[00:20] <cwillu_at_work> pythonirc101, you're aware that twisted has a capable ssh server which will avoid some of the issues I was discussing (although will be slightly more complex initially)
[00:21] <twb> I would be wary of any sshd implementation that isn't OpenSSH
[00:22] <cwillu_at_work> pythonirc101, the actual implementation of crypto is extremely sensitive; it'd be quite rare to have a pure python implementation, and the security of such an implementation would be very questionable given the indeterminate timing of running that code
[00:22] <twb> (Having said that, I do run lshd on prisoner desktops, because unlike sshd, it doesn't Depends: on an ssh client.)
[00:22] <cwillu_at_work> twb, don't confuse the crypto with the protocol
[00:23] <cwillu_at_work> it's quite straightforward to implement the multiplexing of ssh
[00:23] <twb> cwillu_at_work: crypto can be correct and still used improperly and thereby result in weakness
[00:23] <cwillu_at_work> twb, the multiplexing involved is not complicated from a security standpoint, that's the entire reason its there
[00:23] <twb> All I'm saying is I wouldn't use it without careful consideration
[00:23] <cwillu_at_work> you might as well say that you wouldn't trust a python web server that was hosted over https
[00:24] <twb> cwillu_at_work: I wouldn't :-)
[00:24] <cwillu_at_work> twb, then you're silly :p
[00:24] <SpamapS> We're still on this? Why on earth would you use an SSH server unless you want crazy flexibility?
[00:24] <cwillu_at_work> SpamapS, I'm not sure why he's taking that approach either, to be honest
[00:24] <SpamapS> if you want authenticated, encrypted communication, without the saddle of "users" ... TLS is quite a bit more useful.
[00:24] <cwillu_at_work> other than some requirement to use existing daemons (in an insecure way I might add)
[00:25] <cwillu_at_work> given that there are mature implementations of both protocols available though...
[00:25] <SpamapS> sshd has so many extra things
[00:25] <SpamapS> so many places to screw up
[00:26] <cwillu_at_work> most of which really aren't that scary when you're not actually hooking up to real system users
[00:26] <SpamapS> TLS identifies each side (client, optional), and then encrypts. Very simple.
[00:26] <OutOfControl> Hello
[00:26] <cwillu_at_work> again, I don't disagree, it's pretty much guaranteed to be overkill for his purposes
[00:27] <cwillu_at_work> (and he's been told this in both #python and #twisted)
[00:27] <SpamapS> :)
[00:27] <OutOfControl> I am wondering what would be better for a Amazon EC2, Ubuntu Server or Desktop and what is the difference besides Windows Manager in Desktop
[00:28]  * SpamapS goes back to reading his mountain of unread email
[00:28] <SpamapS> OutOfControl: um, what?
[00:28] <SpamapS> OutOfControl: why would you ever run Ubuntu desktop on an amazon instance?
[00:29] <OutOfControl> I'm not sure
[00:30] <SpamapS> OutOfControl: We have official AMI's to run server in EC2
[00:30] <SpamapS> OutOfControl: http://cloud-images.ubuntu.com/
[00:31] <OutOfControl> Thanks
[00:31] <SpamapS> OutOfControl: we also are working on making running specific apps in the cloud easier with juju. https://juju.ubuntu.com/
[00:31] <OutOfControl> Ok
[00:32] <OutOfControl> SpamapS: The server link on the cloud images is just a loop
[00:34] <OutOfControl> Should I use the AMD64 AMI for the micro instance?
[00:48] <tazmania> I have ubuntu-server 10.04 LTS with my nms and radius set up.  Is there a way in which I can create a customized distro based on my set up so that I do not have to go through the painful process of setting up the system?
[00:53] <twb> http://paste.debian.net/150798/ why isn't APT::Default-Release DWIMming?
[00:53] <twb> I told it I want oneiric to be D-R but it isn't using 990 priority for all my oneiric entries, only some.
