Back to Home / #linode / 2006 / 03 / Prev Day | Next Day
#linode IRC Logs for 2006-03-30

---Logopened Thu Mar 30 00:00:00 2006
00:00|-|Eman [] has quit [Quit: A fatal exception 0E has occurred at 0157:BF7FF831 in VXD VMM(01) + 00010E36. The current application will be terminated.]
00:01<@caker>taupehat: no
00:02<taupehat>ok then...
00:03<Spads>taupehat: the commons is simply tragedy-resistant now
00:12|-|Sgeo [] has quit [Quit: Konversation terminated!]
00:13<fo0bar> <-- soon...
00:14|-|Eman [] has joined #linode
00:14<@caker>fo0bar: wow, I was just going to ask for that (seriously)
00:14<fo0bar>a status update?
00:15<@caker>fo0bar: not kidding. no the feature -- have a finnix disc in all the hosts
00:15<fo0bar>yesterday I found the isolinux foo necessary to do that
00:15<@caker>I suppose I could jsut tweak boot prios in the bios, but still .. pretty cool
00:16<@caker>Does finnix *always* eject the CD?
00:16<@caker>that would be a problem on those slim cd-drives without a inject feature
00:16<fo0bar>caker: with 87.0, you can also copy the /FINNIX/FINNIX master image to an ext2/vfat filesystem, copy a few kernels around, and add a "finnix" line to your grub setup
00:16<fo0bar>I need to write a howto for that
00:16<fo0bar>but it sounds like that would work better
00:17<@caker>not if the drives/raid/whatever is kaput
00:17<fo0bar>to answer your question: there is a "noeject" boot flag, but by default it always ejects upon shutdown
00:18<@caker>is that the surprise?
00:18<fo0bar>no :)
00:18<fo0bar>it was just something I came across yesterday
00:19<fo0bar>I also found a neat floppy-based bootloader that can boot from CD, even if the bios doesn't support it
00:19<@caker>So, what's the jist of how that works? It kicks you back into the drive's bootloader?
00:19<fo0bar>oh, no, that was unrelated
00:20<fo0bar>there is an option in isolinux to tell the bios mark the boot attempt as invalid and move on to the next boot device. however, in testing, this is buggy (my first attempted machine bugged out on disk booting and skipped straight to PXE)
00:21<fo0bar>so I used another similar option, which just tells the bios "boot your first hard drive"
00:21<fo0bar>no matter what the boot order was
00:21<@caker>this is all bootloader magic, I assume?
00:22<fo0bar>I assume on the backend... my side of it was 2 lines in isolinux.cfg :)
00:22<@caker>yeah .. pretty cool stuff
00:23<@caker>label 0
00:23<@caker> localboot 0x80
00:23<fo0bar>that's it
00:24<fo0bar>"localboot -1" is the other option I mentioned, but like I said, it didn't work well
00:27<fo0bar> <-- they're totally ripping off my idea of using huge version numbers :)
00:27<taupehat>fo0bar: take a look at my versioning system for idiocylevel.xpi
00:29<taupehat>even the older version
00:29<taupehat>I think my first release was or something like that
00:30<@caker>fo0bar: all Finnix needs to totally blow up is a writeup in Linux [format,journal,magazine] or some other publication/syndicated website
00:30<taupehat> /.
00:30<fo0bar>caker: I send a couple cds to linux journal after 86.2, but they never responded :(
00:30<@caker>my guess is that'll happen soon
00:30<fo0bar>taupehat: I think finnix was in slashdot back in 200
00:30<fo0bar>err 2000
00:31<taupehat>oh ah
00:31<fo0bar>but back then a "livecd" was a big deal
00:32<fo0bar>finnix is basically credited as the second "livecd" ever
00:32<fo0bar>the first being linuxcare's bootable business card
00:33<Spads>well, second bizcard-sized livecd
00:33<Spads>knoppix was first
00:33<fo0bar>then again, back then everybody just used tomsrtbt
00:33<Spads>early lnx-bbc was tomsrtbt on CD-ROM
00:33<Spads>because none of linuxcare's servers had diskette drives
00:33<fo0bar>Spads: 1) finnix was 300MB back then, 2) knoppix didn't come around until late 2001
00:33<Spads>oh right
00:33<Spads>knoppix was after us...
00:33<fo0bar>and even then it wasn't known until 2002-ish
00:33<Spads>who was it though...
00:33<Spads>I guess we were first
00:34<Spads>Yggdrasil was the first livecd
00:34<fo0bar>demolinux came out shortly after finnix, which functionally was the precursor to knoppix
00:34<Spads>that's what i was thinking of
00:35<@caker> <-- these things freaking ROCK
00:35|-|warewolf [] has quit [Read error: Connection reset by peer]
00:35<fo0bar>I'm not counting bootable CDs that were primarily vectors to install the real OS on a disk
00:35<@caker>man, I'm so happy with it
00:35|-|warewolf [] has joined #linode
00:36<fo0bar>if that's the case, yggdrasil, slackware, red hat, mandrake, caldera, etc all came before
00:36<fo0bar>ahh caldera, how I miss thee
00:36<fo0bar>to date, the only distro where you could play tetris during installation
00:36<fo0bar>the OS sucked, but... tetris!
00:36<@linbot>caker: The latest stable kernel is; the latest snapshot of the stable kernel is 2.6.16-git17; the latest beta kernel is 2.6.16-mm2.
00:36<@caker>fo0bar: not going to bother with .1?
00:37<fo0bar>caker: ugh. I'll have to look, but I'm hoping many of the patches in that release were already covered by debian's patchset
00:39<@caker>I'm going to need a new host naming scheme if I keep adding hosts to both at the same time. I've been considering
00:39<@caker>eventually:, etc
00:39<@caker>I could do a subdomain ..
00:39<fo0bar>caker: if I still worked at a datacenter, I'd totally court you :)
00:40<@caker>fo0bar: hehe
00:40<@caker>fo0bar: where now?
00:40<@caker>(do you work)
00:40<taupehat>eh, what's the shortest route to scan a network segment for hosts?
