[00:15] <three18ti> hey, can anyone help me with orchestra?
[00:28] <qman__> !anyone | three18ti
[00:30] <three18ti> qman__, I'll try my specific question again, but it seems that no one knows the answer to my question...
[00:30] <three18ti> Hello, how do I reset the Orchestra login?  I reset the cobbler login, but this does not seem to affect the orchestra web login.  Thanks.
[00:32] <three18ti> !poll
[00:33] <three18ti> !details
[00:33] <three18ti> !gq
[00:34] <qman__> while I don't know anything about orchestra myself, if you're having trouble getting help, fishing with an 'anyone' question isn't going to do any better
[00:34] <qman__> do you know if it's running on apache?
[00:34] <qman__> or something else, or what type of login it is?
[00:35] <qman__> keep in mind also that this is pretty new software and a lot of us have not tried it
[00:38] <qman__> anyway, login credentials are going to generally be stored in one of three ways: standard PAM authentication (/etc/passwd and /etc/shadow, or other PAM methods), a user and password in a database such as mysql, or HTTP basic authentication in an htpasswd file
[01:00] <koolhead17> hi all
[01:03] <twb> And only the former is any good
[01:04] <three18ti> qman__, I do believe that the orchestra package uses Apache to serve the web stuff, however, I am not sure what authentication module it uses...  I reset the cobbler password as it seems that "orchestra" is just the rebranding of cobbler (then you can add Juju etc...).
[01:04] <twb> qman__: it's worth noting that apache's ability to talk pam is pretty crappy
[01:05] <twb> Mostly because of the way libpam makes synchronous challenge-response calls at the wrong place in a web UI interaction :-/
[01:12] <jcastro> roaksoax: awesome, so I guess it came as a surprise to me the default orchestra install blows away machines without a prompt
[01:13] <jcastro> I was testing it with my laptop and next thing I know it was formatting my drive and reinstalling ubuntu-server on it
[01:13] <twb> jcastro: today you learn the value of off-site backups
[01:13] <jcastro> roaksoax: how do I set it to just default to a normal install instead of a preconfigured preseed?
[01:13] <jcastro> twb: right, that's not the point.
[01:13] <three18ti> well, I know cobbler stores credentials in digists; as I understand it "orchestra" is just a meta package that pulls in things like cobbler and sets everything up with post install scripts...  Standard cobbler docs don't apply since Ubuntu seems to mangle the cobbler install (i.e.: the cobbler docs say to use example.com/cobbler/web for the web interface, whereas ubuntu uses example.com/cobbler_web)
[01:14] <three18ti> jcastro, am I understanding you correctly, did you blow away your laptop hdd?
[01:15] <three18ti> jcastro, what credentials did you use to log into "orchestra"?
[01:15] <jcastro> yeah, so I guess by default choosing one of the menu options has a ubuntu-server preseed file
[01:15] <jcastro> which is an automated install
[01:15] <jcastro> which is fine, I just need to know how to turn it off
[02:05] <hansin> I read about "orchestra" the other day. Take what I am going to say here with a grain of salt as I an not fully versed, but it seems like much might be in python. Wouldn't it be nice to have a very lightweight python web server handle this stuff, something you could fire up via SSH on whatever port (kind of like the cherokee web server admin interface) you wanted? Then kill it when done?...
[02:05] <hansin> ...Might be off here, just a thought.
[02:57] <CarlFK> hansin: what's orchestra? url?
[02:59] <lifeless> https://launchpad.net/orchestra
[03:03] <boxybrown> hey guys, I installed the ubuntu-desktop package on an ubuntu-server install, and now X11 forwarding over SSH has stopped working
[03:03] <boxybrown> sshd_config has X11Forwarding set to true, I'm not sure what adding the GUI packages did
[03:08] <CarlFK> lifeless: thanks.
[05:15] <hansin> Sorry if my comment is missing the point of Orchestra. I just know for some things a simple webserver is cool. But I might not have all the info here. Anyway, cool stuff happening in the server space.
[08:35] <tjaalton> kklimonda: could you repeat what was needed for sssd? and file a bug :)
[09:10] <lynxman> morning o/
[09:12] <koolhead17> morning lynxman sirr!! :)
[09:14] <lynxman> koolhead17: ello :)
[09:14] <koolhead17> :P
[09:14] <koolhead17> Daviey, ping!! :)
[09:16] <koolhead17> lynxman, BTW weekend was awesome!!
[09:16] <lynxman> koolhead17: :D
[09:24] <Daviey> koolhead17: hey
[10:32] <kklimonda> tjaalton: mostly a new version :)
[10:53] <eagles0513875_> hey guys has anyone worked with postfix policed before? is that just an install and forget addition for postfix or does it require some configuration
[10:54]  * koolhead17 points eagles0513875_ https://help.ubuntu.com/10.04/serverguide/C/ server guide
[10:55] <ikonia> it requires configuration
[10:55] <koolhead17> hey Daviey
[10:56] <ikonia> eagles0513875_: it's not the toughest configuration, reasonably straight frward, but it does need it, it's not install and forget
[10:56] <eagles0513875_> ok ikonia  thanks :D
[10:56] <ikonia> eagles0513875_: you're not using postfix with mysql or anything like that are you ? just a straight postfix install
[10:57] <eagles0513875_> ya stright postfix install working with dovecot
[10:57] <ikonia> no problem then, it's a pretty straight forward config
[10:57] <eagles0513875_> but i want to setup some sort of spam filtering and noticed that
[10:57] <ikonia> it's not a bad tool
[10:57] <ikonia> I'm sure you'll see a benifit with it
[10:58] <eagles0513875_> :)
[10:58] <eagles0513875_> in the process though of migrating to a linode vps from home server since I'm going to be heading back to the states
[11:00] <eagles0513875_> ikonia: looking at the guides I'm guessing it would be better to setup postfix first prior to setting up dovecot correct seeing as dovecot needs to modify some of the post fix config files
[11:00] <ikonia> certainly setting up postfix first is the best idea, as without it there i no mail store for dovecot to access, and therefore dovecot is pointless
[11:01] <eagles0513875_> ok
[11:03] <eagles0513875_> ikonia: another question I'm guessing its best to follow the server guides for the release that I'm on instead of something like http://www.google.com/url?sa=t&rct=j&q=ubuntu+dovecot&source=web&cd=1&ved=0CCcQFjAA&url=https%3A%2F%2Fhelp.ubuntu.com%2Fcommunity%2FDovecot&ei=6i_KTt_gMs2F-wb4rtQo&usg=AFQjCNE0KlTjn_OdEebV-D0tSUkQULa1kw
[11:03] <ikonia> eagles0513875_: it's up to you
[11:04] <eagles0513875_> ok
[12:06] <Daviey> zul: when you are around, can you give me a shout please? :)
[12:06] <zul> Daviey: barely around
[12:07] <Daviey> zul: ah!  Did you see quantum has hit debian experimental?
