[01:21] <uvirtbot`> New bug: #816754 in nova (universe) "add ovs-vsctl to sudoers file" [Undecided,New] https://launchpad.net/bugs/816754
[01:31] <uvirtbot`> New bug: #816758 in php5 (main) "Cant type some characters in readline()" [Undecided,New] https://launchpad.net/bugs/816758
[04:08] <DanaG> Say, what's the difference between the generic kernel and the server kernel?
[04:08] <DanaG> And how do I get my server to automatically enable the tweaks powertop usually suggests?
[04:11] <DanaG> hmm, seems like laptop_mode may do part of it.
[04:12] <DanaG> Now if only I could tie laptop_mode ac/battery into apcupsd...
[04:12] <twb> DanaG: if you install both, you can just diff /boot/config-*
[04:13] <twb> apcupsd (or nut?) probably has event hooks similar to acpid
[04:14] <DanaG> The tweaks laptop-mode does are the runtime-PM and the ALPM and all that.
[04:15] <twb> laptop-mode does a whole bunch of shit
[04:15] <twb> like spinning down idle disks
[04:16] <DanaG> Yeah, I want it to do write-clumping on AC, but drop back to failsafe on battery.
[04:16] <DanaG> Oh, I can have apcupsd call laptop-mode
[04:19] <DanaG> I'm trying to figure out why my microserver is taking 28% of the capacity of my APC ES 550.
[04:20] <DanaG> Newegg states the capacity is 330 watts; 28% of that is 92 watts.
[04:21] <DanaG> Oh wait.... I have my laptop on it, too.  Silly me.
[04:24] <twb> DanaG: and the LCD monitor too, probably
[04:24] <twb> And the switch
[04:25] <twb> Not much point having all your servers on the UPS if the switch isn't -- the server runnign nut won't be able to tell the other servers to shut down
[04:28] <DanaG> I only have the one microserver, anyway.
[04:28] <DanaG> I run it headless, or use the HP IPMI card.
[04:28] <DanaG> hmm, it's still saying 28% load.
[04:28] <DanaG> I'll bet that's bogus.
[04:29] <DanaG> That's after taking my laptop out from behind it.
[04:31]  * DanaG wishes AMD would make an ES1000 IPMI chip.
[04:32] <DanaG> That way you could have native-res framebuffer on IPMI.
[04:32] <twb> Fuck that shit
[04:32] <twb> Give me a ttyS0 any day
[04:32] <DanaG> Yeah, that's even better.
[04:33] <twb> Good luck getting it these days, tho :-/
[04:33] <DanaG> Do any of you guys have contacts at HP, who would take firmware issues seriously?
[04:33] <twb> It's all "just browse to this totally broken web management UI" crap
[04:33] <DanaG> The Microserver has a couple of issues: ACPI declares wrong KCS base address, and their IPMI firmware doesn't give the serial-over-LAN support the IPMI chip itself supports.
[04:34] <DanaG> Oh, and no option to have external graphics as default while still leaving internal GPU enabled.
[04:42] <lifeless> SpamapS: your lxc branch needs an update ;)
[04:44] <DanaG> So anyway, if any of you guys have real liaison with HP, please try to get them to fix those things!
[04:44] <DanaG> oh, and I had to install a modded BIOS to be able to use AHCI (and hotplug mode) on all ports.
[07:56] <ttx> Daviey: Crowbar open sourced at https://github.com/dellcloudedge/crowbar
[08:01] <twb> What is it?
[08:13] <ttx> twb: Dell's solution to deploy OpenStack on bare metal, somehow competition to Ubuntu's Orchestra
[08:14] <ttx> Dell promised to open source it but it took months
[08:21] <twb> Out of curiosity, how well do (either) deal with heterogeneous hardware?
[08:22] <ttx> twb: no clue. I suspect Orchestra is pretty agnostic, and Crowbar has a reliance on some Dell stuff
[08:23] <SpamapS> lifeless: my LXC branch is effectively dead unless niemeyer decides to reverse his position 180 degrees. :-(
[08:24] <twb> SpamapS: what branch is this?  I'm seriously invested in LXC on lucid.
[08:24] <SpamapS> twb: its for Ensemble
[08:26] <SpamapS> wow.. ifupdown has 106 normal open bugs
[08:26] <SpamapS> wtf!
