[00:02] <Gallomimia> how can i re-activate my ssh server? seem to have mistakenly uninstalled it. i think i reinstalled it now
[00:09] <Ursinha> Gallomimia, if openssh-server is installed, so it should work normally (I guess service activation happens when the package is installed)
[00:09] <Ursinha> hehe
[00:17] <koolhead17> waoo awesome :P
[00:17] <brainysmurf> I see a {crypt}x value for the userPassword field in an ldif file for an ldap; what does that mean?
[00:40] <tdelam> hey guys, i have a web app where I am encoding a video and uploading to a fileshare, I want to mount the drive locally, i have been using smbmount but I am getting permission denied, i can't seem to set the permissions, am i doing something wrong in the mounting?
[00:58] <alazare619> tdelam:  is your user accoutn in the smbmount user group?
[00:59] <tdelam> hmm, no
[00:59] <tdelam> i don't believe so, let me check when it comes back from reboot.
[01:06] <tdelam> alazare619: how do i add my user to that group
[01:06] <jvargas> how do I restore locale to en_US by default?
[01:06] <jvargas> installation was performed using es_CR, and I need to change it for compatibility reasons.
[01:51] <qman__> tdelam, try mount -t cifs
[01:52] <tdelam> i think i finally have it
[01:52] <tdelam> this is pretty finicky
[01:52] <tdelam> I have kids so I can deal with picky but this is daft :)
[01:52] <tdelam> thanks though qman__
[02:02] <koolhead17> neoXsys,
[02:11]  * semiosis is back.
[02:11]  * semiosis is away.  please leave a /msg
[02:11]  * semiosis is away.  please leave a /msg
[02:11]  * semiosis is away.  please leave a /msg
[02:12] <semiosis> oops!  sorry
[02:14] <twb> Don't do it again.
[02:15] <semiosis> twb: ok :)
[02:15] <EvilResistance> semiosis:  yeah, away message broadcasts arent exactly liked on ubuntu irc channels
[02:16] <semiosis> i can't imagine they're liked anywhere
[02:16] <EvilResistance> indeed
[02:17] <semiosis> too tired to be messing with irc client options
[02:17] <semiosis> sorry again
[03:54] <stoned> Hi, I'm on Ubuntu Lucid and I need to install python 2.7.x but repositories only have 2.6.x. What should I do?
[03:55] <stoned> Are there any 2.7 backports? Or do I have to build python from source?
[04:01] <ScottK> There are PPAs that have it, but as with any PPA, check who's PPA it is and decide if you trust them.
[04:03] <stoned> I need to install python 2.7 on lucid. What's the best way to do this?
[04:05] <rbasak> stoned: <ScottK> There are PPAs that have it, but as with any PPA, check who's PPA it is and decide if you trust them.
[04:05] <twb> ScottK: is PPA upload access discretionary?
[04:05] <rbasak> twb: anyone can have their own PPA. You just need to sign the Ubuntu Code of Conduct, that's all.
[04:05] <twb> ScottK: i.e. if I "own" a PPA, can I delegate upload rights to additional people, who maybe you (a user of the PPA) don't trust?
[04:06] <ScottK> You can tell who had upload rights.
[04:06] <stoned> What's a PPA?
[04:08] <stoned> I have to also setup mod python for apache for 2.7
[04:08] <stoned> Should I build from source?
[04:09] <twb> stoned: why do you need python 2.7 on lucid?
[04:09] <stoned> The server we have where everythign is hosted is lucid
[04:09] <stoned> The client needs to host something that requires 2.7 apparently
[04:09] <stoned> 2.6 is what comes w/ lucid
[04:09] <twb> Say to them that this is non-trivial.
[04:09] <stoned> I can't. I have to get this done.
[04:09] <twb> Ask them if it is worth the investment, or if they'd prefer to fix/test their app to work with 2.6
[04:10] <twb> I have had this argument with customers before
[04:10] <twb> For something as deep in the core as python, it will be a headache to backport it to lucid.  Best is to fix their app, middle is to upgrade the VPS to 12.04; worst is to try to backport newer python to lucid
[04:11] <stoned> I could backup and restore the sites/db etc. on a new server install
[04:11] <twb> If it was an "edge" package, the order might be different.
[04:11] <stoned> however, our ip will change
[04:11] <stoned> And they use their own DNS to poitn to our server
[04:12] <stoned> And having them do that takes almost 4-6 weeks (it's a global corporation and paperwork etc.)
[04:12] <twb> Anyway, deciding on which path to choose is up to the customer; you just need to make them aware of the costs & risks of each choice
[04:13] <stoned> Ok. Thanks.
[04:13] <stoned> I let my CEO know. He will talk to them himself.
[04:14] <stoned> Worst case scenario, we built a new vps for them on new ubuntu
[04:14] <stoned> they should just let us handle the domain but blah
[04:14] <twb> You might point out that 10.04's already two years into its five-year EOL
[04:15] <twb> https://en.wikipedia.org/wiki/Three-point_estimation
[04:17] <twb> If you use that technique, you give a biiiiig number for the worst case to demonstrate how iffy the backporting can be
[04:18] <twb> Like ½/2/14 days
[04:18] <stoned> I always do that
[04:19] <stoned> I learned from my mentor that if it takes 10 hours to do something, tell them 60 best case scenario.
[04:19] <stoned> I mean I get you.
[04:19] <twb> Haha, that's just inflation
[04:19] <stoned> Heh.
[04:19] <stoned> Well I don't do any billing so I dunno
[04:19] <stoned> I just job it
[04:20] <twb> I give three-point estimates to my management, who will either make the decisions or bundle it with other estimates and pass the decision up to the customer
[04:20] <stoned> I am reading it
[04:20] <stoned> intersting article
[04:20] <stoned> thanks
[04:20] <twb> It's really no harder than single-point estimates, and it gives management a lot better understanding on your confidence level for a given task
[04:22] <twb> Of course its important to understand that "worst case" is also ONLY an estimate.  It only covers the "known unknowns" as Cheney put it
[04:22] <stoned> how come your nickname is twb
[04:22] <twb> It's just an MIT-style username.
[04:23] <stoned> Initials?