[12:08] <zul> Daviey: yeah just before i was going to bed last night...i just need to fix one thing on mine then i was goin to upload
[12:14] <Daviey> zul: Does it make sense for us to base ours on Debian?
[12:22] <zul> Daviey: not really because our packaging follows the same package scheme, like python-quantum, quantum-client, quantum-server etc etc etc
[12:27] <Daviey> zul: what gain do we get?
[12:28] <zul> ease of maintaining it
[12:29] <Daviey> wait, is it easier?
[12:31] <zul> i think so if we need to update it we can just do bzr bd -S
[12:32] <Daviey> zul: if we use UDD branches, we should still be able to do that?
[12:32] <Daviey> I'm wondering if a common base with Debian makes sense for this package.
[12:33] <zul> Daviey: i dont think it will be updated often enough myself though...im happy with what we have now
[12:33] <Daviey> zul: we don't have anything, "now" :)
[12:34] <Daviey> zul: Right, but we can certainly contrib back to Debian and still enrich if required.
[13:16] <dsirijus> how come i get that there's no nginx (11.04) when running 'apt-get install nginx'?
[13:37] <lynxman> dsirijus: apt-cache search nginx shows nginx as available on my install
[13:37] <dsirijus> lynxman: hm, i0ve managed to install from the nginx ppa
[13:37] <dsirijus> lynxman: but now there's no php5-fpm
[13:37] <lynxman> dsirijus: that should do :)
[13:37] <dsirijus> just a sec, i'll pastie my sources list
[13:38] <dsirijus> (it's custom image for rackspace, maybe they've messed with it)
[13:38] <lynxman> dsirijus: https://launchpad.net/ubuntu/+source/nginx
[13:39] <dsirijus> lynxman: anything fishy in here http://pastie.org/2897800 ?
[13:40] <lynxman> dsirijus: you have the universe repositories commented, nginx is available there
[13:40] <lynxman> dsirijus: uncomment lines 12 to 15 and run apt-get update
[13:41] <lynxman> dsirijus: also 46-49
[13:41] <zul> good morning
[13:41] <dsirijus> lynxman: and what will happen now with my nginx existing install?
[13:42] <lynxman> dsirijus: well if you haven't put it in prod I'd recommend removing the ppa and using universe, it's known to be well maintained
[13:42] <lynxman> dsirijus: but that's just my opinion, I'm sure the ppa is maintained as well, although there's no warranties on that
[13:42] <dsirijus> lynxman: and how do i exactly do that?
[13:43] <dsirijus> i just apt-get installed nginx and started it
[13:43] <lynxman> dsirijus: remove the ppa config file on /etc/apt/sources.list.d
[13:43] <lynxman> dsirijus: run apt-get remove nginx
[13:43] <lynxman> dsirijus: then you can apt-get update and apt-get install nginx
[13:44] <dsirijus> can i purge it somehow?
[13:44] <dsirijus> just add --purge after remove?
[13:44] <lynxman> dsirijus: just use purge instead of remove then :)
[13:45] <dsirijus> hell, this works wonderfully :)
[13:45] <lynxman> dsirijus: :D
[13:52] <dsirijus> so friggin' rockin
[13:53] <dsirijus> i'll have zero to optimized wordpress in half an hour, and not using apache
[14:00] <dsirijus> lynxman: hey, what's the recommended dir for web sites?
[14:02] <lynxman> dsirijus: the default one is /var/www
[14:02] <dsirijus> no it ain't, i don't even have it
[14:02] <lynxman> dsirijus: hmm damn nginx *goes investigate*
[14:03] <dsirijus> so, what, it depends on the server using?
[14:03] <dsirijus> -/usr/share/nginx/www is probably default for nginx
[14:03] <lynxman> dsirijus: looks like yeah, but I'd change it if I were you, I'd rather use /var/www for websites than /usr/share/
[14:03] <dsirijus> okie
[14:08] <zul> greeeeeat
[14:11] <lynxman> zul: moaning :)
[14:54] <roaksoax> jcastro: did you add any systems?
[14:55] <jcastro> what do you mean?
[14:55] <jcastro> in cobbler?
[14:57] <roaksoax> jcastro: yeah
[14:57] <roaksoax> jcastro: or you just instaled orchestra, rebooted the machine and it automatically pxe booteD?
[14:58] <jcastro> yeah
[14:58] <roaksoax> jcastro: uhmmm cause by default it should not automatically select a profile but rather boot locally
[14:58] <jcastro> oh, I selected the profile in the menu
[14:58] <jcastro> I just wasn't expecting the preseed to be automated
[15:01] <jcastro> roaksoax: if I want to add a bunch of deskopy things, I can just plop the normal ISOs in the right iso directory and make profiles right?
[15:02] <roaksoax> jcastro: no, you need to add ubuntu-desktop as package
[15:02] <roaksoax> jcastro: you wont be able to import the desktop ISO and pxe boot from there
[15:02] <jcastro> oh ok so just modify one of the preseeds?