[08:27]  * SpamapS wonders why nobody has taken it upon themselves to rid us of this "literate programming" monstronsity. :-P
[08:27] <SpamapS> twb: are you switching to the natty backport kernel then?
[08:28] <lifeless> SpamapS: surely something is better than nothing ?
[08:29] <SpamapS> lifeless: its apparently just a trivial thing for somebody to do it the "right" way.
[08:29] <lifeless> bitter much ?
[08:29] <lifeless> SpamapS: is there anything I can do ?
[08:30] <twb> SpamapS: no, currently I am just not touching the kernel at all
[08:30] <twb> SpamapS: I will probably have to switch to the backport kernel at some point, or just reroll the normal lucid one to remove that clusterfuck vsftpd "fix"
[08:31] <SpamapS> lifeless: hazmat and flacoste and myself have all presented compelling cases for having both. The "I don't want to maintain two ways of doing it" argument seems to have set and cured.
[08:31] <twb> SpamapS: re ifupdown, there's "ipcfg"
[08:32] <lifeless> that presumes one is  asuperset of the other
[08:32] <lifeless> how will one bring up lxc in dev mode ?
[08:32] <SpamapS> using the "local" provider
[08:32] <lifeless> does that touch your host OS ?
[08:33] <SpamapS> which says you have one machine, that can provide infinite LXC containers... whereas ec2 will be able to provide only one "null" containier
[08:34] <SpamapS> lifeless: its 6 and 1/2 dozen IMO.. but apparently we have to lay the foundation for the future work now.
[08:34] <SpamapS> (a trap I've seen nearly every purist fall into)
[08:34] <twb> There's also "netscript", but I think that's old and abandoned
[08:34] <lifeless> SpamapS: so how much work is it to update the branch ?
[08:35] <SpamapS> twb: I like /etc/network/interfaces, and all that it provides.. but god save me if I ever have to patch ifupdown.
[08:35] <SpamapS> lifeless: couple hours probably.
[08:35] <lifeless> SpamapS: I suspect it will take me longer...
[08:35] <lifeless> SpamapS: I can't wait for the real-deal, whatever it is, to get experimenting with lxc
[08:36]  * SpamapS sees those big blue (blue?) eyes givign him a watery anime stare...
[08:36] <lifeless> SpamapS: so either I need to can the experiment; do it myself, or hopefully some kind soul familiar with the code will do it ;)
[08:36] <lifeless> SpamapS: btw, lxc - and 'public' ips is easy
[08:36] <SpamapS> lifeless: let me see if I can find a gnome on my laptop to do that..
[08:37] <lifeless> set the interface to br0, and setup a br0 with eth0 in it
[08:37] <SpamapS> lifeless: yeah, public IP is simple... assign elastic IP.. nat.. done.
[08:37] <lifeless> of course, this fails horribly on non-AP mode wifi, but thats broken-by-design anyhow.
[08:37] <lifeless> SpamapS: no nat, no elastic IP :)
[08:37] <twb> SpamapS: also the lack of iproute2
[08:37] <SpamapS> oh you mean for local dev.. for local dev .. virbr0 should be fine
[08:38] <lifeless> SpamapS: thats local machine only; I mean for LAN accessibility to the brought up services
[08:39] <SpamapS> lifeless: yeah, for that its the exact same problem space that libvirt lives in
[08:40] <lifeless> yah
[08:40] <SpamapS> lifeless: yeah, public IP is simple... assign elastic IP.. nat.. done.Warning: criss-cross merge encountered.  See bzr help criss-cross.
[08:40] <SpamapS> oops
[08:40] <SpamapS> lifeless: Warning: criss-cross merge encountered.  See bzr help criss-cross.