[04:23] <twb> Xavier Yusuf Zbiegniew ==> xyz
[04:23] <stoned> That's a wicked name if it were a real one
[04:23] <twb> It's historical; hispanic orthography now uses "javier"
[04:24] <stoned> sounds like heavier
[04:25] <stoned> I find Spanish to be a rather funny sounding language, no disrespect intended
[08:39] <lynxman> morning o/
[09:06] <linocisco> anybody tried ubuntu mail server on virtualbox? INternet will be through Host XP
[09:18] <RoyK> linocisco: it should be like running on iron
[09:19] <linocisco> RoyK, umm. it is to test. what do you mean running on IRon?
[09:19] <RoyK> hardware
[09:46] <Tzunamii> bare-metal is the correct phrase :)
[10:18] <ARTSIOM> can someone please explaine me how "su - user" command works. IN documantation is said that "-" option will "Provide an environment similar to what the user would expect had the user logged in directly.". But this is not the case for me on 12.04. When I am loging in directly with the user I have a different $PATH when I am login in with "su - user"
[10:21] <ikonia> ARTSIOM:  the - assume the new users enviornment
[10:22] <ikonia> ARTSIOM: eg: su artsiom just switches user to artsiom, but keeps the current users environment, su - artsiom changes to the user as if you had logged in as that user
[10:23] <ARTSIOM> ikonia: but this is exactly what is not working for me. When I am loging in with artsiom directly I have a $PATH variable, which differs from the one I get when I login as "su - artsiom"
[10:24] <ikonia> what is the path variable for both
[10:26] <ARTSIOM> when I am login in directly I get $PATH which is defined in /etc/environment, when I am login in with su I get - "/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games"
[10:27] <ikonia> where is /usr/local/games etc set, in your .bash_profile (for example)
[10:27] <ikonia> sounds like it's not running a login shell
[10:29] <ARTSIOM>  /usr/local/games etc, is not set in users's  .bashrc and .profile files
[10:30] <ikonia> where is that set then ?
[10:31] <ARTSIOM> it is defined in /etc/login.defs, but I have no idea why it is loaded when using "su - user"
[10:33] <ARTSIOM> I am not even sure that it is really loaded from /etc/login.defs and not from somewhere elese
[10:35] <ikonia> quite surprised by that
[11:19] <RoyK> hm... with encrypted home, google tells me there's an --encrypted-home flag to useradd, but that doesn't seem to be the case for precise
[11:22] <ikonia> RoyK: doesn't core utils have to be built linked against a certain set of libs for that
[11:23] <RoyK> no idea - I just want to create a set of users on this new precise install, and I want their homes to be encrypted, so I'm looking for an easy way
[11:26] <anoman> I have problem with installing ubuntu 12.04 server using usb keyboard
[11:27] <RoyK> anoman: hm... has worked for me a few times... perhaps BIOS is set to USB legacy?
[11:28] <RoyK> not that that should matter
[11:28] <RoyK> just guessing
[11:29] <anoman> I have try but not working
[11:30] <RoyK> please check if usb legacy is set in BIOS - it might help to change that setting's status
[11:33] <anoman> when entering the BIOS settings to run normally, but when it entered the keyboard is not detected in ubuntu
[11:33] <RoyK> but did you check that setting?
[11:33] <anoman> yes
[11:34] <RoyK> and what was it set to? and did you try to change it?
[11:34] <anoman> I do many time for sure
[11:34] <RoyK> ok
[11:34] <RoyK> dunno, then
[11:34] <anoman> USB key board detected
[11:37] <zul> hallyn: i have libvirt 0.9.12 working i just need to clean things up
[11:42] <anoman> some one
[13:16] <hallyn> zul: cool, so do you have that based on upstream, not the debian 0.9.12-rc2 right?
[13:16] <zul> hallyn: upstream
[13:23] <hallyn> i've gotta start by ...  figuring out where to start.  would like to start the whole qemu re-do (as early as possible) but don't think i can do that yet
[13:23] <zul> basically i took whats in the archive right now and a big hammer
[13:25] <hallyn> what could go wrong?
[13:29] <zul> hallyn: not a big enough hammer
[13:49] <mboeru> hello huys
[13:49] <RoyK> http://imgur.com/vWIul
[13:50] <mboeru> has anyone succesfully managed to setup raid using kickstart/kickseed/preseed when deploying any ubuntu sever version?
[13:51] <mboeru> sooo I'm guessing no one
[13:52] <RoyK> !patience
[13:53] <mboeru> oh i am patient :) also searched
[13:53] <mboeru> thanks
[13:57] <ARTSIOM> can someone please explaine me how "su - user" command works. In documantation it is said that "-" option will "Provide an environment similar to what the user would expect had the user logged in directly.". But this is not the case for me on 12.04. When I am loging in directly with the user I have a $PATH, which differs from the one I get when login in with "su - user"
[14:01] <xnox> ARTSIOM: ...hence _similar_ =)
[14:02] <xnox> try `sudo -i -u user'
[14:50] <frankban> hi hallyn, yesterday I attached a patch to bug 994752: it contains lxc-ip and an updated version of lxc-start-ephemeral.
[14:51] <hallyn> frankban: yup, i saw it, haven't opened it yet (was out yesterday)  thanks for that
[14:52] <frankban> hallyn: thank you for taking a look at that
[15:17] <randomDude> I'm trying to pxe boot install _alot_ of workstations, i have pxe working... but preseed REFUSES to work! pxe boot config : http://pastebin.com/0Qu84GiP, preseed : http://pastebin.com/VNZBmXSb
[15:19] <roaksoax> randomDude: you might wanna look at http://bit.ly/uinstall
[15:19] <roaksoax> and compare your preseed
[15:20] <randomDude> i'm looking at that
[15:20] <roaksoax> randomDude: and we cannot help if we don't know what the error is if you don't show us an error
[15:20] <randomDude> well i can't _ show _ the error
[15:21] <randomDude> it keeps saying i don't have a cd inserted
[15:22] <randomDude> i think i've lost 5 hours of my life trying to make this work
[15:22] <roaksoax> randomDude: why don't you just use MAAS?
[15:23] <roaksoax> randomDude: https://wiki.ubuntu.com/ServerTeam/MAAS
[15:25] <randomDude> because i want desktops?
[15:25] <randomDude> not servers?
[15:25] <roaksoax> randomDude: then you just add the ubuntu-desktop metapackage in /var/lib/cobbler/snippets/maas_client_packages and that's it?
