[00:53] <twb> I'm translating /etc/bind to /etc/nsd3.
[00:53] <twb> What's the nsd equivalent of "auth-nxdomain no"?  Do I even need it?
[01:41] <twb> In this named snippet: http://paste.debian.net/110811/
[01:42] <twb> Is "notify yes" applying to the master, or the (presumably downstream) zonedit nameservers?
[04:15] <^Mike> How can I run SMART tests on a HDD?
[04:17] <twb> apt-get install smartmontools
[04:17] <twb> smartctl <something to say "start a short test">
[04:18] <twb> smartctl <something to say "show self-test results">
[04:18] <twb> RTFM; I can't remember what the options are
[04:19] <Datz> something like smartcl -A /dev/sda1 I thought
[04:19] <Datz> er
[04:19] <twb> IIRC -A reports all info
[04:19] <Datz> smartctl
[04:20] <twb> smartctl -l selftest /dev/sda
[04:20] <twb> smartctl -t short # or long, if you have a couple hours
[04:21] <Datz> -A looks lit it reports all smart attributes
[04:21] <twb> You may also need -d ata for SATA disks, if you're on an old system.
[04:21] <twb> Datz: right, but most of those are misleadingly scary
[04:21] <Datz> inaccurate I've noticed for some things
[04:22] <Datz> but right, not tests
[04:25] <^Mike> thanks
[04:26] <^Mike> bah
[04:26] <^Mike> "Device does not support SMART"
[04:28] <^Mike> Well, if you can make any sense of these logs, I'd appreciate the help: http://sprunge.us/bBGh?txt
[04:30] <twb> ^Mike: maybe you're behind a RAID card
[04:30] <^Mike> I doubt it, it's just a cheap external HDD connected to the machine by USB
[04:30] <twb> ^Mike: "unable to read inode block" I reckon you've made the filesystem slightly larger than the partition -- did you dd it from one to another?
[04:31] <twb> Also your name is stupid :-/
[04:31] <twb> \bMike\b: hmm, you're getting it for several partitions, so more likely is just that either the copy is hosed or the disk is dying
[04:31] <\bMike\b> no, but I could well have messed up the partition/filesystem sizes anyways
[04:32] <twb> Oh, and you can't do SMART to a USB enclosure -- if you take the disk out and put it in a SATA slot, then SMART will work
[04:32] <\bMike\b> oh, ok
[04:32] <\bMike\b> huh, that disk also reports "Partition table entries are not in disk order"
[04:50] <`3Shades> Anyone any good at configuring virutal networks? I am having problems remoting into my virtual machine. I can ssh but not remote in.
[04:52] <`3Shades> I am running ubuntu server 10.10 and I have VM's set up and I would like to be able to remote into those as well from outside my network. I have a standard linksys wireless router.
[04:53] <`3Shades> I have googled to no avail, or at least I'm not understanding what I'm doing wrong.
[04:54] <twb> What does "remote in" mean
[04:54] <`3Shades> vnc
[04:55] <twb> Are you running a VNC server in the VM?
[04:55] <`3Shades> I was using remote desktop so...
[04:55] <`3Shades> I set up remote desktop
[04:55] <twb> "remote desktop" often means RDP, not VNC.  Make sure you which you mean
[04:55] <`3Shades> I installed gnome gui on the server then set up remote desktop
[04:55] <`3Shades> I went into admin and clicked the allow remote desktop connections...
[04:56] <twb> `3Shades: last time I looked that will display a prompt on the VM's local display saying "accept connection? [Yes|no]"
[04:56] <`3Shades> allow users to take control of this computer
[04:56] <`3Shades> yes I clicked accpet connections
[04:56] <`3Shades> well first I"m just tring to get into the host
[04:57] <twb> I don't know
[04:57] <`3Shades> See it's weird, because i can ssh in.
[04:57] <twb> If you were using a simple X VNC server, I could help.  I'm not familiar with this GNOME stuff
[04:58] <`3Shades> my dhcp table is set to have the mac address go to an ip address yet when I look up the address you should remote into it says some weird address, but either address I try I cannot remote in.
[04:58] <`3Shades> I know it has something to do with the VM bridge and settings in my linkxys router, but I am lost.