00:40<taupehat>I just setup a cheap wireless AP and can't find its addr
00:40<taupehat>tried ping -b but that doesn't seem to do it
00:40<@caker>taupehat: arp -a, or nmap -sP x.x.x.1-254
00:40<@caker>yeah, that's the other trick
00:41<fo0bar>technically it's a marketing company, but I don't really do anything with that end. I just do IT and sysadmin work for dedicated hosting clients
00:41<fo0bar>but we're not a public dedicated hosting provider
00:41<fo0bar>it's only if we're doing marketing wotzit for you
00:41<@caker>fo0bar: did this involve an acquisition or did you move jobs?
00:42<fo0bar>heh, that's a long and semi-interesting story
00:42<@caker>I thought your company got acquired, at least that's what I can remember ...
00:42<fo0bar>so I worked for a datacenter startup (employee #5 or something). there was also this other company, who while we weren't related, if you go up the VC chain far enough, you'd hit the same company
00:43<fo0bar>but the other company was just exiting "startup" status (this was about 4 years ago)
00:43<fo0bar>and they became our first colo client
00:44<fo0bar>anyways, Redundant (the datacenter) goes belly up, but the datacenter itself still has to be run until the assets can be sold, customers can move away, etc
00:45<fo0bar>I get laid off, then about 4 months later get a call from the other company saying "we've been hired to take over the datacenter operations until the assets can be sold. want to work for us?"
00:45<fo0bar>so I did
00:46<fo0bar>so I spent the rest of the year basically doing my old job in the same datacenter for a different company, until the datacenter itself was sold towards the end of the year
00:46<fo0bar>we moved to a different datacenter, and now I build servers :)
00:47<taupehat>ahh, this should be fun
00:47[~]taupehat phones linksys support
00:48<fo0bar>anyways, I like the subdomain better
00:48<taupehat>the deal seems to be totally bridging
00:48<@caker>me too
00:48<taupehat>but it's got open wireless
00:48<taupehat>can't figure out how to talk to it
00:49<taupehat>(without a windows machine)
00:49<fo0bar>caker: but I'd continue incrementing globally (,, etc), and still keep around as aliases
00:50<@caker>fo0bar: the aliases for now, yes .. curious why you'd keep them unique
00:51<@caker>I guess the subdomain solves the "all fremont hosts" problem
00:51<fo0bar>I fear if you have overlapping numbers, it'll cause confusion ("host127 in dallas is having problems" "OMG, I'm on host 127!" "no, you're on host127 in atlanta")
00:51<@caker>and people can still say "hostXX" without refering to location
00:51<cmantito>you could do Host D127 (dallas) or host A127 (atlanta)
00:52<cmantito>(ie, make the DC part of the host's number.
00:52<@caker>, host12f, host12d ...
00:52<@caker>well, host13f, host14d
00:55<@caker>hmm .. the host load bar doesn't mean much under Xen now..
00:55<@caker>the "host" is really just a Xen domain like a Linode
00:55<fo0bar>it's more to type (which is why I recommended keeping around if you're "in the know"), but the benefit I see to having the full location name as a subdomain is you would put "" in the overview of the LPM, and it would be easier to say on the forums "dallas is having routing problems", instead of "dallas is having routing problems (all hosts that end in 'd')"
00:56<@caker>fo0bar: right
00:56<fo0bar>sadly, most of my argument is based upon reactions to bad things happening :P
00:56|-|Dreamer3 [] has quit [Ping timeout: 480 seconds]
00:57<fo0bar>caker: find some way to measure relative disk traffic maybe?
00:58[~]cmantito pokes caker in the most unannoying way possible for a status update with tha Arch image? ;)
00:58<@caker>cmantito: hehe .. soon, I just built another machine here this evening
00:58<cmantito>cool :)
00:58<fo0bar>the biggest contention is probably still going to be IO, so that would probably be the most appropriate benchmark for machine status
00:59<cmantito>that, and while I'm prodding you, can I get a picture of host52? :P
00:59<@caker>fo0bar: we had two nodes totally thrashing to hell, and no one else could tell
00:59<@caker>cmantito: :)
00:59<cmantito>aw, I meant in the racks ;)
00:59<@caker>fo0bar: could still ssh into Lish, no problems .. was Good Stuff
01:00<Spads>caker: I find my node's load/lag phenomenon puzzling. Reboots seem to improve performance drastically, even though the same number of programs end up being run again.
01:00<fo0bar>cmantito: I may be able to take a picture of a similar supermicro in my company's racks... you could pretend that was your box :)
01:00<Spads>and rngd still pops up and swallows CPU every so often, which is odd
01:00|-|Dreamer3 [] has joined #linode
01:00<@caker>Spads: in UML?
01:01<Spads>right now I'm not even into swap, so whatever
01:01<Spads>and I'm hoping that the I/O scheduler fixes the other problems
01:01<fo0bar>caker: speaking of which, I think I'll do a final backup of my host16 node, then switch over to xen
01:01<Spads>but it's still weird. like what could have been accreting in my UML?
01:01<@caker>fo0bar: cool, submit that form and I'll set it up right now
01:01<Spads> 1279 root 15 0 8044 544 440 S 7.2 0.2 1:44.67 rngd
01:02<Spads> 1279 root 15 0 8044 544 440 S 9.0 0.2 1:45.47 rngd
01:02<fo0bar>caker: done
01:03<fo0bar>rsync baby!
01:05<@caker>fo0bar: migration set
01:05<@caker>(relogin to see it)
01:06<fo0bar>caker: I have an existing finnix profile set up on that account as well... will the migration take that into affect as well?
01:06<fo0bar>(I know finnix isn't available on the xen hosts yet)
01:06<@caker>fo0bar: no -- the profile will go with you, but not the disks
01:07<fo0bar> damn, I took pride on being on host16
01:08<Spads>going xen
01:08<@caker>I haven't out a host online since November because of Xen
01:09<cmantito>caker: which distributor for supermicro do you use?
01:09<@caker>cmantito: mwave
01:09<fo0bar>Spads: the xen public beta is only in HE. you'd have to switch IPs to migrate zork
01:09<Eman>mwave always reminds me of horrible ibm soundcards...