[15:26] <roaksoax> randomDude: http://askubuntu.com/questions/127313/how-do-i-add-different-packages-to-maas-clients-on-installation
[15:26] <randomDude> and learning how to use maas represents a much larger investment of time where as to get this preseed to work *should* be as *simple* as one variable...
[15:26] <randomDude> yeah no, i don't have the time to learn how to use cobbler, orchestra and puppet etc etc
[15:27] <roaksoax> randomDude: have you compared your preseed with the one I linked you?
[15:27] <randomDude> not to mention all the _undocumented_ bugs that go with them
[15:27] <randomDude> roaksoax: that preseed you linked is for 7.04
[15:27] <randomDude> https://help.ubuntu.com/12.04/installation-guide/example-preseed.txt
[15:28] <roaksoax> randomDude: no it is not
[15:28] <roaksoax> randomDude: try this: d-icdrom-detect/ejectboolean true
[15:28] <roaksoax> randomDude: try this: d-i cdrom-detect/ejectboolean true
[15:29] <roaksoax> randomDude: i'm guessing you configured BIOS to boot for PXE first instead from cdrom right?
[15:29] <randomDude> obviously
[15:29] <roaksoax> randomDude: try this: d-i cdrom-detect/eject boolean true
[15:30] <roaksoax> randomDude: even with a minimal preseed file the installer will continue asking questions
[15:30] <roaksoax> randomDude: I recommend you look at http://blog.dustinkirkland.com/2012/01/ubuntu-quick-installation-preseed-link.html
[15:31] <mdeslaur> hallyn: how is bug 999681 a virt-manager bug, if virsh doesn't show them either?
[15:31] <roaksoax> randomDude: go to the line: d-ipkgsel/include string byobu vim openssh-server and add "d-ipkgsel/include string byobu vim openssh-server ubuntu-desktop" and if you follow what's specified there in the blogpost, you should have an installation with no questions asks at all
[15:31] <hallyn> mdeslaur: bc virt-manager isn't actually creating them?  i'm going to mark it as also affecting libvirt, was just in the middle of replying
[15:31] <roaksoax> randomDude: *and* this is *NOT* for a 7.04 as you mentioned above
[15:32] <randomDude> still thinks there is a cdrom to use
[15:32] <hallyn> i like to do things in as many steps as possible
[15:32] <mdeslaur> hallyn: well, he does say the xml files are in the directory...but I'll follow the bug too, thanks
[15:32] <hallyn> oh i missed that bit
[15:33] <randomDude> roaksoax: what am i supposed to do with this blog post you sent me to that just links me to the first url you sent me to ?
[15:34] <mdeslaur> hallyn: sorry, I'm not trying to play the bug bounce game, I was just curious if I had missed something :)
[15:34] <hallyn> no, i'm just doing things in silly orders while catching up on email
[15:35] <roaksoax> randomDude: this is how my pxe file looks like: http://pastebin.ubuntu.com/989101/
[15:36] <roaksoax> randomDude: using a netboot image on a tftp server rather than nfsroot as you are trying to
[15:36] <randomDude> roaksoax: remember i want to install desktops not servers
[15:36] <roaksoax> randomDude: which I suggest you use (a netboot image)
[15:37] <roaksoax> randomDude: installing the desktop is simply adding 1 package to the package list (ubuntu-desktop)
[15:37] <roaksoax> randomDude: and you tell what packages to install in the preseed file
[15:39] <roaksoax> randomDude: i have never pxe booted from a live image so I can't help you with that. I, however, would suggest you use the netboot image, grab the preseed as linked above, add ubuntu-desktop to the package list and that'd be it
[15:39] <hallyn> stgraber: is bug 999187 related to the sources.list but you mentioned before wrt amd64?
[15:39] <stoned> hi
[15:40] <randomDude> roaksoax: i am using the alertnate iso
[15:40] <randomDude> roaksoax: also alot of the settings in that preseed are unsuitable for desktop
[15:40] <stoned> I need to setup mod_wsgi w/ python 2.7 on lucid. I came in here last night and I was told that backporting it to lucid would be a very difficult task. I could try to build python 2.7 from source or install a PPA. How would I then configure mod_wsgi to use the 2.7 python instead? please advise
[15:41] <stgraber> hallyn: yes, that's exactly what I fixed in quantal
[15:41] <randomDude> roaksoax: so i used your preseed, and it still freaks out that it can't mount a cd rom
[15:42] <randomDude> roaksoax: should i press the left key and select <yes> try looking for the cdrom again? or should i press the right key and select <no>
[15:42] <randomDude> roaksoax: does this explain the error enough for you yet?
[15:42] <hallyn> stgraber: was there an open bug for the q one?
[15:43] <stgraber> hallyn: I don't think so, I just saw it when doing something else in the template and fixed it immediately
[15:44] <stgraber> hallyn: btw, for some reason -ubuntu55 was removed from -proposed. I poked cjwatson to have it published again :)
[15:44] <roaksoax> randomDude: again, I have not tested this on an alternate cd. I have tested this on a netboot image
[15:44] <Daviey> zul: you are chairig the meeting?
[15:44] <zul> am i?
[15:44] <hallyn> stgraber: ok thanks, guess we'll add that one to the list to sru in 56 :)
[15:44] <Daviey> zul: rota seems to suggest that?
[15:45] <stgraber> hallyn: well, actually, I think you could just re-upload -ubuntu55 with that extra change as according to LP our current -ubuntu55 was "removed from disk" :)
[15:45] <randomDude> roaksoax: i thought alternate iso was meant to be used with preseed
[15:45] <zul> Daviey: who was the last one?
[15:45] <hallyn> feh.  not sure i have access to my copy of 55 right now
[15:45] <stoned> I need to setup mod_wsgi w/ python 2.7 on lucid. I came in here last night and I was told that backporting it to lucid would be a very difficult task. I could try to build python 2.7 from source or install a PPA. How would I then configure mod_wsgi to use the 2.7 python instead? please advise
[15:46] <stgraber> hallyn: wget https://launchpad.net/ubuntu/precise/+source/lxc/0.7.5-3ubuntu55/+files/lxc_0.7.5-3ubuntu55.dsc https://launchpad.net/ubuntu/precise/+source/lxc/0.7.5-3ubuntu55/+files/lxc_0.7.5-3ubuntu55.debian.tar.gz https://launchpad.net/ubuntu/precise/+source/lxc/0.7.5-3ubuntu55/+files/lxc_0.7.5.orig.tar.gz
[15:46] <stoned> If I build python 2.7 from source, and then build mod_wsgi from source and link against 2.7 python, do you think that would work?