[04:59] <`3Shades> what do you recommend I install then
[04:59] <`3Shades> just vncserver
[04:59] <`3Shades> lol
[04:59] <twb> xvncserver4 IIRC
[04:59] <`3Shades> can I apt-get that
[04:59] <twb> Or just tunnel X over your ssh connection
[05:00] <twb> tightvncserver, on my system, appears to be a simple headless VNC X server
[05:00] <twb> What you have is probably exporting an existing display over VNC, which is more stupider
[05:01] <`3Shades> tightvncserver is already installed
[05:05] <`3Shades> It seems stuck at security type requested
[05:06] <twb> Oh yeah, IIRC the vinagre/vino stuff in GNOME implements some non-standard encryption option
[05:07] <twb> So if you checked "encrypt connection" or whatever, then you can't connect to it except from the GNOME VNC client
[05:07] <twb> And your client will probably just hang showing a black screen
[05:15] <`3Shades> vnc works thanks..
[05:16] <`3Shades> still need to figure out how to route these VM's to a port though
[05:16] <`3Shades> I'll wait to ask until I have another 10 hours on it lol
[08:20] <acegrimm> bonjour
[08:27] <xperia_> hello to all. i have a strange problem with memcached. it give me allways this error here => DataMemCache Error, Purge Old File from Cache Dir
[08:28] <xperia_> where is this Cache Dir of MemCached ? i have searched it but could not find it
[08:39] <twb> xperia_: caches usually go in /var/cache, occasionally /var/lib
[08:40] <whoaski> hello all
[08:42] <whoaski> I just installed ubuntu server w/ apache2, mysql, and rails. tring to configure a developmentserver to bounce code off of
[08:43] <whoaski> I'm new to this and I want to know where i can get information navigating the command line
[08:44] <whoaski> I have the ubuntu server guide
[08:44] <whoaski> it's all installed I just don't know enough to get into it to configure it
[08:45] <xperia_> twb okay thanks will just look at it. /var/cache i checked allready but could not find anything related to memcached
[08:49] <xperia_> twb hmm strange even in /var/lib i can not find memcached ! any other help
[08:51] <twb> xperia_: read the config file?
[08:51] <xperia_> not till yet i will look at it just right now
[09:00] <fakhir> I seem to have a IO bottleneck issue but i dont quite understand what it is telling me.  -> http://pastebin.com/Vz1btnbn
[09:20] <_ruben> fakhir: i assume sdb is a single sata disk?
[09:21] <fakhir> _ruben, yes
[09:21] <fakhir> what i am seeing is high load average but low CPU usage on a MySQL server.
[09:22] <_ruben> then having ~40 random IOs per second is pretty consistent with the 50% utilization
[09:22] <_ruben> a sata disk can do about 100 of 'em per second
[09:22] <_ruben> and a (busy) (my)sql server tends to do a fair bit random IOs
[09:25] <fakhir> any idea why i might be seeing what i am seeing if io is fine?
[09:31] <_ruben> how high is the load average?
[09:33] <fakhir> _ruben,  always above 1 typically ~1.5 . CPU percentage is nearly always 0.
[09:44] <_ruben> iowait is only 6% btw .. or does it go up a lot?
[09:45] <_ruben> and a load of 1 on a 8way system isn't really that much of problem
[09:45] <_ruben> 1 core is "occupied" while the other 7 are idle (on average)
[09:46] <fakhir> ohh i see
[09:56] <Daviey> twb, Interesting that you prefer to maintain your own netfilter rules.
[09:56] <Daviey> Why is that?
[09:57] <twb> Daviey: you mean with respect to blacklisting?
[09:57] <twb> It's a static ruleset
[09:57] <twb> http://paste.debian.net/110844/ is the active ingredient
[09:58] <Daviey> twb, I have a call now, but in an hour or so - i'd like to talk to you further about it.
[09:58] <twb> It'll have to be tomorrow, or trentbuck@gmail.com -- I'm going home shortly
[09:59] <soren> twb: I've never worked out what the difference is between --state and --ctstate.
[09:59] <soren> I guess I never tried to work it out either. :-/
[09:59] <twb> soren: ctstate is "better"
[09:59] <Daviey> twb, ok, thanks!
[10:00] <twb> I *strongly* recommend anyone writing netfilter rules read http://jengelh.medozas.de/documents/Perfect_Ruleset.pdf
[10:00] <twb> ...which is cited by #netfilter
[10:07] <soren> twb: Yeah, the good mr. Engelhardt knows what he's talking about.
