[00:39] <mwhudson> if you've just deployed an openstack for testing purposes, is there some easy way to get some images into glance?
[01:02] <adam_g> mwhudson, grab the *-amd64.img or *-i386.img from http://cloud-images.ubuntu.com/trusty/current/
[01:03] <adam_g> mwhudson, install python-glanceclient, source your OS credentials and run: glance image-create --name trusty --disk-format ami --container-format ami --is-public True --file $path_to_img
[01:03] <mwhudson> adam_g: is there some cli tool for doing that?
[01:03] <mwhudson> ah ok
[01:05] <adam_g> mwhudson, https://launchpad.net/simplestreams is a project that provides the ability to pull ubuntu images and sync to your glance server automatically
[01:07] <mwhudson> oh right
[01:07] <mwhudson> i think i even knew that, somewhere...
[01:42] <semiosis> sarnold: ping?
[03:09] <sarnold> hey semiosis
[03:39] <sarnold> semiosis: ah, very nice bugzilla bug. Thanks :)
[03:53] <arrith> is there a standard way to have different 'views' of a directory, basically excluding certain files and directories? this is to provide different virtual machines guests different files, some of them shared
[03:53] <sarnold> arrith: not easily; check out the 'shared subtrees' support in the linux kernel
[03:53] <sarnold> arrith: Documentation/filesystems/sharedsubtree.txt
[03:54] <arrith> sarnold: ah neat thanks
[03:54] <arrith> iirc i read about a feature like this somewhere, and i think it was something about some plan 9 9p file sharing protocol thing
[03:55] <sarnold> arrith: very similar indeed; plan9 apparently used 'bind mounts' all over the place; you'll do something very similar here, mount --bind but with some 'unshare' stuff beforehand..
[03:56] <sarnold> arrith: oh, hrm, you mean 9p specifically? maybe. I don't know enough about it. :(
[03:56] <arrith> yeah unshare sounds great. i mean i figure there's probably a way to exclude with nfs but this is all on one machine, so that isn't as lightweight as i would prefer
[04:03] <hallyn> in fact, the unshare syscall came about precisely for pam modules to be able to use CLONE_NEWNS to provide polyinstantiated directories for mls.
[04:04] <hallyn> before that you had to clone.  which was useless in pam.
[04:05] <hallyn> (polyinstantiated dirs being sort of what you want)
[04:07] <arrith> ah that's a good search keyword
[04:08] <arrith> i'm finding a lot about namespaces
[04:24] <arrith> seems the most straightforward is maintaining links, bind mounts, or rofs-filtered
[04:24] <arrith> http://askubuntu.com/questions/44925/how-can-i-create-a-filesystem-view-of-a-folder-that-excludes-certain-files
[06:31] <lerra> Hi, I have a netinstall setup based on preseed and 12.04.4, a new laptop I have does not have it's nic present in debian installer. So i tested a 12.04.4 livecd and the nic comes up there so the kernel supports it. When I create my own alt 12.04.4 cd with the preseed on it it gives the same error that it cant find the nic but when I use the normal preseed it seams like the nic works but in a later stage of the instal
[07:08] <rbasak> tgm4883: are you aware that Apache has moved its default directory to /var/www/html? Some changes in config files are accordingly needed.
[07:24] <lordievader> Good morning
[07:32] <joren> Hey, anyone happen to know how to get a dell r610 to pass the disks off to ubuntu for software raid? I've traid everything I can think of at this point...
[07:34] <sheptard> what raid card
[07:37] <joren> PERC 6/i? I thought I saw something saying it was part of the iDRAC6 stuff, I think it's just a random onboard card
[07:37] <joren> but I'll try and dig up more details.
[07:37] <joren> RAID CONTROLLER	 PERC 6i w/ 256MB BATTERY BACKED CACHE
[07:38] <joren> actually, I guess if it's got the battery backup maybe I want to use the perc
[08:14] <rabbel> Hey guys. Did anyone ever tried to run an ssh tunnel (reverse) into the background?