[08:40] <SpamapS> lifeless: 8 conflicts.. not the end of the world
[08:40] <lifeless> I saw a ref to vnet as a bridge interface, but in my libvirt vnet isn't a bridge of its own
[08:40] <lifeless> but an interface connected to virbr0
[08:41] <SpamapS> lifeless: my lxc branch plays dumb and just feeds in these three lines as the lxc config
[08:41] <SpamapS> lxc.network.type=veth
[08:41] <SpamapS> lxc.network.link=virbr0
[08:41] <SpamapS> lxc.network.flags=up
[08:41] <SpamapS> of course, that should be configurable
[08:42] <lifeless> yeah, but thats a different discussion :>
[08:54] <SpamapS> 1 conflict(s) resolved, 5 remaining
[08:54] <SpamapS> die conflicts die
[08:59] <SpamapS> ok.. trunk merged.. but.. does it work? :-P
[09:03] <twb> FFS, my patch panel's wonky or something; wiggling the RJ11 socket fixed the 40% packet loss / 800ms latency
[09:08] <SpamapS> ugh.. they've gone and abstracted everything to kingdom come
[09:12] <twb> I hate that
[09:15] <SpamapS> "We might need another value for booleans.."
[09:15] <SpamapS> ;)
[09:15] <twb> Naming everything that might ever need to vary
[09:20] <alamar> SpamapS: this is why they made ternary ram ;)
[09:22]  * SpamapS now understands why they had to rip out his "don't start a machine if you don't have to code" .. because they've painted themselves into a corner w/ assumptions. tsk tsk.
[09:22]  * SpamapS will play along
[09:38] <lynxman> Daviey: ping
[09:38] <Daviey> lynxman: o/
[09:39] <lynxman> :)
[10:10] <_johnny> anyone know which system is used for www git view on, say, git.videolan.org?
[10:11] <_johnny> ah, gitweb
[10:32] <SpamapS> lifeless: progress!
[10:32] <SpamapS> 2011-07-27 03:32:09,967 INFO Bootstrapping environment 'lxc' (type: lxc)...
[10:32] <SpamapS> 2011-07-27 03:32:09,970 DEBUG Launching Ensemble bootstrap instance.
[10:39] <lifeless> woo
[10:39] <lifeless> SpamapS: ship it!
[10:45] <SpamapS> hrm.. lxc may be broken on my machine
[10:46] <SpamapS> hmm.. maybe cgroup-bin isn't working..
[10:46]  * SpamapS puts the mount back
[10:52] <SpamapS> lifeless: ok, lp:~clint-fewbar/ensemble/lxc-container seems to at least try to start an LXC container properly. Unfortunately, my LXC is broke so will have to monkey with that tomorrow.
[10:52] <SpamapS> lifeless: I think I've finally reached the end of my insomnia.. ttyl
[11:51] <borat> hi
[11:51] <borat> i have an apparmor related question
[11:52] <borat> since apparmor is path based, is there a way to ensure that the file the path points to is actually the file it's supposed to be?
[11:52] <jdstrand> borat: I'm assuming you are referring to symlinks?
[11:52] <pmatulis> http://manpages.ubuntu.com/ seems braindead today
[11:52] <borat> eg ensure that /usr/bin/firefox is actually the executable of firefox and not a link to /tmp/funny-little-script
[11:53] <borat> jdstrand: yes, mainly
[11:54] <jdstrand> borat: apparmor resolves symlinks, so that is not a problem. hard links are treated like any other path, so the hard linked path would need a separate access rule to allow access
[11:55] <jdstrand> (or a separate profile, if you are referring to attachment)
[11:55] <jdstrand> (or alias)
[11:57] <borat> jdstrand: what happens if I just replace the executable? will apparmor detect that or could i place a checksum in the profile or something?
[12:02] <uvirtbot`> New bug: #816934 in samba (main) "package samba-common-bin 2:3.5.8~dfsg-1ubuntu2 failed to install/upgrade: subprocess dpkg-deb --fsys-tarfile returned error exit status 2" [Undecided,New] https://launchpad.net/bugs/816934
[12:07] <jdstrand> borat: if the executable is in the same path, then access is allowed. that said, you would not typically give a confined application write access to binaries it is allow to execute (or itself)
[12:07] <jdstrand> borat: well, 'access is allowed'-- I should say 'access is the same as the profile previously allowed'
[12:09] <zul> ttx: i had a look at crowbar yesterday *giggle*
[12:10] <borat> jdstrand: okay, thanks for the enlightenment :)
[12:16] <skeuds> Hello everyone,
[12:17] <skeuds> I am a beginner on Ubuntu server, i have a dedicated server with ubuntu 10.04 installed and today when i try to login on my server with ssh, i get this error message and the connection is closed :
[12:18] <skeuds> "/bin/bash: Exec format error"
[12:18] <skeuds> If someone have an idea it will be very cool
[12:19] <twb> skeuds: what's the exit status?