[15:46] <Daviey> zul: i have NFI, i just know that your name is now there.. is that wrong?
[15:46] <zul> Daviey: i have no idea
[15:46] <Daviey> zul: i think it was utlemming
[15:46] <hallyn> i see i thought i'td gotten deleted :)  ok thanks
[15:46] <zul> Daviey: ill chair the meeting though
[15:47] <stgraber> hallyn: the binaries have, apparently the source is still around :)
[15:47] <Daviey> zul: super.. we'll i'm not going to be able to be entirely there..  Can you ensure that "Quantal Development".. has a strong focus on blueprint writing.. Would be good to be in a position to be able to review them next meeting.  Ta!
[15:48] <roaksoax> randomDude: yes, but in your particular case, cdrom question being asked sounds to me like something is missing in the preseed, or maybe is due to a kernel argument, or its because you are booting from nfsroot
[15:48] <zul> Daviey: ack
[15:48] <stgraber> hallyn: actually, let me do it quickly, I'll stack an -ubuntu56 entry on top of our -ubuntu55 and use -v to include both entries in the changelog
[15:48] <stgraber> will be cleaner and may avoid some LP weirdness caused by the removal
[15:48] <hallyn> stgraber: ok.  thanks.  (i assume 0081-fix-multiarch-install was the actual fix - but will let you :)
[15:51] <Bilge> So I just upgraded from 10 to 12 LTS
[15:51] <Bilge> There's a few mysterious packages appeared in aptitude that think they have been selected for installation
[15:51] <stoned> What if I install a latest ubuntu chroot inside lucid
[15:51] <stoned> install 2.7 python and apache there
[15:51] <Bilge> Like mbr and g++-4.4
[15:51] <Bilge> Should I install these or is it fine to just deselect the,?
[15:51] <Bilge> them*
[15:51] <stoned> and serve apache out of there and point the traffic there
[15:52] <stoned> Either no one knows, or no one cares to tell me.
[15:52] <stoned> :(
[15:57] <rbasak> Any comment on how to SRU bug 988819 please? I've written up the options I can think of in #6, but I don't like any of them. Is there anything else we can do?
[15:58] <EvilResistance> rbasak:  if you have a patch to solve this then it could be more capable of SRUing
[15:58] <EvilResistance> SRUs usually arent version upgrades, so if you proposed that, well...
[15:58] <EvilResistance> ah
[15:58] <rbasak> EvilResistance: the question is how to write the patch. I've detailed options in the bug.
[15:59] <EvilResistance> rbasak:  was this fixed in Quantal?
[15:59] <EvilResistance> or rather, is this fixed in QUantal
[15:59] <rbasak> EvilResistance: it will be when packages are synced from Debian, yes
[16:00] <EvilResistance> is it fixed in this?  https://launchpad.net/ubuntu/+source/modsecurity-apache/2.6.5-2
[16:00] <EvilResistance> sincie that was synced from Sid recently
[16:01] <rbasak> The fix will need an apache2 merge, too.
[16:01] <rbasak> I should do that.
[16:01] <EvilResistance> should probably fix it in Quantal before you start looking at SRUing to Precise
[16:02] <EvilResistance> the last SRU-ish-related thing i dealt with for Precise (I'm on BugSquad so i see these sometimes), the fix had to exist in Quantal, so...
[16:02] <hallyn> stgraber: say, i forgot to add 'generalize apparmor patch to enable selinux' to the lxc blueprint.  do you object to my adding it to the action list (for myself)?
[16:02] <rbasak> Yes, it's a requirement to have it fixed in Quantal before upload. Doesn't mean that I can't work on both at once :)
[16:02] <EvilResistance> :P
[16:03] <rbasak> Here, the fix for Q is trivial as it's in Debian. The SRU is more involved.
[16:03] <EvilResistance> mh,
[16:03] <EvilResistance> mhm*
[16:06] <stgraber> hallyn: I certainly don't have a problem with you adding more work to your own todo list ;)
[16:07] <hallyn> kthx
[16:08] <EvilResistance> and, there it goes :P
[16:09] <blkperl> can someone set the importance of bug 932064
[16:09] <EvilResistance> blkperl:  to?
[16:09] <EvilResistance> actually lemme see
[16:09]  * EvilResistance is on BUgSquad and can get it triaged if needed
[16:09] <blkperl> EvilResistance: i would say high, it prevents amanda from working on precise
[16:10] <blkperl> works on oneiric and quantal :S
[16:10] <EvilResistance> blkperl:  refer to this first
[16:10] <EvilResistance> https://launchpad.net/ubuntu/+source/modsecurity-apache/2.6.5-2
[16:10] <EvilResistance> erm
[16:10] <EvilResistance> wait
[16:10] <EvilResistance> sorry
[16:10] <EvilResistance> https://wiki.ubuntu.com/Bugs/Importance
[16:10] <EvilResistance> mispaste
[16:10] <stoned> If I build python 2.7 from source, and then build mod_wsgi from source and link against 2.7 python, do you think that would work?
[16:10] <EvilResistance> refer to that page first, then confirm you recommend high
[16:10] <stoned> I need to setup mod_wsgi w/ python 2.7 on lucid. I came in here last night and I was told that backporting it to lucid would be a very difficult task. I could try to build python 2.7 from source or install a PPA. How would I then configure mod_wsgi to use the 2.7 python instead? please advise
[16:11] <blkperl> EvilResistance: ok then yes, it has a severe impact on a small portion of ubuntu server users
[16:11] <blkperl> so importance High please
[16:12] <EvilResistance> blkperl:  does it also prevent the application (and/or dependencies) from working?
[16:12] <EvilResistance> nevermind
[16:12]  * EvilResistance answered his own question
[16:14] <autif> I installed matchbox window manager (on 12.04) - however, after reboot - matchbox did not come up - what else do I need to wire matchbox, so that it starts on reboot?