[15:02] <roaksoax> jcastro: and yes the profiles have a No questions asked preseed
[15:02] <jcastro> orchestra.preseed -> orchestra-desktop.preseed and modify or something
[15:03] <roaksoax> jcastro: yeah just go ahead and modify the preseed for the profile you want to use or add a system
[15:03] <roaksoax> jcastro: yeah you could do that and change the reference to the preseed you wanna use
[15:03]  * jcastro nods
[15:03] <jcastro> yeah I already made the server ones un-automated.  :)
[15:04] <roaksoax> jcastro: lol ok ;)
[15:13] <thafreak> I thought I saw something on the mailing list a while back about removing xinetd from main?
[15:13] <thafreak> I can't seem to find the discussion though...is there a better way to search mailing list archives, or does anyone know about xinetd's future?
[15:24] <WeissWaschbaer> hey.. Can i put a cron script to mount->rsync/mysqldump->umount ?
[15:25] <thafreak> sure...
[15:25] <thafreak> probably want to make sure it mounted though...
[15:34] <RoyK> WeissWaschbaer: mount && mysqldump && rync && umount
[15:34] <jMCg> Maybe even put the whole thing in a make file.
[15:35] <WeissWaschbaer> RoyK: http://paste2.org/p/1788783
[15:35] <WeissWaschbaer> i did this script
[15:35] <RoyK> or do it smarter with some if [ $? -ne 0 ]; then echo "ERROR!!!!!! WHEEWHEEWHEEE!!!" >&2; fi
[15:36] <RoyK> WeissWaschbaer: line 14 could just check the exitcode of mount ($?)
[15:36] <RoyK> if that's non-zero, something's wrong
[15:37] <jMCg> WeissWaschbaer: http://teddziuba.com/2011/02/stupid-unix-tricks-workflow-control-with-gnu-make.html
[15:37] <RoyK> WeissWaschbaer: also, check umount's exitcode before line 28
[15:38] <jMCg> Why not use make?
[15:38] <RoyK> because it's ugly? :)
[15:38] <jMCg> O_o
[15:39] <jMCg> I have an inner urge to call you an idiot, but something tells me to choose my words more wisely.
[15:39] <WeissWaschbaer> jMCg: have to learn it first, long term soltion
[15:40] <jMCg> Now, would you mind elaborating how a simple Makefile in which every operation's return code actually *is* checked, the state is recorded and can be picked up from there - how is that uglier than http://paste2.org/p/1788783 <<< This.
[15:41] <WeissWaschbaer> RoyK: the script does run like this, i just need to make cron execute it..
[15:41] <WeissWaschbaer> for some reason cron is not
[15:42] <thafreak> Make has to be explicitly installed...bash, python, etc are already on most systems
[15:42] <jMCg> ah, right. Linux distros...
[15:43] <thafreak> well, you could write posix compliant shell script then, and most systems (not just linux) default shell will work the same
[15:45] <RoyK> jMCg: this is an ubuntu channel, so, yes...
[15:57] <mgw> Does anyone know why iptables-persistent in oneiric has no status command?
[15:58] <mgw> nm, it looks like the 'status' in previous versions was useless anyway — it just returned
[16:37] <roaksoax> lynxman: any update son bug #874981
[16:37] <lynxman> roaksoax: it's done and waiting for Davieys rubberstamp
[16:45] <roaksoax> lynxman: there's one last Daviey comment
[16:45] <lynxman> roaksoax: which was solved and ironed :)
[16:45] <lynxman> roaksoax: he thought the comments were ours and not the debian package ones
[16:48] <roaksoax> lynxman: oh ok
[16:48] <roaksoax> Daviey: mcollective ready to be merged?
[16:48] <lynxman> roaksoax: :)
[16:49] <lynxman> roaksoax: thought he was dealing with it personally that's why I didn't bother to update on the ticket, apologies
[16:49] <roaksoax> lynxman: no worries ;)
[16:51] <smoser> utlemming, i launched a build of oneiric-server-cloudimg-20111121 (cloud-image) which will have the new oneiric kernel in it.
[16:51] <smoser> (ie, candidate for refresh)
[16:52] <utlemming> smoser: yes, I saw that -- I saw that you'd kicked a few minutes ago when I went looking to see if we had a build with the new kernel in
[16:54] <Daviey> roaksoax: Yeah, sorry - Friday EoD i tried to hand it over to someone else. :)
[16:54] <lynxman> Daviey: 14 hours a day not enough for you sir? jeez ;)
[16:55] <Daviey> roaksoax: http://irclogs.ubuntu.com/2011/11/18/%23ubuntu-server.html#t00:45
[16:55] <Daviey> lynxman: 14?  and the rest!
[16:55] <lynxman> Daviey: like we say in Spain "you work more hours than a clock"
[16:55] <Daviey> lynxman: the time shown is local time, so yes EoD was 00:45. :)
[16:57] <roaksoax> Daviey: yeah I remember but thoguht u were looking for a second opinion rather than an uploader
[16:58] <Daviey> roaksoax: "Would one of you mind reviewing and uploading bug" :)
[17:00] <roaksoax> Daviey: lol ok
[17:02] <lynxman> Daviey: that'd also hint for a second opinion? ;)
[17:02] <Daviey> lynxman: little bit..
[17:02] <lynxman> Daviey: also for drinks, but I see drinks everywhere
[17:04] <roaksoax> Daviey: you might help me
[17:05] <roaksoax> Daviey: quilt has patches-applied, but no record on .pc
[17:05] <roaksoax> Daviey: so when trying to unapply, then it does not aunapply and consequently, it failts to apply
[17:05] <roaksoax> so how can I recover from this situation
[17:06] <Daviey> roaksoax: on an intensive call right now.