01:09<Spads>I know
01:09<fo0bar>Eman: hah
01:10<@caker> for the 3ware cards, and as always: ATP-USA for RAM
01:11<fo0bar>I usually just get "mwave brand ram", which is almost always kingston
01:12<fo0bar>but sometimes a direct hynix branding
01:12<@caker>ok .. duh
01:12<@caker>Mar 30 02:10:56 nova named[10439]: has multiple CNAMES
01:13<taupehat>you get memory from them eh
01:13<taupehat>like memory foam? =P
01:13<@caker> silly
01:13<fo0bar>so, if I click "migrate to", it [ ] will [ ] will not break due to DNS zone problems :)
01:13<taupehat>I know
01:13<taupehat>sometimes google will produce funny results
01:14<taupehat>fo0bar: I migrated a couple hours ago
01:14<taupehat>it's fine
01:14<@caker>fo0bar: no, it's a pull anyhow -- you're ok
01:14<taupehat>email is arriving as it should
01:14<taupehat>too damn much in fact
01:14<fo0bar>taupehat: yeah, but caker has been futzing with the naming scheme just now
01:14<fo0bar>caker: cool
01:15<fo0bar>ooo, my LPM is now "xen-branded"
01:17<taupehat>so how does xen handle IO loading by a particular domain?
01:18<taupehat>like if someone ran updatedb, what's to prevent that person from eating IO capacity from the other domains on that host
01:20<@caker>I guess it doesn't matter which one I make the CNAME .. host56, or host56.fremont
01:20<taupehat>lol... you tested it with a forkbomb?
01:20<@caker>less typing if the .fremont is the CNAME
01:21<taupehat>21:25 < caker> says 189% cpu now :)
01:22<@caker>go go SMP Linodes
01:22<taupehat>oh man
01:22<taupehat>that guy from oklahoma is too damn much
01:23<taupehat>he's at it again!
01:25<taupehat>apparently the man has never head of "google'
02:08|-|spr [] has quit [Quit: Spoon!]
02:19<fo0bar>00:19:00.106522 IP > 7419 A? (36)
02:19<fo0bar> hasn't existed in nearly 2 years
02:20<fo0bar>but I still get thousands of queries per day
02:30|-|Viza [] has quit [Ping timeout: 480 seconds]
02:44<taupehat>I wonder if moving hosts could possibly explain the sudden spike in ssh bruteforce I'm seeing. I can't think of how that might be the case...
02:44<taupehat>fail2ban for the win! anyhow =]
03:16|-|Shaun222 [] has joined #linode
03:20|-|Shaun [] has quit [Ping timeout: 480 seconds]
03:29|-|Shaun222 [] has quit [Ping timeout: 480 seconds]
03:29|-|Shaun [] has joined #linode
03:59|-|jekil [] has joined #linode
04:02|-|Shaun [] has quit [Ping timeout: 480 seconds]
04:31|-|Shaun [] has joined #linode
04:44|-|encode [] has quit [Ping timeout: 480 seconds]
04:46|-|encode [] has joined #linode
05:41<cmantito>root@redbaron:/var/log# rm /var/mail/kevin
06:57<cmantito>finally, an hour dicking around with slackware's cron and I finally get RML client reporting ;)
07:20|-|linville [] has joined #linode
07:50<npmr>sudo is nice
08:05|-|Redgore [~Redgore@] has joined #linode
08:21|-|jekil [] has quit [Quit: Leaving]
08:43<warewolf>cmantito: what was the problem?
08:44<cmantito>slackware's got a crap cron system.
08:44<cmantito>s'not the RML client ;)(
08:44<warewolf>last I used slackware it was vixiecron
08:44<cmantito>s'not the RML client ;)
08:44<warewolf>did you have to 0,5,10,15,20,25 .. etc?
08:44<cmantito>whatever it is, it doesn't work very well.
08:44<cmantito>I got it working with */5
08:45<cmantito>for starters, it doesn't store any cron stuff in /etc.
08:45<cmantito>then I had to restart it a dozen times before it worked
08:45<warewolf>were you using cronab to edit the cron file?
08:45<cmantito>it doesn't save.
08:45<cmantito>I had to edit it by hand in the end.
08:46<warewolf>er crontab
08:46<cmantito>it just kept reverting
08:47<cmantito>turns out, the crontab.$RANDOM files it was making never got moved to the actual crontab file.
08:47<warewolf>you need to use crontab to edit the cron spool file
08:47<warewolf>otherwise the cron daemon will probally detect that the /var/spool/cron/FOO file you frobbed was modified, and revert it.
08:47<cmantito>whatever. it works now :)
08:51|-|Viza [] has joined #linode
09:16<warewolf>hum. strange.
09:17<warewolf>the wiki mark-up in my new wiki I'm adding into RML doesn't seem to actually be working
09:17<warewolf>all it seems to be doing is translating StudlyCaps into wiki-nodes.
09:17<cmantito>what wiki is it?
09:17<warewolf>a ripped apart CGI::Wiki::Kwiki
09:18<warewolf>ripped apart, reimplimented, and retooled :)
09:18<warewolf>so yeah it's my fault in there somewhere :)
09:18<cmantito>I've never managed to have much luck with Wikis.
09:18<cmantito>I gave up ;)
09:18<warewolf>the entire RateMyLinode website is mod_perl
09:18<warewolf>all of it
09:18<warewolf>backed by HTML::Mason
09:19<warewolf>I'm tieing the wiki into the existing authentication crap I wrote
09:19<cmantito>Wikis confuse me.
09:20|-|cherring [] has joined #linode
09:21|-|cherring [] has quit [Quit: ]
09:23<warewolf><b>this is bold</b>
09:23<warewolf>WHY DID YOU ESCAPE THAT
09:25|-|sonorous [] has quit [Ping timeout: 480 seconds]
09:26<cmantito>because it *can*
09:26<warewolf>something's not right
09:26<warewolf>the wiki module isn't loading the wiki formatter module correctly or something
09:27<warewolf>actually it is
09:28|-|sonorous [] has joined #linode
09:29|-|sonorous changed nick to Guest5513
09:29<warewolf> '_formatter' => $wiki->{'_registered_plugins'}[0]{'_formatter'},
09:29<warewolf>it's using the stock default formatter
09:30<cmantito>yayperl :P
09:30<warewolf>(I'm sure you can tell what's wrong with just that line .... )
09:30[~]warewolf haxhaxhax perl
09:30<cmantito>perl is teh sexy :)
09:30[~]cmantito <3 regexes
09:30<Spads>lots of languages have regexes dude
09:30<Spads>hell, awk is hardly anything but
09:31<cmantito>yea, but they're kinda one of perl's specialties.