[16:17] <EvilResistance> blkperl:  refresh the bug, it should be high now
[16:18] <EvilResistance> s/high/Importance: High/
[16:19] <blkperl> EvilResistance: thanks, appreciate the assistance
[16:28] <stgraber> hallyn: ubuntu56 is now in proposed
[16:30] <hallyn> stgraber: cool, thanks
[16:31] <smb> zul, smoser So just to repeat, there likely will still be a linux-virtual and linux-image-virtual meta-package. As there is still those for linux-server and linux-image-server. Just pointing to generic binary packages
[16:31] <smoser> right.
[16:31] <smoser> i know.
[16:32] <smoser> but its still a change. if nothing else, there will no longer be /boot/vmlinux-*-virtual anymore
[16:33] <smb> Yes, true. Just mention it as the discussion about detecting a server install went slightly confusing (at least for me)
[16:33] <pdtpatrick> So here's an interesting question - if ubuntu by default sets up the first user with sudo access rather than create a root account (disabled by default), how can u boot into recovery mode and drop to a root shell if the root account has no password/disabled ?
[16:44] <claude2_> anyone here know how to lock down rpcbind and all the NFS stuff to a specific interface?
[16:47] <pdtpatrick> claude2_: check /etc/default/nfs-kernel-server
[16:47] <claude2_> this is for the client side
[16:47] <claude2_> do i even need rpcbind on the client side?
[16:48] <pdtpatrick>  then /etc/default/nfs-common ?
[16:49] <claude2_> yeah i checked there, but didnt see anything in there. I guess it might not be possible
[16:49] <claude2_>  i can't find anything on google either
[16:49] <pdtpatrick> check the meta package and see what it installed
[16:53] <pdtpatrick> if on the server u only installed nfs-kernel-server then see whether that pulled in anything else
[16:53] <pdtpatrick> apt-rdepends <packagename>
[16:53] <pdtpatrick> or u can search for rpc  .. dpkg -l *rpc*
[17:28] <nijaba> zul: hello.  Are we tracking bug #920197 ?
[17:29] <zul> nijaba: now we are
[17:30] <nijaba> zul: thanks :)
[17:34] <pdtpatrick> hmmm MaaS's python script will hang if u've already have an OS installed rather than starting anew. For some reason it tries to use "maas" user to configure postgres but it never prompted to set that up or grant it privileges nor does it state what credentials it is trying to use so you end up with a Traceback and it does not exit. Not cool
[17:34] <pdtpatrick> will have to run it with strace
[17:35] <RoyK> evening
[17:36] <RoyK> anyone that knows how I can have ubuntu /etc/skel include a symlink, public_html -> /var/www/$username ? or should I script that somehow?
[17:38] <nathwill> RoyK, i think that's easier done with apache directive: UserDir public_html ?
[17:38] <RoyK> nathwill: doesn't work too well with homedir encryption ;)
[17:39] <ikonia> should do
[17:39] <ikonia> as the user will decrypt it
[17:39] <RoyK> while logged in, obviously, but not after logging out
[17:39] <RoyK> since that will umount the user's filesystem
[17:39] <RoyK> for obvious reasons
[17:39]  * RoyK just checked
[17:39] <ikonia> ah, I see what you mean
[17:40] <nathwill> hrm
[17:41] <RoyK> nathwill: less point of having homedir encryption if it's mounted all the time
[17:41] <nathwill> yeah, agreed
[17:42] <nathwill> i'm wondering if you can get /etc/default/useradd to run scripts...
[17:42] <nathwill> looking now...
[17:42] <nathwill> because putting this in /etc/skel i don't think is going to work since the target is different per user
[17:42] <RoyK> yeah, agreed
[17:44] <nathwill> so... i'm seeing discussion of some concept of dynamic symlinks that will create the target if it does not exist...
[17:44] <nathwill> may be worth investigating
[17:46] <nathwill> only other thing i can think of is a cronjob... :(
[17:49] <RoyK> nathwill: or a wrapper script
[17:49] <nathwill> royk, yeah...
[17:49] <RoyK> damn - should have been a trigger for a custom script in useradd
[17:52] <hallyn> zul: would you mind looking at and considering sponsoring http://people.canonical.com/~serge/dialog.debdiff ?
[17:52] <zul> hallyn: sure
[17:57] <Daviey> hallyn: make sure you submittodebian :)
[17:57] <hallyn> Daviey: did
[17:58] <Daviey> \o/
[18:42] <Ryan_Lane> on all of my precise virtual machines, I'm getting soft lockups like this: http://pastebin.com/m7ZfhQJB
[18:42] <Ryan_Lane> not just for nscd, though. for random processes
[18:43] <Ryan_Lane> hm. ignore me. I seem to be getting them for lucid too
[18:57] <resno> i am getting "temporary failure resolving '*.ubuntu.com'". i am unsure how to correc this. 12.04
[18:58] <zul> smoser: ping
[18:59] <zul> Ryan_Lane: can you trigger it reliably?
[18:59] <Ryan_Lane> well, I think it's due to filesystem operations timing out under very high load
[18:59] <Ryan_Lane> we're using glusterfs underneath the instances, and it is performing terribly
[19:00] <zul> Ryan_Lane: ext4?
[19:00] <zul> Ryan_Lane: er...what fs?
[19:00] <Ryan_Lane> ext4 on the instance, then qcow, then glusterfs, then ext3
[19:01] <zul> Ryan_Lane: can you open up a bug in launchpad please than we can get the kernel team to look at it, or even on #ubuntu-kernel
[19:01] <Ryan_Lane> well, I have a feeling it isn't actually a kernel bug
[19:02] <zul> smoser: ^^^
[19:02] <zul> or utlemming: ^^^
[19:02] <Ryan_Lane> and there's so many levels of indirection that I doubt a report is going to be terribly useful
[19:02] <Ryan_Lane> when we switch off gluster I'll report an issue if we're still having it
[19:02] <zul> k
[19:02] <Ryan_Lane> I have a really good feeling gluster is my issue
[19:11] <veebull> hello, having some trouble getting squid3 set up... client machines are supposed to be set to use a proxy for their updates, but I'm still getting a ton of 403 errors whenever they try to use the squid proxy
[19:23] <RoyK> veebull: does manually setting the proxy work? if so, are the clients listed under the squid ACL in its config?
[19:23] <smoser> zul, whats up?