[17:06] <roaksoax> Daviey: ok, let me know when you done please
[17:09] <dkn> using KVM, can i point a disk to an LVM? the /dev/mapper/disc can be mounted on the host system, but i can't see the partition table if i fdisk /dev/mapper/disc
[17:12] <dkn> in other words, instead of using an image file, can i instead add storage to a VM ?Guest using a LVM on the host system? i was thinking i mount the LVM on the host system, load it up with files, then unmount, add /dev/mapper/disc to the device list for the Guest and boot up and mount it, but i have nothing to mount in the VM guest, it just see's vdb with no partitions to mount
[17:13] <SpamapS> dkn: yes, any block device will do. :)
[17:13] <hallyn> yup
[17:15] <dkn> hmmm i guess i'm missing the part where the Guest recognize's the LVM?
[17:16] <RoyK> dkn: if lvm was setup on that device, it should be autodetected during boot
[17:17] <dkn> mmmm no, i setup the LVM on the host machine, the guest VM has no idea that the storage is and LVM
[17:18] <hallyn> bc you're just giving it a blockdev, yes
[17:18] <WeissWaschbaer> my cron is not running my shelscript (it works when I manually run it)
[17:19] <dkn> hmmm
[17:19] <RoyK> WeissWaschbaer: does it email you and tell you something's wrong?
[17:19] <RoyK> WeissWaschbaer: it should...
[17:19] <RoyK> WeissWaschbaer: that is, if root owns the cron job, that email will go to root
[17:19] <WeissWaschbaer> RoyK: I didnt set any email
[17:19] <RoyK> then it'll email root
[17:19] <WeissWaschbaer> how do I check that?
[17:20] <RoyK> su to root
[17:20] <RoyK> type mail
[17:20] <RoyK> or install mutt - a bit better ui
[17:20] <dkn> so i should instead mount the host's VG to get the LVM to show up as an LVM?
[17:20] <WeissWaschbaer> RoyK: not installed
[17:20] <RoyK> dkn: your guest should never see the host's lvm config
[17:20] <WeissWaschbaer> mutt is n-curses?
[17:20] <RoyK> WeissWaschbaer: then install it - apt-get install mutt
[17:21] <RoyK> WeissWaschbaer: don't remember - but it's light and works well
[17:21] <WeissWaschbaer> RoyK: it is a server, no Xorg
[17:22] <RoyK> WeissWaschbaer: mutt isn't an X app
[17:22] <WeissWaschbaer> ok
[17:23] <dkn> oh dear... nvm, just mount the /dev/sda device....
[17:23] <dkn> i was just trying to make it to complicated...
[17:23] <RoyK> :)
[17:24] <dkn> thks Spam & Roy
[17:25] <WeissWaschbaer> No mail for root
[17:25] <RoyK> WeissWaschbaer: apt-get install postfix
[18:17] <roaksoax> lynxman: so it looks good... though shouldn't the mcollective-client.config we filtering for configure and reconfigures only?
[18:18] <roaksoax> lynxman: cause otherwise it would be run at all times
[18:18] <roaksoax> (debconf question)
[18:18] <lynxman> roaksoax: hm, your call chief
[18:18] <lynxman> roaksoax: that'd put some delta from debian though :/
[18:20] <roaksoax> lynxman: so hold on... in the debdiff you attaching there's changes such as the addition of mcollective-client.config
[18:20] <roaksoax> is that done in the ubuntu1 and is a delta
[18:20] <roaksoax> or is that done in the debian package?
[18:20] <roaksoax> lynxman: err let me rephrase the sentence
[18:20] <roaksoax> lynxman: that's done as ubuntu delta
[18:20] <lynxman> roaksoax: that's debian
[18:21] <lynxman> roaksoax: we're narrowing the delta with this release, only delta now is the addition of mcollective-middleware
[18:21] <roaksoax> lynxman: right, but the latest release form debian does not contain that change in their source
[18:21] <lynxman> roaksoax: of course not, because that's on their first release, which was after ours :)
[18:22] <roaksoax> lynxman: right, but where's the *new* debian verison that contains those changes I see as additions in the ubuntu delta
[18:22] <Daviey> roaksoax: Still want me?
[18:23] <roaksoax> Daviey: nope, figured it out :)
[18:23] <Daviey> cool
[18:23] <roaksoax> thanks though
[18:23] <roaksoax> lynxman: changes in description in debian/control seem as ubuntu delta
[18:23] <lynxman> roaksoax: they're not additions, for sure
[18:24] <roaksoax> lynxman: right, but that's why I'm saying I'm confused out of your debdiff as it shows ubuntu delta which you mention are in debian
[18:24] <roaksoax> which are not
[18:24] <roaksoax> that's why :)
[18:24] <roaksoax> (maybe you went through this conversation with Daviey already)
[18:24] <lynxman> roaksoax: because that's the delta from our _latest_ package to the debian one
[18:24] <lynxman> roaksoax: yes I did :)
[18:25] <lynxman> roaksoax: so that debdiff is narrowing the delta from debian, but the debdiff is based on our current one
[18:25] <lynxman> roaksoax: because there was no delta before, since there was no package on their side
[18:25] <roaksoax> lynxman: right, but for example this things:
[18:25] <roaksoax> -Description: Marionette Collective clustering framework - server
[18:25] <roaksoax> +Description: build server orchestration or parallel job execution systems
[18:26]  * lynxman has dejavu about all this...
[18:26] <roaksoax> lynxman: that appear as ubuntu delta, changes made for ubuntu
[18:26] <lynxman> roaksoax: that's the change from Ubuntu to debian
[18:26] <roaksoax> lynxman: right, that's why I;m saying, you are adding those changes
[18:26] <lynxman> roaksoax: based on the Ubuntu one, ys
[18:27] <roaksoax> lynxman: now, --- mcollective-1.2.1+dfsg/debian/mcollective-client.config     1969-12-31 16:00:00.000000000 -0800
[18:27] <roaksoax> +++ mcollective-1.2.1+dfsg/debian/mcollective-client.config     2011-06-27 10:22:02.000000000 -0700
[18:27] <roaksoax> lynxman: mcollective-client.config does *not* exist in ubuntu
[18:27] <roaksoax> and is a delta that we are carrying
[18:27] <roaksoax> correct?