[10:07] <soren> twb: The only thing in that doc that I didn't know was the ctstate thing. I wonder why the --state thing isn't just replaced entirely by --ctstate.
[10:08] <twb> Probably hysteria
[10:08] <soren> Yeah, there's a lot of that going around.
[10:09]  * soren is excited about ipset being included in 2.6.39, by the way.
[10:10] <twb> Eh, DKMS xtables makes that a non-issue
[10:11] <twb> I need xtables for something else anyway, I forget what
[10:11] <twb> Maybe just -j CHAOS
[10:12] <soren> twb: There's a dkms variant?
[10:13] <twb> It's new in Debian
[10:13] <soren> twb: Or are you thinking of the module-assistant thing?
[10:13] <twb> Maybe you suckers don't have it yet
[10:13] <soren> twb: Oh, shiny.
[10:14]  * soren files sync request
[10:15] <soren> done
[10:15] <soren> \o/
[10:15] <soren> twb: Thanks for the hint.
[10:15] <twb> np
[10:15]  * soren smells an openstack patch brewing
[10:16] <twb> Meanwhile flipping nsd3 definitely *will not* DTRT WRT upstart
[10:16] <soren> twb: I didn't fully understand your problem with that.
[10:18] <twb> one moment
[10:23] <twb> soren: http://paste.debian.net/110845/
[10:25] <soren> twb: It gets a new PID when you HUP it? That sounds *incredibly* broken.
[10:25] <twb> YES
[10:26] <twb> soren: it *is* smart enough to update its own pidfile
[10:26] <twb> But even so
[10:47] <twb> If you don't specify an allow-transfers {} list, does bind9 default to allowing AXFRs from 0/0 (i.e. everyone)?
[10:49] <twb> It sure looks like it, which explains a lot
[11:03] <twb> Fuck, I missed tea-time
[11:03] <twb> Looks like I'm having ramen for supper again
[11:04] <soren> "ramen"?
[11:05] <soren> Ah, noodle stuff.
[11:08] <whoaski> good morning
[11:10] <ikonia> twb: easy on the language please
[11:12] <twb> Sorry.
[11:16] <azizLIGHTS> how do i run dropbox as a service? this guide is outdated because start-stop-daemon doesnot exist on 10.04? http://wiki.dropbox.com/TipsAndTricks/TextBasedLinuxInstall/UbuntuStartup
[11:17] <azizLIGHTS> /etc/init.d/dropbox: 60: start-stop-daemon: not found
[11:21] <joschi> azizLIGHTS: it does. it's part of the 'dpkg' package
[11:21] <joschi> azizLIGHTS: you just need either a proper $PATH variable set or use the absolute path to the binary
[11:22] <azizLIGHTS> oh hm
[11:22] <azizLIGHTS> do i have to undo step 2 listed in the guide to edit /etc/init.d/dropbox ?
[11:22] <azizLIGHTS> specifically this line: sudo update-rc.d dropbox defaults
[11:23] <joschi> azizLIGHTS: why would you?
[11:23] <joschi> azizLIGHTS: the "init script" is also lacking a proper shebang
[11:23] <azizLIGHTS> i dont know im new to linux?
[11:23] <azizLIGHTS> what is that
[11:24] <azizLIGHTS> what is the proper shebang i mean
[11:24] <joschi> azizLIGHTS: the first line of a script should contain a pointer to the interpreter which will run the script, e. g. #!/bin/sh to use /bin/sh
[11:24] <azizLIGHTS> oh
[11:24] <joschi> azizLIGHTS: you could also easily use the upstart-script from the dropbox wiki
[11:24] <azizLIGHTS> ugh
[11:25] <azizLIGHTS> so this is wrong?
[11:30] <azizLIGHTS> joschi: ok well putting full path /sbin/start-stop-daemon, and adding #!/bin/sh to first line, now the script works with service dropbox start, confirmed it with ./dropbox.py status
[11:32] <Wise_> I'm trying to create a disk image of my ubuntu install, with clonezilla... but it's telling me that the partitiont able in the disk is illegal/invalid, not supported by parted, with the error "can't have overlapping partitions", anyone know how I deal with this?
[11:32] <Wise_> err, partition table* -_-
[11:37] <azizLIGHTS> ok i edited the dropbox wiki for it, ty for input joschi
[12:15] <MarkAng> How can I see and change the port used by my Ubuntu server to access the internet?