09:31<Spads>and awk's
09:31<Spads>I'll agree that perl got a lot of regex syntax right, where things like POSIX regex syntax are a little clumsy
09:31<cmantito>yes, but perl is just sexier than awk. ;)
09:31<Redgore>which is why alot of things support the perl style
09:32<Spads>no way dude, awk is teh sex-ayyyyy
09:32<Spads>awk is clean and sleek
09:32<Spads>Perl is covered with
09:32<cmantito>awk would have to pay me lots, but I'll use perl anyday ^.^
09:32<Spads>%-shaped carbunkles
09:32[~]Redgore takes two cats and skins them in different way's
09:33<Spads>Redgore: by "way's" do yu mean "way is" or "way has"?
09:33[~]cmantito wants the one that resembles perl.
09:33<Redgore>Spads: I mean as in plural of way
09:33<cmantito>and, don't skin cats. cats rule.
09:33<Spads>perl was hot shit in 1997
09:33<warewolf>it still is.
09:33<Spads>Redgore: plurals never use apostrophes, unless they are possessive plurals
09:33<cmantito>still is.
09:33<Spads>You're right
09:33<Spads>It still is hot shit in 1997
09:33<Spads>I feel like it's 1997 every time I look at Perl code
09:33<Redgore>Spads: perhaps thats why I sucked at college ? :P
09:34<cmantito>Spads: really? quite a lot is written in perl. in 2006.
09:34<Spads>cmantito: and even more is written in better languages. Thank progress!
09:34<cmantito>more is relative, as is better.
09:35<cmantito>Spads: what do you consider 'better'?
09:35<cmantito>PHP? ASP?
09:35<Spads>languages that don't tout all these great syntax features, then frown on you for using them
09:35<Spads>"AUTOVIVIFICATION IS TEH BEST!!!!" "Dude, why aren't you use strict;ing? Autovivification might bite you in the ass!"
09:35<warewolf>auto what?
09:35<Spads>python, ruby... hell, even MONO
09:36<Spads>warewolf: that's larry wall's term for the auto-creation of implied variables
09:36<warewolf>php pisses me off for it's use of "magic"
09:36<warewolf>eg, "magic arrays"
09:36<Spads>used to be OMFG TEH BIG FEECHUR OF TEH PURL
09:36<warewolf>there is nothing "magic" in programming
09:36<cmantito>almost everything I interact with on a daily basis is perl or php.
09:37<Spads>Guh, PHP is even worse
09:37<Spads>and ASP even worse
09:37<cmantito>Spads: gee, php did the same thing with register_globals. iirc, python also did something similar, although I don't know python offhand to confirm that.
09:37<Spads>comparing them to Perl as an example of perl's supremacy is like those qmail dorks who insist on holding up sendmail as the only competition
09:37<Spads>cmantito: python did nothing similar
09:37<cmantito>right, well, I'll be back later.
09:38|-|cmantito [] has left #linode [This is useless.]
09:38<Spads>cmantito: trust me, I'd take Perl over PHP any day of the week
09:38<Spads>man, trumpeting your language is fine
09:38<Spads>but it's always amusing when you clearly haven't looked at the competition
09:38[~]kvandivo yawns.
09:38<Spads>okay, off to alcatraz
09:52|-|jekil [~alessandr@] has joined #linode
09:55|-|cout [] has quit [Ping timeout: 480 seconds]
09:58<warewolf>sweet, wiki is almost done.
10:10[~]warewolf downloads latest WoW update
10:10<warewolf>squee, 1.10, yaay
10:11|-|cmantito [] has joined #linode
10:11|-|nybble [] has joined #linode
10:19|-|Guest5513 changed nick to sonorous
10:35|-|FireSlash [] has joined #linode
10:37[~]caker grumbles something about jury duty
10:39<@caker>I'm thinking if I look dirt enough, they won't pick me
10:41<scott>yeah, that would suck
10:42<@caker>I think so. I'll just explain to them that I run a business with only one employee, and jury duty could potentially jeapordize it
10:42<@caker>of course, the case could be against a computer programmer, or some other conflict
10:44<kvandivo>caker: coroner's jury duty?
10:44<scott>hire kvandivo to go as you
10:45<scott>he has experiance
10:45<kvandivo>i hated my regular jury duty..
10:45<kvandivo>i enjoyed coroner's jury duty
10:45<warewolf>caker- if you tell them that your business will suffer otherwise, your jury duty will only get delayed until a later date
10:45<@caker>warewolf: that's my plan
10:46<@caker>I've always been impressed with the courts system, so I do want to fulfill my civic duty, just not now :)
10:46<@linbot>New news from forums: Emails being ignored in Sales Questions and Answers <>
10:46<warewolf>speaking of customer service :)
10:47<kvandivo>it's already suffereing!
10:48<heidi>well if he can't get a reminder for the password then any replies sent to him from linode wouldn't reach him either
10:49<heidi>since he doesn't know which email account everything is being sent to
10:49<Redgore>probably has his gmail inbox full of large files
10:59|-|scott [] has quit [Quit: leaving]
11:00|-|adamg [] has joined #linode
11:08<fo0bar>kvandivo: Interesting story. I was just reading this:
11:08<fo0bar>a coroner's jury may have helped there
11:08<Redgore>What do gareth gates and harold shipman have in common ?
11:10<Redgore>They both cant finish a sentence
11:10<kvandivo>fo0bar: interesting.. the methodology provided by a coroner's jury would possibly have alerted the coroner to the repetitive occurences..
11:11<kvandivo>each individual jury would probably have found the death to be normal.. but the coroner who's there for years and sees the "big picture" would hopefully get a clue
11:12<fo0bar>kvandivo: yeah, that's what I was thinking
11:13|-|afv-13 [] has joined #linode
11:14<kvandivo>at any rate, i found the jury duty to be interesting.. i left it feeling good.. totally unlike my regular jury experience
11:16<Redgore>a coroner most likely wasnt called in, afterall Shipman did change medical records
11:17<kvandivo>hmm.. in our county the coroner is notified for all deaths, regardless of cause
11:17<Redgore>im not 100% up on how coroners work here in the UK, so I could be wrong
11:18<kvandivo>and the coroner's jury is convened for any case that isn't very plainly natural causes (ie., dying of old age)
11:22<Redgore>they might be notified of all deaths and would be given the records, it would then be down to them seeing something suspicious to launch investigation
12:01<gpd>What is the 'word on the street' about iPods atm? Good time to buy or bad? I forget the site that has tips for apple hardware... ?