[19:24] <zul> smoser:  you mentioned last week at uds that you wanted to see the openstack-ci do some benchmarking on the images, what do you mean by that
[19:25] <smoser> veebull, your squid3 ocnfig is probably just not allowing proxying for those hosts (per its config).
[19:25] <zul> hallyn: i should have libvirt up tomorrow so you can review and play with it
[19:25] <veebull> RoyK, So far I just have the default setup - I think I need to define an acl for the local lan
[19:25] <smoser> veebull, you can just install squid-deb-proxy, which is taylored to that, or, you could just look at the config that it uses.
[19:25] <veebull> here is the (stripped) version of my acl:  http://paste.ubuntu.com/989497/
[19:25] <hallyn> zul: cool, thanks.
[19:26] <veebull> smoser, already have squid-deb-proxy running for the ubuntu-based vms... trying to set up 'regular' squid to work with the RH-based distros
[19:26] <smoser> zul, i meant that if we're running some instances in a per-every-something (commit, day....) then we might as well be running some thing inside those instances, and timing how long that something takes.
[19:27] <zul> smoser: ah ok
[19:27] <smoser> which would potentially give us insight into a kvm performance enhancement, regression, change...
[19:27] <smoser> (but it will unfortunately be hard to isolate the numbers as we have no way that i know of to get a consistent host environment ... ie, one day 2 guests may go to 1 host, the next to a different one)
[19:28] <smoser> veebull, ah... i suggest just removing the RH-based distros.
[19:28] <smoser> thats the easiest thing to do :)
[19:29] <veebull> smoser, true... but probably not going to happen ;)
[19:30] <RoyK> veebull: vi /etc/squid/squid.conf, look for acl
[19:30] <veebull> I've edited their /etc/yum.conf files to point to a proxy @ 'http://192.168.1.1:3128', which is the ubuntu server address on the lan
[19:30] <RoyK> veebull: default config is to only allow access from localhost
[19:31] <veebull> RoyK, ah.  un commented the line to allow squid to use a disk cache, now I need to set up a lan acl, yes?
[19:31] <RoyK> yep
[19:31] <RoyK> create an acl for lan
[19:32] <RoyK> and set allow on that named acl later
[19:32] <veebull> cool, I'll give it a go.
[19:32] <RoyK> veebull: the config file is rather self-explainatory if you try to read it ;)
[19:34] <veebull> yeah... I hate it when people say 'TLDR'... but this may be one of those times.
[19:34] <veebull> more a case of the colors used on the terminal screen I'm on make reading comments very difficult
[19:34] <veebull> on the eyes.
[19:35] <Destreyf> I was wondering if anyone had any idea's on how to handle nodes that do not appear in the dashboard.  I used the enlist feature on Ubuntu ISO, the environment looked like it provisioned, but it hasn't shown up in the MAAS node panel yet.
[19:38] <claude2_> is it a risk to have portmap exposed to the internet?
[19:38] <claude2_> err rpcbind i guess
[19:38] <claude2_> like can it leak anything important?
[19:39] <RoyK> veebull: with vim, :set background=dark
[19:39] <RoyK> claude2_: historically, portmap has been rather bad, but I don't think much bad has happened the last five years, at last
[19:41] <claude2_> RoyK: we've got internal and external interfaces on our NFS clients. I've been searching all day for a way to make it only listen to the internal interface
[19:43] <RoyK> claude2_: ufw/iptables is quite easy to use in that respect
[19:44] <claude2_> yeah I'm trying to do it without iptables first for political reasons
[19:44] <RoyK> I think it uses tcpwrappers as well
[19:44] <RoyK> might be more politically correct ;)
[19:45] <claude2_> haha yeah. I saw that somewhere, but haven't been able to make it work yet. I'll keep fiddling
[19:45] <claude2_> thanks
[19:45] <RoyK> edit /etc/hosts.deny
[19:48] <veebull> RoyK, adding that acl did the trick, thanks!
[19:49] <Gojko> anyone here that can help about installing ubuntu server installation on old HP proliant server
[19:49] <RoyK> !ask
[19:49] <Gojko> !patience
[19:51] <Gojko> !ask
[19:51] <stoned> hey guys
[19:51] <stoned> http://pastie.org/3916869 here is my problem in pastebin. Could anyone kindly take a look and please advise?
[19:52] <stoned> I require a solution to this problem
[19:52] <qman__> stoned, if it's chrooted and running on a different port, you just point to the different port
[19:53] <qman__> in that configuration the ports define which to ask
[19:53] <RoyK> stoned: I don't think "requiring" a solution will work too well in here ;)
[19:53] <stoned> Would you kindly help me find a resouce online that sheds more liehg on this
[19:54] <qman__> building a chroot is complicated and difficult, though
[19:54] <stoned> everywhere it's about mod chroot for apache
[19:54] <Gojko> well,i guess i havent said what is problem,apparently it seems i need drivers for a scsi controler,i have tried with installing CentOS 6.2 but it did not recognise any disks in system,and since i am very uninformed with this , if there is any1 that can point me out any kind of info i can use to solve this,atm i am downloading nev ubuntu server,with hope it will recognise my controller,or
[19:54] <Gojko> if there are any additional drivers needed,anyone could point me out to any info
[19:55] <stoned> not quite that difficult I just need a refresher
[19:55] <stoned> debootstrap latest ubuntu, mount binds the device tree and other things
[19:55] <qman__> you have to build a base environment in a directory
[19:55] <stoned> I can't remember all the things
[19:56] <RoyK> Gojko: those hp servers may come with a raid controller, which should be supported, but will need to be configured to export the drives either individually (JBOD) or as a RAID
[19:56] <RoyK> Gojko: you can usually configure that with the "smart start" CD
[19:56] <qman__> Gojko, depends on how new/old it is and which raid controller
[19:58] <RoyK> qman__: for an "old" HP server, I think those should be supported. My guess is that the hardware RAID isn't configured
[19:59] <qman__> probably, but I've also seen some cases where old SCSI drivers aren't loaded by default because they conflict with other things
[19:59] <claude2_> RoyK: does the nfs client really need portmap/rpcbind anyway?