[08:15] <rabbel> If I try it, my tunnel keeps disconnecting immediately
[08:15] <rabbel> If I run it in the foreground, it just stays there and does not disconnect...
[08:17] <joren> rabbel, can you just stick it in a screen?
[08:18] <rabbel> joren: is that the only solution?
[08:19] <joren> probably not, but it's not something I've tried.
[10:28] <tiblock> Hi. I have debian's init script, will it work in ubuntu?
[10:29] <lordievader> tiblock: Yes. At least the scripts I wrote do :)
[10:29] <tiblock> lordievader, thank you
[10:43] <RoyK> rabbel: I run them in the background
[11:52] <hadifarnoud> I see some services were not affected by heartbleed bug. is it just because of OpenSSL version, or they use an alternative (if there is any)?
[11:53] <bekks> hadifarnoud: All libssl version unaffected and services using them arent affected.
[11:54] <hadifarnoud> bekks: so libssl is the alternative
[11:55] <bekks> hadifarnoud: No. libssl is part of openssl.
[11:55] <bekks> hadifarnoud: And the libssl component is/was affected.
[11:55] <bekks> !sslbug | hadifarnoud
[11:57] <Patrickdk> hadifarnoud, curl wasn't affected
[11:57] <Patrickdk> hadifarnoud, and the *issue* is not limited to services
[11:58] <hadifarnoud> Patrickdk: did not know that.
[11:58] <Patrickdk> wget is affected
[11:58] <Patrickdk> php is affected
[11:58] <Patrickdk> any client or server that uses ssl/tls, is affected
[11:58] <hadifarnoud> Patrickdk: all VPNs. shit, that's huge
[11:58] <Patrickdk> if your browser uses openssl, and connects to a server, that server could compromise you
[11:59] <Patrickdk> not all vpn's
[11:59] <Patrickdk> it's limited to ssl/tls vpn's
[11:59] <mardraum> not even most vpns, heh
[11:59] <Patrickdk> ipsec is not compromised, pptp is not
[11:59] <Patrickdk> ppp I mean
[11:59] <hadifarnoud> Patrickdk: openvpn is I guess
[11:59] <Patrickdk> yes
[12:01]  * Patrickdk doesn't know any browsers that use openssl though
[12:05] <hadifarnoud> Patrickdk: IE probably :D
[12:06] <hadifarnoud> I did upgrade all packages. version is still 1.0.1 but hopefully it has patched
[12:06] <Patrickdk> dpkg -l | grep libssl
[12:06] <Patrickdk> then compare that version with the openssl entry from
[12:06] <Patrickdk> !usn
[12:08] <hadifarnoud> ubottu: I do have affected version (1.0.1-4ubuntu5) but apt update&upgrade says nothing to update
[12:09] <hadifarnoud> nevermind. I have the patch.
[12:54] <moparisthebest> how can I add another ipv6 address to a running machine?
[12:54] <moparisthebest> I added a 'iface eth0:1 inet6 static' block in /etc/network/interfaces and ran /etc/init.d/networking restart
[12:54] <moparisthebest> but that didn't do it
[12:58] <rbasak> Don't restart networking. That can break things (eg. in Trusty).
[12:58] <rbasak> ifdown the interfaces you want to change, edit /etc/network/interfaces, then ifup the interfaces you changed.
[12:59] <rbasak> I'm not sure if this is the cause of your problem or not.
[13:00] <rbasak> (I hope it's self-evident that ifdown will disconnect you if you run it over the interface you're taking down)
[13:02] <moparisthebest> yea it's on a remote machine with only one interface, so I can't really do that :/
[13:02] <moparisthebest> this is 12.04 Precise btw, if that matters
[13:16] <zul> coreycb: glance rc2 should be out today (hopefully) do you want to handle it? ill handle swift rc2 if its out today
[13:16] <coreycb> zul, sure sounds good
[13:48] <jamespage> clre
[13:48] <jamespage> coreycb, zul: https://code.launchpad.net/~james-page/swift/fixup-upgrades/+merge/215423
[13:48] <jamespage> zul, that can go with the swift rc2 imho
[13:48] <jamespage> but lets get it pushed into the lab so we can test it
[13:50] <zul> jamespage:  done...merged it as well
[13:50] <zul> jamespage:  wait...hold on...my tarmac is broken...gimme a sec
[13:53] <coreycb> jamespage, +1 thanks!