12:03|-|jfaulkne [] has quit [Quit: Lost terminal]
12:04<@caker>gpd: I got a 4GB nano about two months ago -- love it
12:04<gpd>is the screen scratching issue fixed?
12:04<@caker>dunno, I bought a case that came with a clear plastic stick on thing to protect it
12:05<gpd>there seem to be rumors and counter rumours about April 1st announcements about new video pods
12:05<@caker>yeah, 30th anniversary and all
12:05<@caker>worth it to wait a few days, IMO
12:05<gpd>ya - concluding the same myself
12:06<kvandivo>have iPods been around that long? Geez.. time flies
12:29<afv-13>Host Load: low
12:29<afv-13>always good to see :D
12:40|-|spr [] has joined #linode
12:57|-|cout [] has joined #linode
12:58|-|Dreamer3 [] has quit [Ping timeout: 480 seconds]
12:59|-|Dreamer3 [] has joined #linode
13:09|-|FireSlash [] has quit [Read error: Connection reset by peer]
13:10|-|macdan [] has joined #linode
13:52<@caker>I'M A FREE MAN!
13:52|-|spr [] has quit [Quit: Spoon!]
13:53<gpd>No - you are number 1.
13:53<gpd>I am number 6.
13:53<@caker>They called my name #5 out of about 50 people for the grand jury
13:54<gpd>5th first to be removed?
13:54<gpd>you must have been dirty...
13:54<@mikegrb>gpd: nah
13:54<@mikegrb>gpd: he had slept with the defendants daughter
13:54[~]mikegrb runs
13:56<@mikegrb> <-- spiffy
13:56[~]xptek tracks it right to mikegrb's door
13:56<@mikegrb>nah, just goes to city center
13:57<xptek>Google Earth knows all
13:57[~]xptek launches it and follows the truck
13:57<@mikegrb>worked out nicely for me
13:57<@mikegrb>weather caused a flight delay so the package couldn't be delivered by 10:30
13:58<@mikegrb>apple called fedex to find out the status, fedex told apple it would be delivered tommorow so apple refunded my $60 for overnight shipping
13:58<@mikegrb>then a few hours later the fedex page said it was here in pensacola
13:58<@mikegrb>and an hour ago out for delivery
13:58<@mikegrb>so looks like
13:58<xptek>The only time Apple and Fedex have ever given me anything on time was the Mac OS X release.
13:58<@mikegrb>I will get it today afterall
13:58|-|spr [] has joined #linode
13:59<xptek>Otherwise there's always some stupid delay
14:16|-|Marcel [] has quit [Quit: Leaving]
14:25<fo0bar> <-- odd, it skipped the LA portion of it
14:29[~]caker looks at getting initrd support going for Xen (and with it, Finnix)
14:31<@caker>fo0bar: the final version of finnix (iso,initrd,etc) somewhere yet?
14:31<@caker>fo0bar: also, have you managed to patch a xenified kernel with those patches it requires?
14:31<fo0bar>caker: not yet, official release is tomorrow, but I'll be able to get it to you this evening
14:32<fo0bar>for patches, yes, I patched the 3.0.0 release's 2.6.12 kernel with unionfs and squashfs
14:32<warewolf>ah, the super secret hookup
14:32[~]warewolf goes back to hacking on RML again after playing some WoW
14:32<fo0bar>caker: the xen initrd hasn't changed between versions though
14:33<@caker>fo0bar: ok, I'll be able to use the same initrd for both, right?
14:34<fo0bar>and I'm 99% sure my xen initrd and my uml initrd are the same, but I'm checking now
14:34<warewolf>(caker's mind ... ) .oo( Less work for me! )
14:37|-|Dreamer3 [] has quit [Ping timeout: 480 seconds]
14:38<fo0bar>caker: looks like it. so just use the initrd from the 86.2 release on the linodes
14:39<fo0bar>caker: you did eventually push 86.2's iso/initrd from host11 to the rest of the hosts, right?
14:39<@caker>fo0bar: nope... ok, so host11's finnix initrd is what I should use.
14:39<fo0bar>caker: bastard :P
14:40<warewolf>this is so cool to see the FIGHT between two "vendors" :)
14:41<@caker>fo0bar: woohoo .. 58 total finnix config profiles
14:41<fo0bar>let me know when you're ready to start looking at the kernel... 2.6.16 will require a different patch for unionfs
14:41|-|Sgeo [] has joined #linode
14:42<@caker>fo0bar: off hand, know how to pass an initrd to a domU?
14:42<fo0bar>yeah, one sec
14:43<@caker>ramdisk = "/boot/initrd-xen"
14:43<fo0bar>that's it
14:45|-|Dreamer3 [] has joined #linode
14:55<fo0bar>oh, and it looks like both unionfs and squashfs are planning on making bids to get accepted in mainstream kernel
14:56<fo0bar>appearantly squashfs tried in the past and was shot down before
15:04|-|FireSlash [] has joined #linode
15:05<warewolf>hey caker, you around?
15:05<warewolf>want to do me a flavor?
15:09|-|nybble [] has quit [Quit: Leaving]
15:24<@caker>huh .. I'm surprised you can cat a device file
15:24<@linbot>New news from forums: firehol - integer expression expected in Linux Networking <>
15:24<@caker>guess that makes sense
15:24<warewolf>block or character, yes
15:24<warewolf>there are some that need to be memmapped or something
15:24<warewolf>I forget which
15:25<warewolf>they return "Invalid Address" when you try to cat them
15:34<warewolf>"Teen faces Child Porn Charges For Posting Pics Of His Friends [having sex on his weblog]"
15:36|-|Redgore [~Redgore@] has quit [Remote host closed the connection]
15:38|-|Redgore [~Redgore@] has joined #linode
15:39|-|macdan [] has quit [Quit: Got Itch?]
15:50<@linbot>New news from forums: Unscheduled shutdown in General Discussion <>
16:02|-|Redgore [~Redgore@] has quit [Quit: A geek without purpose - | SMDC-Network IRC -]
16:04<afv-13>kernel BUG at mm/rmap.c:486!
16:04<afv-13>Kernel panic - not syncing: BUG!
