david what is ED and HE?
19:43david Hurricane Electric?
19:43@caker ED is Equinix Direct (basically, they peer me onto everyone through them)
19:44@caker and HE represents a secondary provider
19:44david oh
19:44david caker: you can probably lose the cross connect between the two core switches
19:44david caker: although, maybe you'd want it
19:44david to stop traffic hitting more useful links in the event of a failure
19:45@caker david: any usefullness in a cross-connect between the two routers? .. I couldn't really think of one
19:45david caker: heartbeed
19:45david heartbeat
19:45@caker david: that's something you don't want to do through the switches?
19:45david caker: I always heartbeat over a dedicated VLAN, but I'm a prude
19:45mikegrb I usually heatbeat in my chest
19:46* caker beats mikegrb
19:46mikegrb yes like that
19:46david caker: you can, but I like it to be independent of any other device
19:47* mikegrb too
19:47mikegrb would hate to have a pace maker and have to worry about microwaves and stuff
19:48david caker: if not a sepeate VLAN, I'd at least heartbeat over a serial cable
19:48david caker: I just find it easier to use a cross-over ;-)
19:48@caker david: was just wondering about that
19:49* mikegrb wonders what would happen if I dropped my ibook in the bath water
19:49@caker david: so in the event of one fails, the other brings up the gateway IPs?
19:49mikegrb don't think it would affect my heartbeat other then being pissed
19:50david caker: brings up the other inside address, yes
19:50mikegrb and perhaps tries to tell the other to restart and/or page you
19:50david caker: depending upon your internal routeing, you might even not want to heartbeat
19:50david caker: and just run both routers in 'active' mode
19:50david and use OSPF on the inside
20:13--- <<-- mikegrb2 [] has quit (Remote host closed the connection)
20:21--- ---> jrm [] has joined #linode
20:22@caker bbiaf
20:34--- <<-- jrm [] has quit (Ping timeout: 480 seconds)
20:45--- <<-- SupaDongzu [] has quit (Remote host closed the connection)
20:45--- ---> SupaDongzu [] has joined #linode
20:54--- <<-- duckpuppy [~d8a0ca03@] has quit (Quit: CGI:IRC (Ping timeout))
20:59--- <<-- harshy [] has quit (Quit: I quit for now)
21:05--- <<-- emcnabb [] has quit (Ping timeout: 480 seconds)
21:43--- ---> jrm [] has joined #linode
21:45jrm anybody know what the problem at HE was earlier today?
21:53@caker jrm: I'm thinking someone misconfured a BGP table, since we saw some interesting route-loops
21:53mikegrb I had fruit loops
21:53mikegrb for breakfast
21:53mikegrb other day
21:53mikegrb now I go have a cigaretet
21:53mikegrb I earnedd it
22:09heidi hrm
22:14jrm cool
22:21--- <<-- jrm [] has quit (Quit: Client exiting)
22:50--- <<-- gkire [~ducky@] has quit (Quit: )
22:52--- ---> emcnabb [] has joined #linode
23:55--- <--- UML_ChanLog [~stats@] has left #linode (Rotating Logs)
23:55--- ---> UML_ChanLog [~stats@] has joined #linode
---Logclosed Thu May 20 00:00:19 2004