[12:16] <MarkAng> We're planning on using it to create and distribute images to computers using PXE.
[12:17] <MarkAng> But it isn't able to reach the ubuntu archives, like the ones specified in the sources.list file.
[12:18] <joschi> MarkAng: that'll be on destination port 80/tcp (http). the source port is random
[12:19] <MarkAng> The reason I'm asking is that we're operating under a company-wide network that has restrictions on certain ports. For instance, the port used by uTorrent is blocked. Is it possible to instruct my server to use or not use a certain port?
[12:21] <MarkAng> Anyone?
[12:22] <_ruben> if they're blocking based on source port, hit 'em with a cluebat, a lot
[12:22] <_ruben> if they're blocking port 80, you should probably use the proxy they provide
[12:23] <MarkAng> They're not blocking port 80, as I'm able to access web pages. \
[12:24] <shauno> do you have a specific error from trying to reach the archives?
[12:24] <_ruben> what does your sources.list look like then? those tend to list http urls
[12:25] <MarkAng> They do. But whenever we issue a command that should use the repositories inside to update and/or install programs (in this case, a GUI for the text-based server)
[12:26] <MarkAng> t says it cannot reach the destination files.
[12:26] <MarkAng> Cannot reach/cannot find.
[12:26] <pmatulis> "a GUI for the text-based server"?
[12:27] <MarkAng> Yes. It's command-line as you know, and we know there are GUI skins available.
[12:28] <MarkAng> We try to install one to make life a little easier for us, but so far no good.
[12:28]  * Pici scratches his head
[12:28] <Pici> Which package are you trying to install?
[12:28] <MarkAng> Hang on, let me check
[12:28] <MarkAng> ..
[12:28] <pmatulis> MarkAng: you most probably need to configure a proxy for apt-get
[12:28] <MarkAng> We use either of the following commands, but neither one works:
[12:28] <MarkAng> sudo apt-get install xubuntu-desktop
[12:29] <MarkAng> sudo aptitude  install --no-install-recommends ubuntu-desktop
[12:31] <MarkAng> Both commands should fetch a desktop-like environment and overlay it over the command-line.
[12:31] <MarkAng> But like said before, they don't work.
[12:33] <MarkAng> Guys?
[12:34] <pmatulis> MarkAng: well.  is your company using a proxy or not?
[12:34] <MarkAng> I think they are, yes. Im an IT intern, so I'm not sure what it is.
[12:34] <pmatulis> https://help.ubuntu.com/community/AptGet/Howto#Setting%20up%20apt-get%20to%20use%20a%20http-proxy
[12:37] <Caribou> you would need a statement like "Acquire::http::proxy "{proxy URL}"; in /etc/apt/apt.conf file
[12:44] <MarkAng> It's a large organisation, spanning multiple cities and with hundreds of schools. The only persons that could tell me what the proxy is are out of reach.
[12:46]  * _ruben puts a sticker MarkAng's server: Warranty void if desktop packages are installed
[12:52] <Caribou> MarkAng: don't you have a web browser somewhere that would be configured with the same proxy ?
[12:59] <MarkAng> I just asked an admin that returned to the office, and we have no proxy.
[13:00] <pmatulis> MarkAng: confirm that your server can reach the internet in general?
[13:01] <Caribou> you might want to try 'wget http://www.google.com' to see if you can hit the Internet
[13:03] <pmatulis> Caribou: are you dumming down my comments or do you not see what i'm writing?
[13:04] <Pici> w3m is installed by default I believe.
[13:04] <Caribou> no, I'm just adding a suggestion to your comment : he could confirm that he can reach the internet by doing that. Sorry if it seems rude
[13:05] <pmatulis> Caribou: alright
[13:28] <RoAkSoAx> morning all
[13:30] <SpamapS> RoAkSoAx: howdy
[13:31] <RoAkSoAx> SpamapS: how's it going man, up early today huh?
[13:37] <RapidTraffic> Webmasters, you interested in some HQ, 24h ip unique, content and geotargeted traffic?
[13:39] <SpamapS> RoAkSoAx: always.. the baby knows just when to wake me up so I can't go back to sleep :p
[13:43] <RoAkSoAx> SpamapS: look the bright side.. you can now say you wake up earlier than everybody else her ein the US >P
[13:43] <RoAkSoAx> (everybody else on the team)
[13:50] <zul> SpamapS: its genetic ;)
[14:04] <medberry> Caribou, howdy
[14:05]  * Caribou waves back
[14:39] <yann2> hello! what is the most appropriate package to provide the command "mail" - mailutils?