16:12|-|linville [] has quit [Quit: Leaving]
16:13<@caker>fo0bar: ok, initrds and the iso are working .. just need the kernel juice
16:13<@caker>RAMDISK: Couldn't find valid RAM disk image starting at 0.
16:13<@caker>maybe not...
16:24<fo0bar>one sec
16:24<warewolf>god damn do I ever hate CSS
16:24<fo0bar>caker: ok, you'll need this:
16:24<fo0bar>that's the same as last time
16:24<@caker>fo0bar: that'll patch against 2.6.16?
16:25<fo0bar>then this:
16:25<fo0bar>untar it, and there's a script in there to patch the kernel. point it at your kernel tree
16:26<@caker>crap .. got a .config laying around?
16:26<@caker>a xen one would be nice
16:26<@caker>guess I could start from my uml config
16:27<fo0bar>I may still have the one from 2.6.12, but that would be difficult to migrate
16:27<fo0bar>why not just start from the regular 2.6.16-xenU?
16:27<@caker>ok, I've got one from the 2.6.14-finnix kernel
16:27<@caker>true dat
16:29<@caker>fo0bar: do you recall if Xen wants that ramdisk uncompressed?
16:29<fo0bar>I've never tried compressing it, as it's already (squashfs) compressed
16:30<fo0bar>for xen that is
16:30<fo0bar>uml/normal kernel will accept a gzipped squashfs partition
16:30<fo0bar>so I assume xen would
16:38<@caker>ok .. makes sense I got that error above if the initrd is squashfs without kernel support
16:38<@caker>compiling now
16:41<fo0bar>oh yeah
16:41<fo0bar>did the unionfs patch script apply cleanly?
16:41[~]mikegrb sends fo0bar a power strip as a gift
16:49|-|FireSlash [] has quit [Read error: Connection reset by peer]
16:53|-|toxicbomber [] has joined #linode
16:53<toxicbomber>Hey can you go to this link to get me more points to get me more mp3's? Thank you if you can!
16:54|-|mode/#linode [+b *!*robotmonk@*] by mikegrb
16:54|-|toxicbomber kicked [#linode] mikegrb [toxicbomber]
16:57<fo0bar>mikegrb: dude, all he wanted was to get more mp3s!
16:57<@mikegrb>I got him akilled
16:57<@mikegrb>I'm so mean
16:57|-|mode/#linode [-b *!*robotmonk@*] by mikegrb
16:59|-|flatronf700B [~flatronf7@] has quit [Ping timeout: 480 seconds]
17:00<@caker>fo0bar: here goes nothing
17:01<@caker>wait for it...
17:01<@caker>Welcome to Finnix for Xen!
17:02<@caker> WARNING: Currently emulating unsupported memory accesses **
17:02<@caker> ** in /lib/tls glibc libraries.
17:02<fo0bar>oh wait
17:02<@caker>that's the 86.1 image
17:02<fo0bar>but you're using the 86.2 initrd, correct?
17:03<fo0bar>I could have sworn I modified that to remove /lib/tls before the pivot_root
17:03<@caker> ** Offending process: chroot (pid=1) **
17:03<fo0bar>one of the first things the main startup scripts did was remove /lib/tls, but it was still not the first process
17:03<fo0bar>one sec
17:05<afv-13>got that kernel panic again :(
17:05<@caker>afv-13: doing what?
17:05<@caker>afv-13: which kernel?
17:05<afv-13>it's right after something about squid
17:06<@caker>afv-13: ouch
17:09<@linbot>New news from forums: Announcement: Xen Public Beta in Xen Public Beta <>
17:10<@caker>afv-13: that looks like a bug in 2.6.15-vanilla... I'll try to get a 2.6.16-um kernel out asap
17:10<fo0bar> <-- caker
17:10<@caker>fo0bar: move that one into place?
17:11<fo0bar>I could have SWORN I did this "way back" in 86.2
17:11<fo0bar>oh well
17:11<@caker>18:11:16 ERROR 403: Forbidden.
17:11<kvandivo>"Xen is feature complete! " that's quite a statement
17:11<fo0bar>mksquashfs likes to make images 700
17:12|-|valen2 [] has joined #linode
17:12<@caker>fo0bar: will that initrd work on the rest of the hosts?
17:13<afv-13>thanks caker
17:17<@caker>fo0bar: ok, that initrd is on all the hosts now
17:18<@caker>fo0bar: worked, no tls message
17:18<fo0bar>huh, it dies out after kernel load when I try to boot finnix on the xen host
17:19<@caker>fo0bar: oh, you need to set the profile to use the finnix kernel
17:19<@caker>fo0bar: it got changed when you migrated
17:19<fo0bar>I thought I did
17:19<@caker>everything gets defaulted to the xen kernel during uml->xen migration
17:20<fo0bar>ok, let's try that again
17:21<fo0bar>holy CRAP that was fast
17:24<fo0bar>caker: I don't know if this is xen-specific (I don't think I ever tried it on a UML linode)... but while booted into finnix, I clicked that "reboot" button in the LPM. it shut down finnix normally, then booted my "default" OS
17:25<fo0bar>err, the "reboot" on the right
17:25<@caker>fo0bar: buttons on the right boot the default profile, ones in the main body boot the selected profile
17:25<@caker>that default setting is kinda useless
17:26<fo0bar>you get fo0bar's Seal of Intuitiveness! ;)
17:26[~]mikegrb seals fo0bar in a glass jar
17:31<fo0bar>ok, this is kickass
17:31|-|lucca [] has joined #linode
17:31<fo0bar>finnix in xen mode + fast host = instant bootup
17:32|-|hydride [] has joined #linode
17:32<hydride>are you guys gonna offer a bsd package any time soon?
17:33|-|valen2 [] has quit [Quit: System/IP migration.]
17:33|-|hydride [] has quit [Quit: ]
17:34|-|hydride1 [] has joined #linode
17:34|-|hydride1 changed nick to hydride
17:38<fo0bar>caker: BTW, *thank you* for putting ^D in lish
17:38<@mikegrb>he didn't do that, bish
17:38<@mikegrb>I did
17:38[~]fo0bar just logged out of 17 of his 24 open putty windows :)
17:38<@mikegrb>and you are welcome
17:38<fo0bar>mikegrb: oh, you suck.
17:39<fo0bar>mikegrb: what did you do? install a Readline perl module?