[19:59] <RoyK> claude2_: iirc yes
[19:59] <qman__> yes
[20:07] <Gojko> huh,alot info at start,RoyK i bought this server as "used" so i dont have any CD or documentation that should or should not have came with it
[20:08] <Gojko> @:qman__ i can check what controler is in,but my best guess so far is googling compatibility and trying different versions of distros
[20:08] <qman__> Gojko, have you configured the raid? it prompts you to press a key combo during boot to do so
[20:09] <Gojko> it prompts me to press <ctrl-S> if that is what you mean,ill check it and be back in few sec
[20:09] <qman__> I have a compaq proliant DL380 which more or less works out of the box
[20:10] <qman__> pentium 3 era, only things missing are fan control
[20:10] <claude2_> yeah most likely it will work fine
[20:10] <qman__> not important in a data center or server room
[20:10] <RoyK> Gojko: you can find that smartstart cd with a quick google
[20:10] <qman__> hugely important when you're sitting in front of it
[20:16] <Gojko> well i have checked a bit,server is arround 10 years old so... ill try to find that startsmart CD
[20:17] <claude2_> i think you can just download the newest one?
[20:17] <claude2_> like its not model-specific i dont think
[20:19] <Gojko> well as i said,i am very uninformed about linux,but,since this is for my gamehouse i am holding i want to make it to work as dedicated counterstrike server for my clients.
[20:19] <Gojko> i found this startsmart CD and downloading it
[20:20] <RoyK> try that first :)
[20:20] <Gojko> link is a bit long,if anyone wants to check am i on the right path
[20:20] <Gojko> http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=15351&prodSeriesId=316537&swItem=MTX-597d7cb6b45d493285e27c1412&prodNameId=3279705&swEnvOID=181&swLang=8&taskId=135&mode=4&idx=1
[20:20] <RoyK> I'd say try it
[20:20] <RoyK> it's quite possible that none of us has your hardware
[20:22] <qman__> that should do
[20:22] <Gojko> well, i cannot afford some fancy expencive server so this one will be more than good to do it,and i dont want to put windows on it,i wand best for my customers ;)
[20:22] <qman__> though I have to ask, counter strike 1.6? otherwise it may not be fast enough
[20:22] <qman__> game servers are a heavy CPU load
[20:23] <Gojko> yes 1.6
[20:23] <qman__> ok
[20:23] <Gojko> WELL
[20:23] <Gojko> ON INTEL G620
[20:23] <Gojko> soz caps
[20:23] <RoyK> !caps
[20:23] <RoyK> :)
[20:24] <Gojko> 2 servers+cafe software+AVP+ deepfreeze console and i have servers on 1000 FPS so i think i can hold 4 servers here if on linux
[20:24] <Gojko> any1 can give opinion
[20:25] <RoyK> what sort of server is this?
[20:27] <Gojko> dedicated CS server
[20:27] <Gojko> only thing it will serve for
[20:28] <RoyK> just wondering about cpu+memory+ec
[20:28] <RoyK> just wondering about cpu+memory+etc
[20:28] <Gojko> in your opinion,can it hold 4 srv on 1000 FPS,since i will not use it for anything else
[20:29] <RoyK> you still haven't said what sort of cpu it has or how much memory it's got...
[20:31] <Gojko> 1 Gb RAM,xeon 2.8
[20:31] <Gojko> it is ml350 g3
[20:32] <RoyK> not a whole lot, but then, I don't know much about gaming servers
[20:32] <qman__> single or dual xeon, how many cores
[20:32] <qman__> may not be enough RAM for four, not sure
[20:33]  * RoyK somewhat doubts those CPUs are multicore
[20:34] <Gojko> i think single,i can get another 1 Gb
[20:34] <qman__> one single core may not be enough either, but it'll be close
[20:34] <RoyK> - Up to 1 GB of 2-way interleaving capable PC2100 DDR SDRAM, with Advanced ECC capabilities (Expandable to 8GB)
[20:34] <Gojko> it is delivered with 1 gb,supports up to 8
[20:35] <qman__> best to just try it and find out, since you already have the server
[20:35] <RoyK> most cheap laptops may beat that :)
[20:35] <qman__> monitor resources and load up the servers
[20:36] <RoyK> qman__++
[20:36] <qman__> true, but most cheap laptops don't have raid, remote management cards, or cost $20 like my P3 did
[20:37] <RoyK> indeed, but then, if you don't need hw raid or remote management, a cheap mobo, cpu and memory will go a long way
[20:38] <qman__> yep
[20:38] <RoyK> sw raid may be just as good as hw raid in most cases
[20:38] <Gojko> well as qman__ said, it was BANG for a BUCK ;)
[20:39] <qman__> I'm running a minecraft server on just such a setup, a sempron "single core" am3
[20:39] <qman__> which is actually dual core, and running at ~3.5GHz instead of the default 2.7
[20:46] <Gojko> well i guess i gotta try to get it up and running first,it is easiest for me to put load on it,so ill first download all of these isos and after that i will try to get it working
[21:31] <blendedbychris> if i am getting mount.nfs: requested NFS version or transport protocol is not supported what should i be looking for
[21:31] <blendedbychris> mount -o proto=tcp,nfsvers=3 -t nfs sld-web-4:/srv /srv.gluster
[21:39] <pdtpatrick> blendedbychris: maybe it wants to use version 4 instead of 3 ?
[22:02] <blendedbychris> pdtpatrick: no?
[22:04] <pdtpatrick> what version of NFS do u have installed?
[22:04] <blendedbychris> pdtpatrick: it's just nfs-common on precise so i pressume its v3 and v4
[22:04] <blendedbychris> from man
[22:05] <rockets> How is this information generated on login? Is there a way I can edit what shows up? http://pastie.org/3917485
[22:06] <pdtpatrick> rockets: look in /etc/update-motd.d
[22:07] <rockets> pdtpatrick, thanks!
[22:25] <rockets> pdtpatrick, actually it appears to be non-trivial to edit. hmm
[22:25] <blendedbychris> pdtpatrick: is it true that nfs-common should support either version?
[22:25] <pdtpatrick> right but why are you trying to use version 3 ? does it work with version 4 ?
[22:26] <blendedbychris> gluster doesn't support v4
[22:26] <blendedbychris> http://download.gluster.com/pub/gluster/glusterfs/3.2/Documentation/AG/html/sect-Administration_Guide-GlusterFS_Client-NFS.html#id1720347
[22:27] <pdtpatrick> worked fine on my glusterfs setup. Instead of using nfs
[22:27] <pdtpatrick> i used mount.glusterfs server1:/test-volume /mnt/glusterfs
[22:27] <blendedbychris> what?