[14:41] <yann2> there is also bsd-mailx...
[14:41] <hallyn> i think i usually install mailutils
[14:42] <hallyn> dunno why.  works, though.
[14:42] <soren> I used to use mailx.
[14:42] <genii-around> yann2: It used to be something like bsd-mailutils but I think just the mailutils now
[14:42] <soren> In acient times, at least.
[14:42] <soren> For the last decade, I've used mutt, though.
[14:42] <soren> ..but if I needed "mail", I'd use mailx. No particular reason.
[14:42] <yann2> soren, so used I, but http://pastealacon.com/27001
[14:43] <soren> yann2: bsd-mailx == mailx, IIRC.
[14:43] <patdk-wk> well, mailx == rh mail, and most other peoples mail command
[14:44] <yann2> ok, I ll go for that
[14:44] <yann2> thanks
[15:05] <bricas> I set up some logrotate config files for my web server logs (monthly rotation), it's been running well over the last 6 months -- except for some reason it decided to rotate things yesterday (15th). Anyone else experience this? Any hints as to how I might diagnose why?
[15:18] <Daviey> bricas, Depends on the rotate rule.... Did it rotate on filesize, rather than date?
[15:20] <bricas> Daviey: http://nopaste.snit.ch/37851 # changed real domain to example.com for the paste, but everything else is straight copy+paste.
[15:24] <Daviey> bricas, how odd... Nothing obvious is jumping out at me.
[15:25] <bricas> Daviey: yeah, i didn't think my config was off -- like i said, it's been running well for the last 6 months. on the 1st of each month, it seems to fire.
[15:25] <bricas> Daviey: there's no obvious place that would log something like "logrotate running .. rotating log $foo because of $bar" ?
[15:26] <Daviey> not really..
[15:27] <bricas> of course not :)
[15:48]  * patdk-wk teachs bricas to use symlinks :)
[15:50] <bricas> patdk-wk: ...? rather than a copy of the log, perhaps? :)
[15:50] <patdk-wk> ya :)
[15:51] <bricas> patdk-wk: i only did a copy because i was afraid someone was going to mess that file up, honestly. paranoid am i. :)
[15:58] <hallyn> Daviey: if you'd like to chat sometime about https://blueprints.launchpad.net/ubuntu/+spec/ubuntutheproject-community-n-server, please ping me.
[16:00] <RoAkSoAx> hallyn: hallyn include me on your conversation ^^
[16:04] <Daviey> hallyn / RoAkSoAx: Yes!  That would be wonderful.. I'm so pleased you are both enthusiastic about that one. :)
[16:09] <RoAkSoAx> Daviey: :)
[17:34] <SpamapS> http://wiki.spamaps.org/
[17:40] <patdk-wk> 1.7sec page load time :(
[17:43] <jiboumans> soren, ttx, are you guys likely to head the celery way for messaging? http://wiki.openstack.org/UseCeleryAsQueueManager
[17:52]  * RoAkSoAx goes out for lunch
[18:17] <SpamapS> patdk-wk: no surprise when your mysql db is on a t1.micro ;)
[18:25] <patdk-wk> SpamapS, actually, it's more of an issue of only using 1 dns name for the server :)
[18:25] <patdk-wk> 500ms per request, and they are getting delayed :)
[18:30] <SpamapS> so its not 1 dns name, but that you didn't have the dns name cached
[18:31] <SpamapS> my resolver responds in 50 - 100 ms once its cached..
[18:31] <SpamapS> your browser shouldn't be requesting it multiple times
[18:35] <_ruben> hm, a dig +trace wiki.spamaps.org doesn't go beyond the .org servers
[18:39] <zul> jiboumans: probably not but check on #openstack
[18:39] <jiboumans> zul: will do
[18:48] <hallyn> kirkland: around ?
[18:49] <hallyn> kirkland: the vgabios package's links file is still sending links to /usr/share/kvm.  is tehre a reason for that?  or can i fix that to /usr/share/qemu?
[18:49] <hallyn> kirkland: bc otherwise if you isntall vgabios first and then qemu-kvm, then qemu-kvm doesn't find vgabios
[18:50] <kirkland> hallyn: howdy
[18:51] <kirkland> hallyn: please put it in /usr/share/qemu
[19:05] <orudie> what is the minimum ram required for ubuntu server 64 bit ?