[12:19] <skeuds> I have this message :
[12:20] <skeuds> Last login: Wed Jul 27 13:55:18 2011 from xxxxxxxxxxxxxxx.fbx.proxad.net
[12:20] <skeuds> Connection to xx.xxx.xxx.xxx closed.
[12:20] <skeuds> oups
[12:20] <skeuds> Last login: Wed Jul 27 13:55:18 2011 from xxxxxxxxxxxxxxx.fbx.proxad.net
[12:20] <skeuds> "/bin/bash: Exec format error"
[12:20] <skeuds> Connection to xx.xxx.xxx.xxx closed.
[12:20] <twb> That is not the exit status
[12:20] <skeuds> How can i get the exit status ?
[12:21] <twb> echo $?
[12:22] <skeuds> ok i get it : 1
[12:22] <twb> Not helpful.  Try "ssh foo -t pwd", where "foo" is the destination.
[12:22] <skeuds> Ok i try
[12:23] <skeuds> Same error :
[12:24] <skeuds> "/bin/bash: Exec format error"
[12:24] <skeuds> Connection to xx.xxx.xxx.xxx closed.
[12:24] <skeuds> Exit status 1
[12:24] <twb> OK, I don't know, that system is totally fucked.
[12:24] <twb> It sounds like you have managed to boot an x86 kernel with an x86-64 userspace or something
[12:25] <twb> Have you been messing with your kernel?
[12:25] <skeuds> I have juste installed a clean copy of ubuntu 10.04. No manual modification
[12:25] <Pici> When did you install it?
[12:26] <skeuds> 1 year ago
[12:26] <Japje> ssh user@xxxxxxxxxxxxxxx.fbx.proxad.net -c "/bin/sh uptime"
[12:26] <Japje> see is sh is broken
[12:26] <Japje> or /bin/dash
[12:26] <twb> Uh, it's not -c
[12:27] <Pici> skeuds: You just installed Ubuntu 10.04, or you installed it a year ago?
[12:27] <skeuds> I installed it 1 year ago
[12:27] <twb> Japje: "/bin/sh uptime" is not a cipher spec
[12:27] <Japje> twb: the -c flag was command right?
[12:28] <Pici> I wonder if the server was comprimised
[12:28] <twb> No
[12:28] <Japje> damn
[12:28] <Japje> my memory needs an upgrade
[12:28] <Japje> didnt need the -c :P
[12:29] <skeuds> With : ssh mailto:user@xxxxxxxxxxxxxxx.fbx.proxad.net "/bin/sh uptime"
[12:29] <twb> Sadly sshd still uses system(3) instead of execvp :-/
[12:29] <skeuds> i have the same error
[12:29] <skeuds> "/bin/bash: Exec format error"
[12:29] <twb> skeuds: that's because uptime is not a program
[12:30] <twb> Er, not a sh script in the working dir
[12:30] <twb> skeuds: you should just reinstall the thing
[12:30] <ewook> so how about you point out the full path to uptime
[12:30] <twb> ewook: it'll still use system
[12:30] <twb> You *can't* opt out of system()
[12:31] <skeuds> So i must reinstall all the system ?
[12:31] <twb> skeuds: unless you can work out why bash is pissed off
[12:32] <skeuds> Ok thank you for your help
[12:32] <skeuds> :)
[12:32] <_ruben> no physical/kvm/whatever access to the box?
[12:32] <ewook> errr
[12:33] <ewook> just go ssh whatever@something "/usr/bin/uptime"
[12:33] <ewook> if your user has a default shell, that is
[12:34] <_ruben> but if that default shell happens to be broken....
[12:34] <ewook> true, then you do have an issue.
[12:34] <skeuds> I have a rescue system
[12:34] <skeuds> boot on a network operating system
[12:35] <skeuds> To make maintenance operation
[12:35] <skeuds> perhaps i will try it before reinstall
[12:36] <ewook> but, ssh whatever@something "/bin/sh" should get you another (in a crazy way) shell.