[18:27] <lynxman> roaksoax: that comes from debian, yes
[18:28] <lynxman> roaksoax: we're trying to unify the packages
[18:28] <lynxman> roaksoax: I'm in contact with hggh (the debian maintainer)
[18:28] <Daviey> roaksoax: that was the same thing i thought :)
[18:28] <roaksoax> lynxman: right, so given that case, wouldn;t it be better to wait for a released version in debian to then just sync the packages?
[18:29] <lynxman> roaksoax: no
[18:29] <lynxman> roaksoax: because... we have mcollective-middleware, which we're adding
[18:29] <Daviey> well, the merge is done - we should try and close the gap sooner rather than later
[18:29] <lynxman> roaksoax: and was in our package
[18:29] <lynxman> roaksoax: and will be in the debian package as soon as hggh catches it
[18:29] <roaksoax> ok
[18:30] <lynxman> roaksoax: so when he gets the changes back the delta will be nil
[18:30] <roaksoax> lynxman: right
[18:30] <lynxman> roaksoax: then we can sync :)
[18:30] <roaksoax> lynxman: i understand, but these changes come from a debian package not yet even uploaded
[18:31] <roaksoax> lynxman: so technically, they are ubuntu detal
[18:31] <roaksoax> delta
[18:31] <lynxman> roaksoax: call them dorothy if you want :D
[18:31] <lynxman> roaksoax: hehe
[18:31] <roaksoax> lynxman: hehe ok, dorothy, from my personal point of view, should be documented as delta, and then when debian releases, and we sync (or merge if there's any delta left) that's noted in the changelog
[18:32] <roaksoax> lynxman: and for ease of understanding
[18:32] <roaksoax> lynxman: so you wont have to answer questions like these here after
[18:32] <roaksoax> lynxman: cause others might also get confused
[18:33] <lynxman> roaksoax: noted
[18:35] <roaksoax> lynxman: so debian/control needs to be fixed, there's a few typos
[18:37] <roaksoax> lynxman: its minor though
[18:37] <lynxman> roaksoax: :)
[18:40] <roaksoax> lynxman: other than that looks good man
[18:41] <roaksoax> lynxman: so thanks for having this done
[18:41] <lynxman> roaksoax: thanks for reviewing this, it was a nightmare to merge
[18:41] <roaksoax> lynxman: yeah I can see that :)
[18:41] <lynxman> roaksoax: ;)
[18:41] <roaksoax> lynxman: but, as a recommendation, it would be better if the fdebian maintainer sets a git repo for you and him to work on so collaboration gets improved
[18:42] <roaksoax> lynxman: that's what we did with the debian maintainer of the ha stuff
[18:42] <lynxman> roaksoax: that'll happen soonish I hope
[18:42] <roaksoax> cool ;)
[18:42] <lynxman> roaksoax: great recommendation!
[18:42] <roaksoax> ;)
[19:31] <Daviey> adam_g: For cobbler-enlist, we are looking to do discovery in this order,  1) static hostname , 2) avahi discovery , 3) manual input.  Do you think this logic should be part of the tool, or a standalone app?
[19:32] <zul> i think it should be apart of the tool
[19:32] <adam_g> Daviey: cobbler-enlist should take '-s http://$server/cobbler_api' as one required argument. it shouldn't matter if it gets that from d-i, avahi, wherever.
[19:33] <Daviey> adam_g: Yes, but where does the $server come from?
[19:34] <Daviey> within the cobbler-enlist tool, or elsewhere?
[19:34] <adam_g> Daviey: 1) static hostname, 2) avahi discovery, 3) manual input  , i'd prefer that takes place elsewhere
[19:34] <Daviey> adam_g: still shove it in the same package?
[19:34] <Daviey> cobbler-discover ?
[19:35] <adam_g> and we can keep the enlist tool generic and indifferent to what discovery method is being used
[19:35] <Daviey> ok
[19:36] <adam_g> Daviey: correct me if wrong but this could all be scripted in packaging around the tool, no?
[19:37] <Daviey> adam_g: Yes, but avahi is actually easier for us to handle in c
[19:37] <Daviey> we have prior art.
[19:38] <adam_g> oh. im still a  noob. i thought there were easy shell utilities around avahi
[19:38] <Daviey> adam_g: There are, but IIRC - it linked against too much for us to create a udeb.
[19:38] <adam_g> ah
[19:38] <adam_g> hmm
[19:38] <adam_g> we want it to be somewhat automatic though, right? ie, try avahi or fall back to user input?
[19:38] <Daviey> adam_g:  I can kick off that part, if it helps.
[19:39] <Daviey> yeah
[19:39] <Daviey> adam_g: we are blocked on smoser, bug 893189 :)
[19:43] <adam_g> Daviey: can you think of anything we have out there that currently uses C libraries to interface with avahi?
[19:44] <Daviey> adam_g: Yeah
[19:44] <Daviey> adam_g: do you want to kick it off?
[19:44] <Daviey> I was going to if not.
[19:44] <adam_g> Daviey: yeah, i can take a look.
[19:45] <adam_g> i suppose we'd need to first reenable that cobbler code?  was that disabled by us or upstream?
[19:45] <b0gatyr> I have a server log entry that looks like this: 111.111.111.111#23444: ns1.somedomain.com , how can I make a new file with only IP and ns1.somedomain.com?
[19:46] <b0gatyr> in other words I would like to move the trailing #23444:
[19:46] <DW-10297> Howdy, I'm probably foolish in thinking this but didn't I read a long time ago that Ubuntu 11 has a really nice virtualization/private cloud thing built into it?
[19:46] <b0gatyr> or anything in between # and :
[19:47] <Daviey> adam_g: upstream, simply re-enabling it isn't enough.
[19:47] <DW-10297> or is that a different version of ubuntu/distro?
[19:47] <Daviey> I tried that, needs another event
[19:47] <Daviey> git blame helped
[19:56] <roaksoax> Daviey: the discovery needs to be done in C?
[19:57] <bhm> Hello chaps, I have a server with root enabled (as the only means of accessing the server) and am trying to get myself created as a user on the server, but the classic "adduser bhm admin" responds "the group admin does not exist".