[13:53] <jamespage> coreycb, that appears to be my achilles heal this cycle
[13:53] <jamespage> coreycb, got one wrong in neutron yesterday as well
[14:10] <tgm4883> rbasak, yea we noticed that, but wouldn't a different site still be in a different folder under /var/www (eg. /var/www/mythweb ) ?
[14:11] <rbasak> tgm4883: I can only guess here. Can you tell me what exactly is broken?
[14:12] <tgm4883> rbasak, just a sec, I just woke up and am reading backlog from last night. superm1 may have fixed it late last night
[14:16] <tgm4883> rbasak, I can't get him now, but looking at his comments and his commit it looks like it's fixed. Thanks for looking, sorry to bother you
[14:17] <rbasak> tgm4883: np
[14:20] <DevilsOwn> Hello! How are ya'll
[14:21] <rabbel> 12:42:50 < RoyK> rabbel: I run them in the background > I managed to do it, with -N and -f :-)
[14:21] <DevilsOwn> How do you guys have a GUI during install and none after install?
[14:22] <patdk-wk> no
[14:22] <patdk-wk> there is no gui
[14:22] <jibel> could anyone have a look at bug 1306575 found during an upgrade test from Precise to Trusty ?
[14:22] <jibel> the base system was Precise amd64 + all the package in main
[14:23] <DevilsOwn> How do you guys have a GUI during install and none after install
[14:23] <patdk-wk> there is no gui
[14:23] <verdeP> how do I create a named fifo pipe? like say I have some program that constantly outputs stuff to stdout and I want to pipe it to some command that will take that output and make a file in /tmp/fifopipe that I can access that stream from, so I can use it as a file name for daemons etc...but the file is just an access point to the pipe, like it doesn't grow in size
[14:23] <DevilsOwn> but I can use the mouse and see colors and logos during the boot process for the first time... what is that then?
[14:23] <Pici> verdeP: have you looked at the mkfifo manpage?
[14:23] <patdk-wk> DevilsOwn, in what version?
[14:24] <patdk-wk> I didn't see that at all even in trusty
[14:24] <verdeP> Pici: ah looking now ^^
[14:24] <DevilsOwn> Ubuntu Server 11.04 onwards
[14:24] <patdk-wk> hmm, not in my trusty iso's
[14:24] <patdk-wk> pici, you mean mknod?
[14:24] <DevilsOwn> trusty as-in which numerical version?
[14:24] <patdk-wk> 14.04
[14:25] <patdk-wk> and I know there isn't a gui in 12.04, installed that millinos of times
[14:25] <RoyK> rabbel: mhm
[14:25] <DevilsOwn> aah... but every server version from 11 through 12 that I have used has the GUI during install with a purple background and message boxes and choose what to install questions?
[14:26] <patdk-wk> that isn't a gui, that is ncurses
[14:26] <DevilsOwn> aha completely ncurses?
[14:26] <Pici> patdk-wk: I didn't, but that should work just as well. They look to have very similar arguments.
[14:26] <DevilsOwn> does ncurses let'cha use the mouse?
[14:26] <patdk-wk> no
[14:27] <rbasak> Is it ncurses, or whiptail+newt? Anyway, same concept.
[14:27] <patdk-wk> gpm would
[14:27] <rbasak> There is some mouse support in terminals. I've seen an urwid-based package use it.
[14:27] <rbasak> No idea about the current installer.
[14:27] <verdeP> oh mknod looks nice as I can buffer it a bit, which could be really useful as this will be for streaming audio
[14:27] <DevilsOwn> and is it a custom made installer?
[14:27] <rbasak> It's based on Debian's installer.