[19:06] <genii-around> orudie: I have a box here with 192Mb it runs on
[19:07] <cloakable> I've seen a minimal install use 32Mb
[19:07] <orudie> so 64 bit is recommended ?
[19:08] <genii-around> if your cpu is 64 bit
[19:08] <_ruben> 128MB seems to be the documented minimum
[19:51] <kirkland> hallyn: you've been on a bug triaging tear :-)
[19:51] <kirkland> hallyn: my bug-mail inbox is *full*  :-)
[19:53] <kirkland> hallyn: also, i see the vgabios merge proposal ... don't you have upload rights to vgabios?
[19:53] <kirkland> hallyn: or do you just want a sanity check?
[19:58] <geekbri> question, is there something lightweight that can be installed if i only want to do outgoing mail from my server and dont care about incoming mail?
[20:02] <SpamapS> geekbri: smail might be what you're looking for
[20:05] <geekbri> SpamapS: thanks
[20:16] <jsemar> I need to set up an smtp server to act as a relay for my ec2 instances, can one of you guys point me to a decent tutorial for that?
[20:18] <cloakable> apt-get install postfix
[20:18] <cloakable> Follow directions.
[20:19] <jsemar> thanks cloakable
[20:19] <jsemar> thats awesome
[20:19] <cloakable> jsemar: ubuntu is pretty easy to use :)
[20:20] <jsemar> yeah i know, but your answer is useless, obviously i can install postfix, but i need for the ec2 instance to have its postfix .cnf file send mail to my postfix installation on probably linode
[20:21] <jsemar> so theres authentication, and all sorts of things besides 'sudo apt-get install postfix'
[20:22] <cloakable> By default ubuntu postfix uses system accounts. You want smarthosting.
[20:22] <_ruben> well, you didn't mention authentication in the first place
[20:23] <cloakable> Also, does your linode have a static ip? Use that for authentication.
[20:23] <_ruben> no experience with ec2, but if the ips are static, you can just use those as restrictions
[20:23] <cloakable> yea
[20:29] <jsemar> you wouldnt recommend using sasl/smpt auth
[20:32] <_ruben> depends on the scenario
[20:40] <geekbri> damnit, is there anyway to find out why another mail server is refusing your connection... I know i must be missing something that is causing it to make me marked as spam and refusing my connection but they just say error 550 connection refused
[20:41] <_ruben> contact its postmaster
[20:46] <Patrickdk> heh, ip's for ec2 are static per instance
[20:47] <Patrickdk> so if you use ebs fs's, and stop/start, ip will change
[20:48] <Patrickdk> kind of like a cable modem :)
[20:48] <Patrickdk> as long as it doesn't go down, your likely to have the same ip forever
[20:48] <Patrickdk> but no guarrentee :)
[20:54] <geekbri> well yeah that i know
[20:54] <geekbri> but i have an elastic IP assigned to the instnace
[20:56] <SpamapS> jsemar: given that your instances will be ephemeral, you'll probably want to use sasl+tls
[20:57] <SpamapS> jsemar: Its similar to the config I use for my laptop to relay through a smart host.
[20:57] <RoyK> natty installation was quite nice, btw, at least on the desktop - it starts installing in the background before asking stupid questions like username/timezone etc :)
[20:57] <SpamapS> geekbri: EC2 ips, even the elastic IPs, are pretty much banned by now
[20:57] <SpamapS> too much abuse out there.
[20:58] <SpamapS> RoyK: I think Maverick did that too..
[20:58] <geekbri> i guess that makes sense... its just unfortunate when you want to send automated emails to yourself :)
[20:58] <SpamapS> does.. shouldn't refer to the stable release in the past tense.. ;)
[20:58] <RoyK> SpamapS: oh - dunno - haven't installed a maverick desktop yet - only upgraded to it...
[21:01] <geekbri> you know whats nice?
[21:01] <geekbri> ubuntu on ec2, i haven't found a server distro with as much support as it yet.
[21:23] <ccm> Hi there
[22:03] <hallyn> kirkland: I don't yet have the upload rights
[22:03] <hallyn> (I test with a vmbuilder push that I want to do every morning :)
[22:39] <kirkland> hallyn: still around?