17:39<@mikegrb>we were already using one
17:39<@mikegrb>had to modify the perl module
17:41[~]warewolf fumes at his now broken website layout
17:59|-|afv-13 [] has quit [Quit: work in 3 hours, might as well sleep to pass the time]
18:02<@caker>hydride: probably not -- there aren't decent tools for Linux to manage bsd filesystems
18:04|-|snpsnp [] has joined #linode
18:05<lucca>Hm, anyone familiar with migrating images between datacenters?
18:05<lucca>I got the email saying shutdown, resize, and hit the button
18:05<lucca>but... I can't find the button?
18:06<@caker>lucca: for a xen beta test?
18:06<lucca>yes, exactly
18:06<@caker>ok.. logout and log back in, I must have missed yours
18:07<lucca>ah, that made a difference, thanks!
18:23<@linbot>New news from forums: resizable console in Xen Public Beta <>
18:27|-|adamg [] has quit [Ping timeout: 480 seconds]
18:28|-|snpsnp [] has left #linode [Plink!!]
18:29|-|FireSlash [] has joined #linode
18:33[~]gpd wonders why he is not prepared to switch IP address
18:33<gpd>What is the downtime during the switch? Doesn't change of primary DNS servers takes 24-48 hours?
18:34<lucca>dns issues vary depending on registrar and when they batch dump
18:34<gpd>any idea for godaddy?
18:35<lucca>sorry, no...
18:35<gpd>do you have secondary DNS and backup MX mail on another server
18:35<lucca>well, my linode Is the backup dns :p
18:37<gpd>yeah - same with me - so when the IP changes everything is broken until the DNS change propogates - which could be 2 days!
18:37<gpd>or am I just being paranoid?
18:39<lucca>well in my case I just make sure no production services other than dns are pointing at the linode, whack the button, wait for dns changes to propagate, then move them back
18:39<lucca>but yes, if your linode is your primary webserver, mx, etc, things can get inconvenient
18:39<lucca>one thing you can do
18:39<gpd>I think I'll wait for Xenification of Dallas... should be 3 weeks
18:40<gpd>go on...
18:40<lucca>is make sure your non-DNS services, like http, mail, etc
18:40<lucca>are all specified with very low TTL
18:40<lucca>i.e. 5minutes or similar
18:40<lucca>and make Sure you have two dns's and the one not being moved is up-to-date
18:41<lucca>then when you move, you can rapidly update all of your A/CNAME records to the new ip (using the other dns)
18:41<lucca>people won't see your moved dns immediately, but that's why there are two
18:47|-|nybble [] has joined #linode
18:49|-|nybble [] has quit [Quit: ]
19:02|-|jekil [~alessandr@] has quit [Quit: Leaving]
19:10|-|khedron [] has joined #linode
19:11[~]khedron wonders if there is an issue w/ host27
19:11<@caker>huh .. pings, but no response to commands
19:12<@caker>Mar 30 19:55:41 host27 kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000020
19:22<@linbot>New news from forums: Reboot: host27 in System and Network Status <>
19:31<khedron>why the extra restart?
19:32<@caker>kernel upgrade
19:33<khedron>yay, it's coming back up -- thanks caker
19:42<khedron>huh -- boot froze during inetd, but a reboot fixed that
20:02<anderiv>hey all - quick question. I'm normally a gentoo guy, but have started playing around w/ ubuntu for desktop stuff. Ubuntu's loading a bunch of kernel modules at boot, some of which I'd like to prevent it from loading. How would one do this in ubuntu? I've checked /etc/modules and there are only a few modules listed there...
20:02<anderiv>...there must be some other daemon that's running at boot to load up the rest that aren't listed in /etc/modules.
20:03<khedron>look at udev
20:04<gpd>what kind of modules do you not want loaded?
20:06<anderiv>specifically the snd_emu10k1 driver. I have another USB audio device that I want to work with, and would rather just get the "normal" sound card out of the way.
20:06<anderiv>I would pull the sound card out, but right now, I'm about 15 miles away :-)
20:07<gpd>you can set the default alsa card in gnome
20:07<gpd>but i accept your point
20:08<anderiv>gpd: would I have to reboot after changing the default sound card for it to take effect?
20:08<gpd>not if you do it through the gnome admin panel
20:09<anderiv>okay - that's what I did then.
20:09<gpd>you have to restart the application probably
20:09<gpd>and some apps can set the sound card within
20:09<gpd>eg. VLC
20:12|-|FireSlash [] has quit [Ping timeout: 480 seconds]
20:13<gpd>anderiv: if you are 15 miles away - why do you want sound anyway? using a usb microphone to spy on your spouse?
20:13[~]gpd ducks
20:15<anderiv>heh - good point :-)
20:16<anderiv>FYI I'm at home w/ the spouse currently :-)
20:18<gpd>You can also blacklist modules to prevent them being loaded... but this requires a reboot
20:19<gpd>but you could rmmod them for now i suppose
20:19<anderiv>gpd: I just discovered the blacklist...
20:20<anderiv>that's what I did and I'm currently rebooting.
20:22|-|Kelsey23 [] has left #linode [Leaving]
20:24<anderiv>heh - well I've done it now...she's not coming back up. Doh.
20:24<anderiv>I guess I won't be working on this any more tonight :-)
20:24<gpd>might be a fsck?
20:24<gpd>i think default is 30 reboots
20:25<anderiv>nah - the box was just installed this afternoon.
20:26<anderiv>to explaing further...I picked up a Griffin Radio Shark and am trying to get that working in ubuntu.
20:27<anderiv>yep - that's it.
20:27<gpd>Tivo for radio
20:27<anderiv>yep - exactly.
20:27<gpd>drive over and get it?
20:28<gpd>is there anything of AM/FM worth recording?
20:28<anderiv>nah - it's raining cats and dogs...
20:28<anderiv>gpd: mostly public radio programs.
20:28<anderiv>...that's about all I'm interested in.
20:28<gpd>you can get a few via podcasts - but not many I suppose
20:28<anderiv>yes - that and MN public radio.
20:28<anderiv>mostly the weekend shows.
20:29|-|khedron [] has left #linode []
20:30<gpd> <-- keep you amused until tomorrow ;)
20:31<anderiv>heh w00t. The box had just grabbed a different addy from DHCP. I checked the DHCP server's logs and tried the last few addresses it had handed out.