[12:37] <skeuds> Nop ewook, i have always "/bin/bash: Exec format error" error message
[12:37] <ewook> skeuds: well, don't do bash then
[12:37] <_ruben> if bash is the default shell, any other shell would be spawned by bash, and if bash is broken...
[12:38] <ewook> doh..
[12:38] <ewook> didn't think of that ofcourse.. but scp should still work? so replacing whatever "default" you have should be doable, right?
[12:38] <_ruben> if you allow root logins to overwrite /etc/paswd ... :)
[12:39] <_ruben> +s
[12:39] <ewook> lul
[12:39] <twb> This is why sashroot
[12:39] <ewook> try to think in 30 degrees C ... naah..
[12:40] <ewook> physical access?
[12:40] <skeuds> No physical access
[12:40] <ewook> *_*
[12:40] <ewook> bleh
[12:41] <ewook> glhf, since I'm just running around in circles due to the heat, I'll do no good here.
[12:42] <skeuds> ^^
[12:42] <uvirtbot`> skeuds: Error: "^" is not a valid command.
[13:33] <jamespage> zul: I need an opinion on how todo something - around?
[13:34] <zul> almost always
[13:34] <zul> but if its a bad opinon i will deny that i said it
[13:34] <jamespage> zul: great
[13:35] <jamespage> zul: so this is with respect to jenkins plugins - some of which I aim to package
[13:36] <jamespage> when a plugin is installed it requires a restart of jenkins to get it noticed
[13:36] <zul> right
[13:36] <jamespage> however it would be possible to install a number of plugins all in one go; and I don't really want to restart jenkins after each  - I want todo it after they are all installed
[13:37] <jamespage> is this posible? or should I leave it up to the admin to schedule the restart?
[13:37] <jamespage> I've left it as the second option ATM - mainly because it means you can install the plugin(s) and then enable them at a later date when the jenkins instance is not busy
[13:38] <zul> jamespage: well you can do a trigger a restart of jenkins when you do a package install but that might be a bit hairy
[13:38] <zul> right with my admin hat on i would do the second part
[13:38] <jamespage> that was my thinking as well - did not want to get to clever with the package
[13:39] <jamespage> I'll add some notes into the README.Debian to state that a restart of jenkins is required
[13:39] <jamespage> that should keep folks happy.
[13:40] <zul> cool beans
[14:13] <raubvogel> In https://help.ubuntu.com/10.04/serverguide/C/openldap-server.html, under Consumer Configuration,  olcRootPW should match the same one in the master/provider ldap server, right?
[14:48] <mendel_> guys how do I reinstall a pci card
[14:48] <mendel_> I've added some drivers to the blacklist and want to reinstall it
[15:02] <uvirtbot`> New bug: #817034 in php5 (main) "PHP SOAP Extension Not Working" [Undecided,New] https://launchpad.net/bugs/817034
[15:04] <dori922> is UEC 11.04 compatible with unupdated UEC10.10 nodes?
[15:09] <jamey-uk> I've just installed 11.04 server on my new machine, upgraded all the packages and then noticed the kernel images were being held back. So I did a dist-upgrade, which sorted it out but it reinstalled GRUB… seemingly incorrectly as I now get the GRUB rescue screen on boot. I've tried using the Rescue mode of the CD but it seems that my UEFI machine has meant the disk was formatted with GPT which it doesn't seem to support. Wha
[15:47] <RoyK> jamespage: I thought ubuntu/grub was supposed to support GPT
[15:47] <RoyK> jamespage: big disk?
[16:01] <uvirtbot`> New bug: #817074 in etckeeper (main) "Please merge etckeeper 0.56 (main) from Debian unstable (main)" [Wishlist,Confirmed] https://launchpad.net/bugs/817074
[17:20] <Daviey> uhoh
[17:35] <hallyn> kirkland: getting ready to push qemu-kvm0.14.1 merge.  Tests fine here.  (I almost, once AGAIN, pushed with the roms!)  It's a pretty trivial merge, but if you want to take a look first gimme a shout
[17:36] <kirkland> hallyn: dude, I trust you -- if you want *me* to look at it, though, I'd be glad to
[17:37] <kirkland> hallyn: as for the roms problem, add a check in debian/rules
[17:37] <hallyn> good idea, think i'll do that before pushing.  thx
[17:37] <kirkland> hallyn: or, rather, add rules in the clean target to prune them *every* time
[17:37] <hallyn> btw, there is a tests/pi_10.com program in there (binary)
[17:37] <kirkland> hallyn: actually, a check is probably good, since it's the tarball
[17:37] <kirkland> hallyn: wtf is that?