[22:28] <blendedbychris> i am trying to specifically use nfs
[22:29] <pdtpatrick> r u on solaris ?
[22:29] <blendedbychris> no
[22:29] <blendedbychris> trying to use nfs because it can help io issues
[22:29] <blendedbychris> instead of mounting using the gluster client
[22:29] <pdtpatrick> I'm only seeing vers=3 mentioned under solaris
[22:30] <pdtpatrick> does this not work for u ? mount -t nfs server1:/test-volume /mnt/glusterfs
[22:30] <blendedbychris> no that times out presumably because it's trying to use udp
[22:31] <blendedbychris> these guys were saying i had to use portmapper
[22:31] <blendedbychris> i wonder if the nfs port is wrong
[22:31] <blendedbychris> tried  :38467 no luck
[22:31] <pdtpatrick> do u get this message ?
[22:31] <pdtpatrick> requested NFS version or transport protocol is not supported.
[22:31] <blendedbychris> yes'
[22:32] <blendedbychris> OR if i don't specify the protocol it just times out
[22:32] <blendedbychris> mountproto to be specific
[22:33] <pdtpatrick> mount -o tcp -t nfs server1:/test-volume /mnt/glusterfs
[22:33] <pdtpatrick> btw -- what happens when u run "showmount -e <ipaddr>"
[22:33] <pdtpatrick> do u see any exports ?
[22:33] <blendedbychris> ^^ that's not a valid option
[22:34] <blendedbychris> clnt_create: RPC: Program not registered
[22:34] <blendedbychris> presumably i can use a hostname in that command/
[22:35] <blendedbychris> either way same issue
[22:35] <pdtpatrick> if show mount returns nothing - then there's probably nothing being exported?
[22:35] <blendedbychris> what do you mean by that?
[22:35] <pdtpatrick> on the server ur trying to connect to  -- what does /etc/default/nfs-kernel-server say ?
[22:36] <pdtpatrick> it would tell u the ports and such
[22:36] <pdtpatrick> then in there -- do u have anything in /etc/exports ?
[22:36] <blendedbychris> pdtpatrick: gluster runs the nfs server
[22:36] <blendedbychris> it's a nfs server on it's own
[22:36] <pdtpatrick> thanks for that clarification
[22:40] <pdtpatrick> well I'm out of ideas then.  maybe ask in #gluster  and share your fix when found please
[22:40] <pdtpatrick> also thought current versions of NFS used tcp by default
[22:40] <pdtpatrick> TCP is the default transport protocol used for all modern NFS implementations.   It  performs
[22:40] <pdtpatrick>        well  in  almost  every  conceivable  network  environment  and provides excellent guarantees
[22:40] <pdtpatrick>        against data corruption caused by network unreliability.  TCP  is  often  a  requirement  for
[22:40] <pdtpatrick>        mounting a server through a network firewall.
[22:41] <pdtpatrick> sorry for spam .. was supposed to be paste bin
[22:41] <blendedbychris>  yea i was bugging those guys and it seemed like they were hinting towards "not our problem, it's your lack of nfs knowledge"
[22:41] <blendedbychris> heh
[22:51] <blendedbychris> pdtpatrick:
[22:51] <blendedbychris> doh
[22:51] <blendedbychris> http://www.gluster.org/community/documentation/index.php/Gluster_3.2:_Using_NFS_with_Gluster
[22:53] <pdtpatrick> lol what ?
[22:53] <pdtpatrick> u hadn't brought the volume set up ?
[22:53] <pdtpatrick> :(
[22:53] <blendedbychris> missed the option
[22:53] <blendedbychris> no
[22:53] <blendedbychris> nfs is disabled by default apparently
[22:53] <pdtpatrick> well that's interesting
[22:53] <pdtpatrick> so what makes u think u get better io ?
[22:54] <blendedbychris> Using a platform's native NFS client will result in best performance when reading many small files (web serving). However, no automatic fail-over will happen in case of brick failure.
[22:54] <blendedbychris> docs say so? heh
[22:54] <pdtpatrick> on small files yes
[22:54] <pdtpatrick> because gluster using xattr
[22:54] <blendedbychris> well that's my usecase
[22:54] <pdtpatrick> i c
[22:54] <pdtpatrick> well looks like ur all set :)
[22:54] <blendedbychris> indeed
[22:54] <blendedbychris> thanks for your persistence :)
[22:55] <pdtpatrick> np, glad to have "tried" to help
[23:07] <sako> hey guys, i rolled my own repo with reprepro, what's the easiest way to pin it so apt prefers any package in my custom repo over all others?
[23:09] <qman__> sako, make your version numbers higher than those in the public repos
[23:09] <sako> HUH
[23:09] <sako> oops sorry for caps
[23:10] <sako> isn't there a way to pin all packages from a repo with 999 or whatever the highest # is so that apt will always install it over other repos?
[23:14] <qman__> sako, not that I'm aware of, you set that per-package when you create them
[23:14] <sako> hmm i don't think so.. i found the man file
[23:14] <sako> man apt_preferences
[23:15] <sako> thanks
[23:18] <nathwill> sako: crunchbanglinux.org/forums has some pretty detailed apt-pinning guides for debian
[23:19] <nathwill> i imagine a lot of that would apply
[23:35] <three18ti> In a local network, how can I configure DNS so I can address each host by host name instead of by IP?  Initially, I had thought if running a dedicated DNS server, but this will require configuring the DNS for each new guest, and I'd like something more automatic.  I would prefer not to use DHCP for addressing my hosts.
[23:39] <ikonia> you need a name service such as dns / ldap and you either need to use dhcp to update, or your need to manually manipulate the name service record for each new host
[23:45] <postpac1> i'm running 12.04 desktop but having some raid/mdadm issues and #ubuntu suggested someone here might have more experience in that area.  basically my raid 5 was working perfectly, i upgraded to 12.04 from 11.10, and as soon as i rebooted everything went to hell.  mdadm is saying all 3 disks have wrong raid level and can't assemble mbr metadata
[23:58] <three18ti> ikonia, ok, thanks.  That's kinda what I was thinking, but was hoping someone had some magic.