[14:28] <DevilsOwn> cool! thanks guys! tata for now!
[14:35] <verdeP> meh :( the command I pipe to it just says broken pipe /:
[14:35] <verdeP> I tried like command | mkfifo somename, then I also tried mkfifo somename then command | somename
[14:36] <verdeP> the docs are really sparse for both mkfifo and mknod btw xD
[14:36] <verdeP> and they don't explain the [Major][Minor] arguments either for mknod
[14:37] <verdeP> could I use >> to my named fifo? o.o
[14:38] <patdk-wk> major/minor aren't use for pipe
[14:38] <verdeP> odd if I go tail myfifo for e.g., it causes the program that I'm >> from to start, but as soon as I stop accessing that file, the program says broken pipe
[14:39] <verdeP> so like something has to be reading the pipe for it to use it?
[14:39] <patdk-wk> did you read the man page
[14:39] <verdeP> yes I know they are for block device
[14:39] <verdeP> and yes I did
[14:39] <patdk-wk> minor/major MUST NOT be used when using type p FIFO
[14:39] <verdeP> I know
[14:39] <verdeP> wtf
[14:39] <patdk-wk> tht isn't how you use a pipe
[14:40] <patdk-wk> | pipes to a program
[14:40] <patdk-wk> you need to use >
[14:40] <patdk-wk> mkfifo somename; command > somename
[14:41] <verdeP> how do I make the command think the pipe is always being accessed? cause I just tried with a single > and it does the same thing
[14:42] <patdk-wk> it can't
[14:42] <patdk-wk> it's a fifo
[14:42] <patdk-wk> first in, first out
[14:42] <patdk-wk> you can't put more in, tillyou take some out
[14:42] <patdk-wk> if you wanted it to think it was always being dumped, use /dev/null
[14:43] <verdeP> if I send it to /dev/null, can I still sample the output with another program at the same time?
[14:43] <zul> coreycb/jamespage: ok glance uploaded to saucy-proposed
[14:43] <verdeP> or transparently switch to another program taking the output and back to /dev/null when it doesn't need it, so the original command always keeps sending
[14:44] <jamespage> zul, saucy?
[14:44] <zul> jamespage:  yep
[14:44] <zul> jamespage: 2012.2.3
[14:45] <verdeP> I mean the program does give me the option to write out just as a file, but that file will get really huge, if I could just keep deleting part of the file...idk /: not sure the best way to handle this
[14:45] <verdeP> its all live streaming stuff
[14:46] <jamespage> zul, okies
[14:46] <verdeP> and I don't want to record more than whats needed for buffer
[14:46] <jamespage> zul, Daviey was wanting for something todo earlier so I pointed him at you previous uploads :-)
[14:47] <zul> jamespage:  cool...thanks...yeah unemployment must suck ;)
[14:48] <verdeP> hmm I still see a way to do it, I'll just add the named fifo to my ffmpeg server config, then only start the server after I use command >, thanks patdk-wk and Pici
[15:07] <moparisthebest> I added a file /etc/init/deluged.conf
[15:07] <moparisthebest> and made a symlink /etc/init.d/deluged -> /lib/init/upstart-job
[15:08] <moparisthebest> however, everytime I do anything with it, it says 'start: unrecognized service'
[15:08] <moparisthebest> do I need to do something else? never added an upstart job before
[16:17] <baggar11> Anyone have a good PPA on 12.04 for Libvirt backports?
[16:18] <bekks> baggar11: There are quite a few: https://www.google.de/search?client=ubuntu&channel=fs&q=ubuntu+precise+ppa+libvirt
[16:19] <baggar11> bekks: You have experience with any?
[16:20] <bekks> baggar11: No. I am using vbox since ages.
[16:22] <baggar11> bekks: libvirt supports vbox too :)
[16:22] <bekks> It wants to. All vital functions are unsupported or buggy.
[16:23] <bekks> Thats why I am not using it, since I dont need another layer between me and vbox when scripting vbox using VBoxManage, which works remotely, as well.