[17:37] <hallyn> dunno. windows executable?
[17:37] <hallyn> it's tiny
[17:38] <hallyn> lemme see where it's used
[17:39] <hallyn> hm, i'd previouslyi removed it (for 0.14.0), but it had shown back up somehow
[17:39] <hallyn> it's used by tests/qruncom.c
[17:40] <kirkland> hallyn: hmmf;  remove it
[17:40] <kirkland> hallyn: and poke aligouri and figure out how to build it
[17:40] <kirkland> hallyn: i'm always surprise how/why he ships all these binaries with the upstream source
[17:42] <hallyn> kirkland: dont' blame him    http://repo.or.cz/w/qemu.git/history/HEAD:/tests/pi_10.com
[17:42] <hallyn> added in 2004  by bellard, never changed since :)
[17:43] <kirkland> heh
[17:48] <robbiew> Daviey: hey...did we ever do the flip with ntpd and ntpdate?
[17:48] <Daviey> robbiew: no, not happend yet
[17:48] <Daviey> still TODO
[17:49] <robbiew> cool deal
[17:52] <Daviey> robbiew: i was concerned ntpdate (via ubuntu-minimal) and ntp would conflict, but it seems both can be installed safely.
[17:54] <zul> SpamapS: can you ack landscape-client in the SRU queueu
[17:54] <SpamapS> zul: i didn't see it there
[17:54] <zul> but i uploaded it
[17:54] <SpamapS> oh, only lucid
[17:54] <SpamapS> I was looking in natty :p
[17:55] <zul> oh...
[17:56] <Daviey> robbiew: so, seems ntp binds to 0.0.0.0 .. we need to make the default only bind to localhost to instal it by default.
[17:56] <robbiew> RoAkSoAx: ever been to Boston?
[17:56] <RoAkSoAx> robbiew: nope but wanted to go there for quite a while :)
[17:57] <robbiew> RoAkSoAx: well...the week of August 29th, you'll get your wish...talk to kirkland ;)
[17:58] <kirkland> robbiew: ;-)
[17:58] <kirkland> RoAkSoAx: we'll take the plans private
[17:58] <kirkland> robbiew: thanks!
[17:58] <RoAkSoAx> robbiew: hehe cool, will do ;)
[17:59] <zul> RoAkSoAx: its not actually boston btw ;)
[18:00] <RoAkSoAx> zul: lexington ? :)?
[18:00] <zul> yep
[18:00] <Daviey> RoAkSoAx: make sure you take some tea with you, there is a shortage there.
[18:00] <RoAkSoAx> zul: close enough, just 30 mins :)
[18:00] <patdk-wk> stop dumping it into the bay
[18:01] <RoAkSoAx> Daviey: hehe will do.. I;m a tea drinker
[18:02] <Daviey> good stuff!
[18:06] <aleperalta> Hello all, I'm using hardy in a server and I've just installed the postgresql-8.4 backports, and it's running fine, but now I need postgresql-8.4-postgis but that isn't supplied by the backports. I've also checked out the ppa UbuntuGis but postgresql-8.4-postgis isn't supplied by the ppa. What choices do I have besides compiling?
[18:07] <Daviey> aleperalta: consider backporting it yourself?  Others might appreciate it :)
[18:07] <patdk-wk> aleperalta, beg, plead, file a backport request
[18:08] <aleperalta> Daviey: nope, I can try... any how-tos on that?
[18:09] <aleperalta> patdk-wk: hehe :-D I don't think they'll listen to prayers.. :-)
[18:09] <SpamapS> zul: done
[18:10] <zul> SpamapS: thanks
[18:44] <dkn> anyone know how to make the /var/log/auth.log longer? i can only see a couple days back, when i want to see a couple months back
[18:45] <dkn> duh... got it.... auth.log.1
[18:46] <ahasenack> hi, does anybody know why a kvm natty guest would have the directory /sys/bus/xen ?