[19:57] <bhm> Any good advice?
[19:58] <roaksoax> Daviey: cause koan was able to "discover" a cobbler server already so I was thinking maybe we could extend koan to do some of the stuff we need
[19:59] <roaksoax> adam_g: btw.. did you test latest orchestra in precise?
[20:00] <adam_g> roaksoax: it wasn't built yet by the time i logged off for the night on friday, but i was running the same version locally for a day or two
[20:00] <roaksoax> adam_g: ok cool
[20:03] <Daviey> roaksoax: It's really easy to do avahi discovery in pythong :)... The problem is, at the moment, we do not have python avaliable in the instaler.
[20:04] <Daviey> roaksoax: BTW, cobbler preseed retrieval failed for me when i tried it on saturday, it complained it didn't have an ssl cert
[20:04] <Daviey> "not found" style error in the logs
[20:04] <Daviey> We may have dropped the generation, or i did something silly, either way :)
[20:04] <roaksoax> Daviey: yeah that's fixed in precise already
[20:05] <Daviey> cool
[20:05] <roaksoax> Daviey: how did you drop that?
[20:05] <roaksoax> Daviey: that's generated by orchestra-logging-server
[20:05] <Daviey> roaksoax: how did i do what now?
[20:05] <Daviey> Ah, i might not have had orchestra-logging-server installed
[20:05] <roaksoax> Daviey: it is the same approach the thing is that if the file was not found it failed, so I just did a try/except and return something approprite to not fail if not found
[20:06] <Daviey> cool
[20:06] <dkn> i'm playing with FOG and i need to change the password in one of the config files... but the password is hash'd it's a ~digit hex, how can i change the password with a new hash instead of typing it out in plain text in the config file?
[20:07] <Daviey> roaksoax: I'd quite like a review of, https://launchpad.net/~davewalker/+archive/cobbler-testing/+packages
[20:07] <Daviey> roaksoax: i split out the theme into a different package, and made it detected at run time.
[20:08] <Daviey> roaksoax: it's not quite polished yet, but input appreciated
[20:08] <roaksoax> Daviey: cool, will look at it
[20:08] <Daviey> ta
[20:08] <adam_g> trivial bug #892409 is also pending review. :)
[20:08] <roaksoax> Daviey: and yeah, if we enable cobbler avahi publication then we could technically use koan like "koan --server=DISCOVER"
[20:09] <roaksoax> Daviey: byut the feature seems buggy so I guess that's why its been deprecated maybe?
[20:11] <Daviey> roaksoax: i did dig out the commit which disabled it, and it was something along the line sof making startup more efficeient or something
[20:11] <Daviey> seemed unrelated.
[20:12] <roaksoax> Daviey: weird then, I guess we'd have to check with upstream for that
[20:16] <Daviey> roaksoax: it was disabled in 2009, not removed, just commented oit
[20:16] <bhm> ...solved...
[20:17] <roaksoax> Daviey: maybe its been implemented differently?? i can't remember if I managed to get koan to automatically discover a server or not
[20:17] <Daviey> roaksoax: ah, scrub that
[20:17] <Daviey> roaksoax: "Consolidating services to use just one port,"
[20:18] <Daviey> and avahi opened another port, i guess.
[20:19] <Daviey> roaksoax: https://github.com/rubenk/cobbler/commit/f24ffa0d4d997d6c536ca7d89d6bfbc1b9dce07f
[20:30] <roaksoax> Daviey: looks good the webui stuff, are you planning to upstream this?
[20:31] <Daviey> roaksoax: planning to upstream the cobbler patch, but not the theme
[20:32] <roaksoax> Daviey: of course lol
[20:32] <Daviey> bug 893308, is on our hit list.. Anyone that wants to grab that, fee free.
[20:41] <adam_g> Daviey: we shouldn't need to have support for broadcasting avahi within cobbler. looking at squid-deb-proxy, couldn't we just advertise via an entry at /etc/avahi/services ?
[20:46] <Daviey> adam_g: Hmm, maybe.. you think that is cleaner than fixing the support cobbler used to have?
[20:48] <SpamapS> cobbler had some avahi stuff, but it went away?
[20:49] <Daviey> SpamapS: yah
[20:49] <Daviey> SpamapS: bug 893189
[20:50] <SpamapS> There's a movement to rip out unmaintained stuff going on right now in cobbler's dev list
[20:50] <Daviey> SpamapS: does bug 893308 interest you? :)
[20:51] <SpamapS> Daviey: heh, it does sound like a challenge.. but I wonder.. something 151kB, needs to be smaller?
[20:54]  * RoyK wonders what some people are smoking
[20:55] <Daviey> SpamapS: on cd, no - netboot, yeah
[21:33] <kpettit> can anybody recommend a good replacement for putty that can do tabs and go full screen in windows?  I'm forced to Citrix into a windows machine to get to my linux ones and need something better than putty/cygwin.
[21:41] <kpettit> ah nevermind, I found a good one.  mRemoteNG
[21:45] <Dulcin> When I generated opendkim keys the public key included 'r=postmaster;' but it does not seem to get validated here http://dkimcore.org/c/keycheck and I think that's why it doesn't show up when I dig for my txt records either
[21:46] <Dulcin> anyone familiar with this?
[21:49] <ersi> Dulcin: How long ago did you setup the TXT record?
[21:49] <ersi> Might take an hour before it pops up AFAIK
[21:50] <Dulcin> About 40 hours ago
[21:51] <ersi> Huh, shoulda popped up by then
[21:52] <Dulcin> That's why I'm wondering why it hasn't yet. I added it to the DNS admin panel of my domain hosting company, but when I use the validator it's complaining about the r=postmaster
[21:52] <Dulcin> maybe I'm wrong though ersi:
[21:53] <Dulcin> should 'dig txt knoep.nl @ns1.hostnet.nl' retrieve the record
[21:53] <Dulcin> or are DKIM records only retrieveable through the selector?