[17:14] <DevilsOwn> how do the devs create the ISO for Ubuntu Server?
[17:14] <jrwren> afaik, they don't. some automated system does it.
[17:14] <DevilsOwn> I have a heavily customized version of the server with asterisk configured, how do I convert it into an ISO?
[17:17] <DevilsOwn> I have a heavily customized version of the server with asterisk configured, how do I convert it into an ISO?
[17:24] <jrwren> http://askubuntu.com/questions/83617/can-i-build-a-ubuntu-iso-from-a-manifest
[17:24] <jrwren> DevilsOwn: ^^
[17:26] <DevilsOwn> thanks mate le'mme check that... appreciate it
[17:33] <DevilsOwn> If I customize the LiveCD will it be persistent also?
[17:49] <jrwren> anyone every use iptables match-set ?  it seems to not be matching for me.
[19:10] <jrwren> ah, src,dst is and not or.  ipset & iptables match-set works
[19:19] <jdstrand> zul: fyi, commented in the heat MIR
[19:20] <zul> jdstrand:  cool thanks
[19:23] <Havenstance> trying to install grub on a system with a RAID1 config it fails every time. says nothing but executing "grub-install /dev/sdb' failed this is a fatal error
[19:23] <Havenstance> 13.10 server
[19:24] <Havenstance> do i have to do a separate boot partition outside of the Software RAID1?
[19:39] <zul> jamespage: when you get a chance can you have a look at some heat fixes based on the MIR review https://code.launchpad.net/~zulcss/heat/heat-mir/+merge/215494
[19:48] <smoser> roaksoax, have you (or will you) uploaded maas ?
[20:47] <rostam> hi I am using 12.04 LTS and running tftp-ha on my ubuntu box. How could I enable the log messages on tftp-hpa? thx
[21:32] <GrueMaster> Has anyone actually tested the trusty daily amd64 server image on an actual server?  Partman had no clue as to the 120GB sata drive in my Intel servers (I tried a few different off-the-shelf systems).  Saucy found them just fine (even if it did install grub to the usb flash drive used for image installation).
[21:33] <GrueMaster> I'd file a bug, but since this was a failed install, I don't have any (wouldn't know what to file on anyways - always get the wrong packages).
[21:34] <GrueMaster> *any log files (ment to say).
[21:35] <sarnold> GrueMaster: I'd try filing against debian-installer, lacking any better ideas..
[21:36] <sarnold> GrueMaster: if the desktop installer also fails, that's the ubiquity installer
[21:40] <GrueMaster> I understand that.  My main issues (aside from the obvious bugs) is that my bug reports, no matter how detailed, almost always get rejected without logs.  Kind of frustrated witht he whole bug filing process (still).
[21:41] <GrueMaster> If I get a chance next week, I'll try to get a screen capture and file one.
[21:42] <sarnold> GrueMaster: completely understood
[21:43] <sarnold> GrueMaster: can you fire up another console while running the installer to run the ubuntu-bug command? it may be able to capture some logs
[21:44] <GrueMaster> I'll see if I can.  It will be a little while, as I am busy with something else atm.  (more "gotta have this now" crap).  :P
[21:44] <sarnold> GrueMaster: man it's too nice a friday afternoon for -that- :(
[21:46] <GrueMaster> I'm hoping to finish by 4, tee off at 5.
[21:48] <sarnold> wooooo
[22:59] <hallyn> smb: i rudely subscribed you to bug 1218959.  IIUC backporting those two commits to every post-3.5 kernel would fix the bug.  (and be much better than working around it with a udev script, since udev script will miss custom-made bridges)
[23:38] <Kihokki> I'm going to give ssh access to few users but how do I prevent them browsing each others homedirs? Just chmod?
[23:40] <sarnold> Kihokki: that's probably the easiest approach
[23:41] <Kihokki> sarnold: chmod 0750 fits for this purpose?
[23:42] <sarnold> Kihokki: assuming your users do not share groups, yes
[23:42] <Kihokki> sarnold: thanks, I'll try this