[18:46] <ahasenack> and not always, I don't know yet what happened, we are still investigating it
[18:47] <ahasenack> as we used that directory to report the vm as being of type xen, but in fact it's kvm
[19:14] <Slyboots> I've got an absoluty bewildering problem with Ubuntu-server and using iSCSI
[19:15] <Slyboots> If I try to access the same iscsi lun from two locations at once.. the server resets the network connection.  But instead of going back to its static IP it pulls an address from the DHCP pool
[19:15] <Slyboots> ... *why*
[19:20] <Slyboots> This is seriously pulling a kibosh on its usage as a server, I've already set it as a static interface in /etc/networks/interfaces
[19:20] <Slyboots> So I dont get it
[19:23] <hallyn> kirkland: do you think a check for binaries should be done from the clean: or the build: rule?
[19:23] <kirkland> hallyn: hmm, clean, probably
[19:23] <hallyn> eh, the clean rule already forcibly removes them
[19:23] <Slyboots> Might it be worthwhile to strip out the dhclient totally?
[19:23] <Slyboots> But I would like to *solve* the problem, and sto pit crashing when I access iSCSI at all x.x
[19:23] <hallyn> I'm going to stop it removng them
[19:28]  * Slyboots headscratches
[19:29] <Slyboots> Im really strugglign to also find out why iscsi is causing the network card to crash
[19:29] <Slyboots> I've been throught eh logs but.. Im not seeing anything usful
[19:33] <Slyboots> And I cant bloody remove dhclient "Virtual packages like dhclient can not be removed"
[19:33] <Slyboots> wtf
[20:42] <raubvogel> I am installing slapd and it is not asking me for the admin password I want it to use. How can I enter it or force it to be asked? FYI, I can completely wipe the ldap thing if needs to be (and I know which directories need to go).
[20:51] <oCean> !google | oCean
[20:51] <g0t> Results for | oCean on Google:
[20:51] <g0t> --
[20:51] <oCean> g0t: you should disable that script
[21:09] <jcastro> kirkland: where do you keep all those nifty preseed config files? the ones for automating installs
[21:09] <kirkland> jcastro: i mostly use http://bit.ly/uquick
[21:09] <kirkland> jcastro: it's under bzr control in lp:bikeshed and lp:orchestra
[21:10] <kirkland> jcastro: you can do a quick/dirty/unattended install by a) booting a server cd and b) adding the following to the kernel command line "local=en_US priority=critical url=http://bit.ly/uquick"
[21:13] <jcastro> ah that URL is what I was missing, thanks.
[21:18] <zroysch> hello
[21:19] <zroysch> I have an mdadm raid-0 that recently had a drive fail, or at least marked as faulty. I just want to replace it.
[21:19] <zroysch> does anyone know of a resource that will specify what exactly I need to match on the existing drive to make it work properly?
[21:19] <zroysch> rpm? cas? cache?
[21:21] <zroysch> sorry I meant raid1. it is a mirror
[22:01] <qman__> zroysch, mdadm doesn't care about any of that
[22:02] <qman__> just create a raid partition at least as large as your existing one, and add it to the array
[22:02] <qman__> performance is determined by the least common denominator
[22:02] <qman__> so you want a drive equally as large or larger than the one it's replacing, and equally fast or faster
[22:31] <zroysch> qman__: ok, thank you
[23:50] <LeChacal> hi, question my isp has port 80 blocked for incoming connections (like for a web server), I know that I can change the port that apache looks for but then to get to the web site you have to type example.com:88 if the port was changed to 88. Is there something I can do to get around my ISP blocking 80 but not have to include the port number at the end of the url? googling isn't coming up with much thank you
[23:51] <qman__> you can't get around having the port number on your URL
[23:51] <qman__> what you can do, though, is sign up for something like dyndns, and create a redirect
[23:51] <qman__> and have your.website.com redirect the user to 1.2.3.4:88
[23:51] <qman__> or other.website.com:88
[23:51] <LeChacal> i didn't think that DNS held port data
[23:52] <qman__> it doesn't
[23:52] <qman__> but dyndns.com and some other sites offer that service
[23:52] <qman__> for free
[23:52] <qman__> so, your site is still :88, but you have a URL that will redirect users without the port number
[23:53] <LeChacal> hmm didn't know that, thank you, i am reading the dyndns site now