[21:53] <Dulcin> that would make sense actually
[21:53] <ersi> I'd name the txtrecord "@" though
[21:54] <Dulcin> shouldn't it be selector._domainkey.domain ?
[21:54] <ersi> Or oh, sorry hehe
[21:54] <ersi> I'm just used to a certain DNS panel
[21:56] <patdk-wk> depends on what you want, the dkim key, or the policy
[21:56] <Dulcin> I'm just trying to get DKIM to work
[21:56] <Dulcin> first time I'm doing it
[21:57] <patdk-wk> what is you dkim email header look like?
[21:57] <Dulcin> You mean if I test it? or how I set it up?
[21:57] <Dulcin> because I haven't set it up then I think
[21:57] <patdk-wk> heh?
[21:58] <patdk-wk> what domain?
[21:58] <Dulcin> knoep.nl
[21:58] <Dulcin> what I did was set up postfix, and created the dns record
[21:58] <Dulcin> and set up opendkim
[21:58] <patdk-wk> what selector?
[21:58] <Dulcin> knoep.nl
[21:58] <patdk-wk> that is a domain
[21:59] <Dulcin> yeah i used it as selector as well
[21:59] <patdk-wk> ok, I see that, but I see no policy
[22:00] <Dulcin> but you do see my record? http://paste.ubuntu.com/745420/
[22:00] <Dulcin> isn't that the policy? or am I confusing something
[22:00] <patdk-wk> no
[22:00] <patdk-wk> the policy is kind of like spf
[22:01] <patdk-wk> selector is just keys you use
[22:01] <Dulcin> I see
[22:01] <Dulcin> so there's my mistake
[22:01] <Dulcin> I have an spf record set, so I need to set up an additional record for DKIM then
[22:02] <patdk-wk> normally people just do _domainkey.example.com TXT "o=~"
[22:02] <patdk-wk> and _policy._domainkey.example.com TXT "o=~"
[22:02] <patdk-wk> the _policy one is for dkim
[22:03] <Dulcin> and without _policy. ?
[22:03] <Dulcin> _domainkey.example.com TXT "o=~"
[22:03] <patdk-wk> that would be domainkey policy
[22:04] <Dulcin> ahhh
[22:04] <Dulcin> no I see
[22:04] <Dulcin> now*
[22:04] <patdk-wk> http://www.sendmail.org/dkim/surveyFortune1000
[22:04] <patdk-wk> hmm, wonder what the _ssp is, have to look at that
[22:04] <Dulcin> so one more question you might know:
[22:05] <Dulcin> I tried to set up senderID as well
[22:05] <patdk-wk> good luck :)
[22:05] <Dulcin> I've emailed microsoft to add me, and they did
[22:05] <Dulcin> but when I check my hotmail, I end up in spam and senderid=temperror
[22:05] <Dulcin> even though it has been over a week since they've added me
[22:05] <patdk-wk> did you setup your senderid spf record correctly?
[22:06] <patdk-wk> I dunno what you mean by, they added you
[22:06] <patdk-wk> there is nothing to add
[22:06] <Dulcin> from what I understood, the senderid framework belongs to microsoft and is kept manually?
[22:07] <patdk-wk> http://www.microsoft.com/mscorp/safety/content/technologies/senderid/wizard/
[22:07] <patdk-wk> senderid is nothing more than an adjustment on how spf works
[22:07] <Dulcin> yeah I did that one
[22:07] <Dulcin> and then I did this one: https://support.msn.com/eform.aspx?productKey=senderid&page=support_senderid_options_form_byemail&ct=eformts&wa=wsignin1.0&st=1&wfxredirect=1
[22:07] <patdk-wk> did you add the dns records?
[22:07] <Dulcin> yeah I have 1 spf record
[22:08] <patdk-wk> ya, an spf record, but no senderid record
[22:08] <Dulcin> aha!
[22:08] <patdk-wk> spf2.0/pra == senderid record
[22:08] <patdk-wk> yours says v=spf1 == spf record
[22:08] <Dulcin> so I'm wondering though, if there's nothing to add, why did microsoft e-mail me back "We have successfully added your domain to the Sender ID program. This may take up to 2 business days to be fully replicated in our systems. If you have any questions regarding this please let me know."
[22:09] <Dulcin> (and more after that)
[22:10] <patdk-wk> how evil
[22:10] <patdk-wk> that wizard says  Idon't need an spf2.0 record
[22:10] <patdk-wk> but I had to have one, or everything went to spam folder :)
[22:10] <SpamapS> Daviey: wget-udeb already has ssl support btw
[22:11] <patdk-wk> dulcin, I can't remember how senderid records work exactly
[22:11] <patdk-wk> but based on mine, it's the same as the spf record, except for the first part is change from v=spf1 to spf2.0/pra
[22:11] <Dulcin> so I should duplicate my spf record for this?
[22:12] <patdk-wk> ah yep, that is what the wizard gives me now
[22:12] <patdk-wk> that is what I do
[22:12] <patdk-wk> if I remember correctly
[22:12] <patdk-wk> if a spf2.0 record is found, the spf1 is ignored
[22:12] <patdk-wk> but if spf2.0 is missing, it uses spf1
[22:12] <Dulcin> Ah cool
[22:12] <patdk-wk> but spf1 was only meant to over source ip, but microsoft expended it with senderid to match envelop and from addresses, causing issues
[22:13] <Dulcin> so I just found out for some reason my DNS panel does not allow me to add the txt records for some reason it doesn't like the name  _policy._domainkey - so my next question - which is what I'm fearing a bit
[22:13] <Dulcin> if I want to do DNS on my server, i have to set the nameservers on my domain hosting company to my server and set up BIND9 somehow?
[22:14] <patdk-wk> sure
[22:14] <patdk-wk> _domainkey NS ....
[22:14] <Dulcin> I tried bind9 once, got scared and used the admin panel
[22:14] <patdk-wk> I dropped bind around bind8
[22:15] <Dulcin> to replace it with?
[22:15] <Daviey> SpamapS: yeah, we need to look to make it smaller, before it is actually used.