[22:40] <hallyn> yup
[22:41] <hallyn> kirkland: what's up?
[22:41] <kirkland> hallyn: just sent you the kvm auto install magic
[22:41] <hallyn> ah, thanks.
[22:41] <kirkland> hallyn: essentially, it's:
[22:41] <kirkland> boot an ubuntu server ISO and append "priority=critical locale=en_US url=http://bit.ly/ubuntu-nqa" to the kernel  command line
[22:42] <kirkland> hallyn: noting that we might need to tweak the preseed file which is pointed to by url=http://bit.ly/ubuntu-nqa"
[22:42] <kirkland> nqa is my acronym for "no questions asked"
[22:42] <hallyn> and so i should test it and send it on to the kvm-autotest folks?
[22:43]  * hallyn is watching grub compile and his memory is failing him
[22:45] <kirkland> hallyn: right
[22:46] <kirkland> hallyn: what i've never quite been clear on
[22:46] <kirkland> hallyn: is what/how they want to call this
[22:46] <kirkland> hallyn: note that the preseed pointed to by that bit.ly link should probably go into kvm-autotest's source control
[22:46] <kirkland> hallyn: and it'll need to be web-retrievable from somewhere, perhaps a raw git print
[22:46] <kirkland> hallyn: and you can shorten that with bit.ly or the like
[22:47] <kirkland> hallyn: does that make sense?
[22:47] <hallyn> are bit.ly links permanent?
[22:47] <oraqol> hey yall, quick question, what exactly is Ubuntu Enterprise Cloud?  Is it a virtualization platform?  I've got a quadcore amd 6 gig rig, is that enough to support it?
[22:47] <hallyn> (that is, until .ly goes down :)
[22:48] <kirkland> hallyn: yeah
[22:48] <kirkland> hallyn: sure
[22:48] <kirkland> hallyn: full urls are fine too in there
[22:48] <kirkland> hallyn: and if it's called by a script, there's no need to obfuscate it behind a url shortener
[22:49] <hallyn> kirkland: ok, i'll take a look at that next week.  (i'm out rest of this week).  thanks
[22:49] <kirkland> hallyn: if you're typing it in by hand (which i do often do to quickly get a vm installed, hands off)
[22:49] <kirkland> hallyn: it's nice to have a bit.ly link :-)
[22:49] <kirkland> hallyn: sure
[22:49] <hallyn> you control your dns don't you  so you could shorten p.c.c/~kirkland/preseed.cfg
[22:49] <kirkland> hallyn: heh
[22:49] <kirkland> hallyn: i reckon i could
[22:50] <kirkland> hallyn: i'm at austin java, at the moment though
[22:50] <kirkland> hallyn: and i haven't pwned their dns....yet.... :-)
[22:50] <hallyn> ah   been back to thunderbird?
[22:50] <kirkland> hallyn: i haven't, actually
[22:50] <kirkland> hallyn: i just finished the bag of beans i bought there, though, so i was thinking about it yesterday
[22:50] <kirkland> hallyn: i didn't pick you to be one for march madness
[22:51] <hallyn> huh?
[22:51] <hallyn> you'd have been right
[22:51] <kirkland> *robbiew is taking the rest of the week off to watch basketball :-P
[22:51] <RoAkSoAx> kirkland: maybe magic can be put into cobbler to install images with koan using that presseed
[22:51] <hallyn> oh :)
[22:51] <kirkland> RoAkSoAx: either there, or in orchestra
[22:51] <hallyn> unrelated :)
[22:51] <kirkland> hallyn: i know, i was kidding
[22:52] <kirkland> hallyn: i assume you care none whatsoever for NCAA basketball :-)
[22:52] <hallyn> right, but apparently our esteemed prez does
[22:53] <oraqol> no love?
[22:53] <hallyn> so i found 'conboy' for the n900 today.  it does tomboy format, and syncs to ubuntu one!  yay
[22:53] <hallyn> oraqol: kirkland could answer you best i think :)
[22:53] <RoAkSoAx> kirkland: definitely in orchestra. But with koan we can easily install KVM images using it on the host, but right now the use of koan to install Ubuntu images is broken :)
[22:53] <kirkland> oraqol: what's your question?
[22:54] <hallyn> (just hit up-arrow twice i think)
[22:54] <oraqol> is it a virtualization platform?