20:31<gpd>trying to hide from you no doubt
20:32<anderiv>FYI - looks like blacklisting snd_emu10k1 worked like a charm. Thanks.
20:32<gpd>xlnt - off to play squash - bye
20:38|-|nenolod [] has joined #linode
20:38<nenolod>how would I go about reporting an abusive IRC user on linode?
20:39<@caker>nenolod: msg me, please
20:39<@caker>or better yet, send an email to abuse .... at ...
20:39<nenolod>caker: ta
20:41[~]encode hopes nenolod doesnt find him abusive
20:41<nenolod>encode: no idea who you are :>
20:42<encode>that makes two of us
20:53<taupehat>I think I've lost the ability to move =[
20:53<taupehat>my house is a stark vista of cardboard skyscrapers
20:53<taupehat>on a happier note, that's the lot of the lot, so now it's time to begin unpacking
20:54<taupehat>and the xenode is performing very nicely
20:57<nenolod>yeah, how are the new xen-based nodes performing? i might get myself one for experimentation
20:58<taupehat>I'm approaching 24 hours on mine
20:58<taupehat>works fine
21:00|-|flatronf700B [~flatronf7@] has joined #linode
21:06|-|valen [] has joined #linode
21:06|-|valen changed nick to valen2
21:12<nenolod>taupehat: groovy
21:25<hydride>hey, I heard there was talk of releasing a freebsd package
21:25<hydride>is this true?
21:25<efudd> <- amazing
21:25<xptek>That would be orgasmic.
21:25<hydride>a freebsd package?
21:25<hydride>I know
21:26<hydride>I'm looking at the site, it sounds cool, I'll do it if I can get it in FreeBSD
21:27<xptek>The Linux ones aren't bad at all either.
21:27<hydride>I used linux for a while back in the day
21:27<hydride>but I'd rather stick with one system
21:28<xptek>Yeah, I'd probably get another Linode again if they offered FreeBSD, but there's no way to do FreeBSD with Xen (That I know of) or UML.
21:29<hydride>so it's not possible for them to offer a FreeBSD package?
21:29<xptek>It is, but they'd have to create something to manage FreeBSD jails or something similar.
21:29<hydride>Someone I go to school with has an account here and he mentioned something about a future FreeBSD package
21:31<xptek>Bleh, I'm not sure how much freedom FreeBSD jails would give you anyway
21:31[~]xptek shrugs
21:31<hydride>if I got a Slack linode and they released a freebsd version, would I be able to switch?
21:32[~]xptek points to caker and mikegrb
21:32<hydride>are they even here?
21:32<@caker>efudd: very cool -- does that thing have a mini turbine in it?
21:32<efudd>no. that's a a large scale electric in fact.
21:32<efudd>the sound you heard was primarily just the blades vs. the air.
21:33<@caker>BSD Xenodes are only a possiblity when there are tools for Linux to manage BSD filesystems
21:33<efudd>(there aren't tools on linux to manage BSG filesystems?)
21:33<@caker>none that are r/w
21:33<hydride>so that's a negative on a future bsd package?
21:33<@caker>that I've found, anyway .. last I looked was a few weeks ago
21:33<efudd>Well, BSD and Linux are very.. well.. "seperate camp".
21:34<efudd>makes sense that no one has fully crossed that gap.
21:34<@caker>hydride: I think people are working on them ... and yes, you'd be able to change between Linux and BSD
21:36<efudd>XM radio now has a new channel, 84, "Chill Techno"
21:36<efudd>i'm digging.
21:36[~]caker gave up on XM a while ago
21:36<efudd>in favor of...?
21:36<@caker>nothing .. FM/ipod/cds
21:36<efudd>I can not, and will not, EVER, go back to FM/AM. :)
21:36<hydride>I like having my xbox which can just tune in to any online radio station
21:37<@caker>I expected the "24 hour Descendents" channel, etc
21:37<@caker>it just doesn't work that way
21:37<efudd>ipod/cds is too.. well, static for my tastes. What I'd really like is just the ability to stream to my car :)
21:37<hydride>you guys don't offer domains do you?
21:37<efudd>caker, i almost gave up on XM during the christmas season. I listen to just channel 82 (techno).. they replaced it with christmas music..
21:37<efudd>I caled to voice my opinion.. they said "We can turn off your radio until december 25th."
21:37<@caker>hydride: no, that's a crowded marketplace, anyhow
21:37<efudd>.. which just pissed me off. :)
21:38<@caker>efudd: hah!
21:38<@caker>efudd: and I assume credit you during that time?
21:38<efudd>no. just turn it off and not charge me.
21:38<efudd>so "credit" by proxy of that. sure.
21:38<efudd>honestly, I wanted a free month. Just cause. Call it customer service. I honestly didn't listen to it....
21:39<efudd>hrm. 5m/d average on ratemylinode shos me at a grand total of 2 ios.
21:39<efudd>remind me to churn later.
22:03<taupehat>highly amusing
22:08|-|schultmc [] has quit [Ping timeout: 480 seconds]
22:20|-|adamg [] has joined #linode
22:59|-|VS_ChanLog [] has left #linode [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #linode
22:59<@caker># ./ | grep TLS
22:59<@caker>Host TLS support detected
22:59<@linbot>New news from forums: Kernel: with NPTL/TLS support in Announcements <>
23:04<lucca>I'd love to test that out
23:05<lucca>I have an app that needs NPTL in order to use threads
23:05<@linbot>New news from forums: NPTL in Feature Request/Bug Report <> || binfmt in Xen Public Beta <>
23:05<@caker>lucca: well, Xen already does nptl...
23:05<lucca>(SBCL, an ANSI Common Lisp implementation)
23:05<@caker>unless you have another node
23:05<lucca>well I'll get that cooking right away
23:05<encode>dont forget to preheat the oven
23:06|-|vodka_ [] has joined #linode
23:07|-|vodka [] has quit [Ping timeout: 480 seconds]
23:09<lucca>actually I'll re-expand the fs so I'm not at 100%, heh
23:10<warewolf>I fucking hate CSS.
23:10<warewolf>I am going to bed.
23:10<warewolf>sadly, this will not solve the problem :(
23:14|-|Sgeo [] has quit [Remote host closed the connection]
23:38|-|cout [] has quit [Ping timeout: 480 seconds]
---Logclosed Fri Mar 31 00:00:48 2006