[22:40] <SpamapS> Daviey: heh, when recompiled on precise, the udeb grows
[22:40] <SpamapS> -rw-rw-r-- 1 clint clint 152K May 17  2011 orig.wget-udeb_1.12-3.1ubuntu1_amd64.udeb
[22:40] <SpamapS> -rw-rw-r-- 1 clint clint 159K Nov 21 14:40 wget-udeb_1.13.4-1ubuntu1_amd64.udeb
[22:41] <Daviey> crap.
[22:41] <SpamapS> trying with -Os
[22:42] <Daviey> SpamapS: that is probably because you aren't using package mangler locally.
[22:42] <Daviey> ah, not sure it would do much on a udeb
[22:42] <Daviey> hmmm
[22:43] <SpamapS> does the default sbuild use pkgbinarymangler ?
[22:44] <Daviey> no
[22:44] <SpamapS> Well either way, I don't think it does much to binaries
[22:44] <Daviey> but it wouldn't be that for a udeb.
[22:44] <SpamapS> it hasn't been rebuilt since 11.04
[22:44] <SpamapS> so gcc 4.6 could just be making bigger code
[22:45] <Daviey> rocking
[22:45] <SpamapS> -Os is supposed to value code size over all
[22:45] <SpamapS> trying that
[22:45] <Daviey> \o/
[22:45] <SpamapS> I don't see any config options to tweak..
[22:46] <SpamapS> Daviey: where are we going to use wget that we couldn't use a library?
[22:47] <SpamapS> -rw-rw-r-- 1 clint clint 142K Nov 21 14:46 wget-udeb_1.13.4-1ubuntu2_amd64.udeb
[22:47] <SpamapS> compiled with -Os
[22:47] <hallyn> Daviey, so i'm looking at MIR for freeimage (needed for spice).  Noone seems to maintain it.  Do I have to say "we'd rather not maintain but will if we have to"?  :-)
[22:47] <hallyn> or is there some other way to phrase that?
[22:47] <hallyn> (common in MIRs)
[22:47] <SpamapS> hallyn: is it an optional dependency?
[22:48] <hallyn> hm, i don't think so, but will check
[22:48] <Daviey> SpamapS: sorry, what do you mean?
[22:48] <hallyn> on the bright side zaroo bugs :)
[22:48] <Daviey> hallyn: is it /really/ required?
[22:48] <Daviey> gah, SpamapS is ahead of me
[22:48] <SpamapS> Daviey: just wondering if there's a library call that can be used instead of a big giant 300+k binary like wget.
[22:49] <Daviey> SpamapS: well i think it would involve re-inventing the wheel TBH, and higher technical obverhead
[22:49] <hallyn> Daviey, SpamapS: well I think it is
[22:49] <hallyn> but it's a 2nd or 3d level depend
[22:49] <Daviey> SpamapS: Currently the busybox built-in wget is used, so it's a cheap replacement
[22:50] <Daviey> SSL validation by hand, sounds less fun
[22:50] <SpamapS> Daviey: ahhh so its already being done that way but it is lacking.
[22:50] <SpamapS> Daviey: ok, so a whopping 10kB drop for -Os ..
[22:50] <Daviey> yeah, busybox wget doesn't spport ssl apparently
[22:51] <Daviey> SpamapS: anything interesting in --$options ?
[22:51] <Daviey> or adding --without-* or somethig?
[22:51] <Daviey> (configure options)
[22:52] <SpamapS> Daviey: thats what I'm looking at now.. how those affect the code.
[22:52] <hallyn> I dunno, we'd need our own custom cegui-mk2 package to not build-depend on freeimage, though I don't see where enable-freimage is given during configure
[22:52] <SpamapS> Daviey: --disable-threads might help
[22:53] <Daviey> SpamapS: ooo
[22:53] <hallyn> lemme try building under precise with and without it i guess.  maybe i'll open a bug to drop that.
[22:53] <Daviey> hallyn: If we can avoid it, i think it would be best.  It's not the sort of package we'd normally opt to support, as you spotted.
[22:54] <Daviey> seems to be well maintained to date, as you saw.
[22:54] <hallyn> right, well, that's why the others i was willing to MIR (like svgalib), they look well maintained in debian
[22:54] <hallyn> this one, not
[22:54] <hallyn> but again, to not MIR it, we'll have to get cegui-mk2 package changed
[22:55] <Daviey> hallyn: incidently, we are ahead of Debian for freeimage
[22:55] <Daviey> so /someone/ cares about it
[22:55] <hallyn> i'm not so sure
[22:55] <hallyn> depends which 'it' you mean
[22:55] <hallyn> cegui-mk2, yes.  spice calls out to that
[22:56] <hallyn> gotta run, bbl
[22:56] <Daviey> hallyn: no comment why we jumped ahead of debian on the bug.
[22:56] <Daviey> hallyn: o/
[22:57] <hallyn> Daviey, i know!
[22:57] <hallyn> i was considering submitting it back to debian,
[22:57] <hallyn> but wanted to find the ppl involved in that bug first
[22:57] <hallyn> (but couldn't)
[22:57] <hallyn> anyway, bbl
[22:57] <Daviey> hallyn: catch you later
[23:04] <SpamapS> no threads just disables linking to libpthread .. there aren't any ifdefs
[23:17] <jacobw> i'm considering moving from asterisknow (centos based) to running asterisk on a virtualised ubuntu server for ease of management
[23:17] <jacobw> is there anything i should know about running asterisk on ubuntu?
[23:23] <jacobw> information about asterisk can be unreliable, so i'm poking a channel where people may have tried what i'm planning to do
[23:28] <Daviey> jacobw: Using freepbx?
[23:31]  * Daviey goes awol
[23:43] <jacobw> Daviey: sorry, yes, using freepbx.
[23:44] <jacobw> dang, missed the reply.
[23:58] <patrickmw> is there a way to add a new OS Version in Orchestra (precise)?
[23:58] <patrickmw> Or is that a code change?