[22:54] <hallyn> uec
[22:54] <kirkland> oraqol: UEC = Ubuntu Enterprise Cloud
[22:55] <kirkland> oraqol: i suppose you could call it a virtualization platform; your hardware is enough to support it, yes
[22:55] <hallyn> oraqol: it's intended for making clusters.  what is your intended use?
[22:55] <kirkland> oraqol: https://help.ubuntu.com/community/UEC/CDInstall#STEP%201:%20Prerequisites
[22:55] <oraqol> right now im running two vms on vanilla ubuntu server, one of which is a pbx, should i just be using uec as the host os?
[22:56] <kirkland> hallyn: i'll commit/upload vgabios now, if you like
[22:56] <kirkland> oraqol: do you often need to spin new vm's up and down?
[22:57] <oraqol> no, they tend to be always running
[22:57] <hallyn> kirkland: please do, thanks
[22:57] <kirkland> oraqol: probably not necessary in your case
[22:57] <kirkland> oraqol: kvm + libvirt is probably enough
[22:57] <oraqol> ok cool, thanks guys, much more responsive than the #ubuntu chan
[22:57] <oraqol> ttyl!
[22:58] <hallyn> all right one more attempt at building grub2 in my vm.  then i give
[22:58] <hallyn> up
[22:58] <kirkland> hallyn: oh
[22:58] <kirkland> hallyn: actually, we need one more thing to make this vgabios work
[22:59] <hallyn> ?
[22:59] <kirkland> $ dpkg -S /usr/share/qemu/vgabios.bin
[22:59] <kirkland> qemu-common: /usr/share/qemu/vgabios.bin
[22:59] <kirkland> hallyn: qemu-common puts symlinks in that particular path
[22:59] <kirkland> hallyn: so we'll have dpkg conflicts
[22:59] <hallyn> hold on - maybe not (checking)
[23:00] <hallyn> oh
[23:00] <hallyn> so why does qemu-common create that link?  if it doesn't provide the file?
[23:00] <kirkland> hallyn: so actually, i think we just need to remove vgabios/debian/links entirely
[23:00] <kirkland> hallyn: okay, here's the logic ...
[23:00] <hallyn> well then qemu-common needs to add a bunch more
[23:00] <hallyn> (vgabios.qxl.bin, etc)
[23:00] <kirkland> hallyn: vgabios puts its bios files in a path that it "owns", /usr/share/vgabios/*
[23:01] <kirkland> hallyn: qemu-common depends on vgabios to be functional
[23:01] <kirkland> hallyn: and qemu, itself, expects those vgabios files to be in /usr/share/qemu/*
[23:01] <kirkland> hallyn: so it should be up to qemu to put symlinks in its owned path to the data it needs
[23:02] <kirkland> hallyn: and if qemu needs to add more links, then so be it
[23:02] <kirkland> hallyn: but it should be up to qemu package to do that
[23:02] <kirkland> hallyn: makes sense?
[23:02] <hallyn> sure
[23:02] <hallyn> it does make sense
[23:03] <hallyn> so scratch that merge request :)
[23:03] <kirkland> hallyn: okay, i'm going to fix that in the branch and upload
[23:03] <hallyn> which branch?
[23:03] <hallyn> both should be done at once, probably, so if you wanted to wait until monday...
[23:04] <hallyn> i.e. you can't remove the links from vgabios without fixing qemu-common, without biting some people, right?
[23:05] <kirkland> hallyn: hmm, no, i don't think that's a problem
[23:05] <kirkland> hallyn: what extra links are needed in qemu-common?
[23:06] <kirkland> hallyn: vgabios uploaded
[23:06] <kirkland> hallyn: i'm looking at qemu now
[23:08] <hallyn> Well, vgabios has all those oterh bios names now...
[23:09] <kirkland> hallyn: ah, yes, i see
[23:09] <kirkland> hallyn: i'll get those fixed now
[23:10] <hallyn> kirkland: thanks!
[23:15] <kirkland> hallyn: gotta run now, but i'll get it uploaded tonight
[23:15] <kirkland> hallyn: bzr is really slow over public wifi :-(
[23:17] <hallyn> kirkland: s/over.$//
[23:18] <hallyn> oh no did i say that?
[23:18] <hallyn> kirkland: thanks, ttyl
[23:19] <hallyn> kirkland: btw, it helps to preseed (if you will) your laptop...  I have one shared bzr repo which makes subsequent re-fetches of a tree pretty fast no matter what