[00:54] <twb> Wait a minute, it looks like it's not applying them to Translation-en only, which is fucking strange and wrong, but I don't really care
[00:55] <tazmania> How do I or can I create a customized ubuntu-server distro?
[00:56] <twb> tazmania: jigdo
[00:56] <tazmania> twb: jigdo? a new distro?
[00:57] <smw> tazmania, what are you trying to do with this "custom distro"?
[00:58] <tazmania> smw: So that I don't have to go through the same painful and lengthy process of setting up the system
[00:59] <tazmania> smw: and also for the engineers who will take over the project next time
[00:59] <twb> So you want an SOE
[00:59] <smw> tazmania, it seems like an easier way would be to just make a script that does all the changes
[00:59] <tazmania> What's SOE?
[00:59] <twb> standard operating environment
[01:00] <twb> Basically all your desktops end up configured the same way, same set of packages installed, same corporate wallpaper, etc.
[01:01] <twb> https://en.wikipedia.org/wiki/Standard_Operating_Environment
[01:01] <tazmania> smw: I do have the script which takes care of setting up additional packages.  But I still need to manually set up the mysql and other database
[01:01] <twb> For the base install you want to read up on preseeding and/or kickstart
[01:01] <tazmania> twb: yes. all the servers will have exact identical imaage
[01:01] <twb> For post-install change management you should look at puppet/chef/cfengine
[01:02] <tazmania> I just wondered how mythbuntu did it - myth-TV package + ubuntu desktop
[01:02] <twb> Alternatively you can prepare a read-only / emphemeral image, the provision umpteen of them speaking to a persistent SQL backend -- this is basically what Amazon EC2/S3 is
[01:03] <smw> tazmania, I have made custom disks before. It was awhile ago and I think jigdo did not exist
[01:03] <twb> AFAIK mythbuntu does it differently
[01:03] <twb> smw: jigdo has existed for decades
[01:03] <smw> twb, in that case I was an idiot ;-)
[01:03] <twb> smw: it was introduced about the same time as DVDs IIRC
[01:03] <twb> To avoid having to host DVD-sized images on cdimage.debian.org
[01:03] <smw> tazmania, I just made my own deb repo with everything I needed and put it on the cd.
[01:04] <twb> smw: you don't even need a CD, just put it on your httpd
[01:04] <smw> tazmania, then I made a custom preseed file and had some scripts that packaged it all into an iso
[01:04] <smw> twb, yeah
[01:04] <tazmania> So I can now do it with jigdo
[01:04] <smw> twb, for my situation, the CD made more sense
[01:05] <twb> preseeding is documented in an appendix of the installation guide; apt-get install installation-guide-amd64
[01:05] <smw> twb, we gave them out and they put ubuntu on refurb computers to be donated to people
[01:05] <twb> Ah, OK
[01:05] <smw> twb, our windows licenses only worked for non-profits
[01:06] <smw> twb, so we used ubuntu to give out functioning systems
[01:06] <twb> Not freedos ? ;-)
[01:06] <smw> lol
[01:07] <tazmania> so I should try out jigdo
[01:07] <twb> tazmania: preseeding + puppet is more likely what you want
[01:08] <tazmania> let me check it out
[01:10] <tazmania> twb: is it two different packages or in one source?
[01:10] <OutOfControl> How do I update my Ubuntu Server in EC2 without rebuilding it?
[01:10] <twb> preseeding isn't a package, it's part of the normal installer
[01:11] <twb> puppet is a separate package that does a different job
[01:11] <tazmania> ok thanks.
[01:11] <tazmania> I will check out both jigdo and puppet as recommended
[01:11] <twb> OutOfControl: I don't know, google for "ubuntu + cloud"
[01:12] <twb> OutOfControl: it's some trendy buzzword bollocks, so you will need to filter out all the naff noise
[01:12] <OutOfControl> Ok
[01:13] <OutOfControl> https://help.ubuntu.com/community/EC2StartersGuide is good
[01:14] <twb> I can't comment
[01:15] <twb> I like my OSes to run where I can go kick the hardware when it inevitably shits itself
[01:18] <twb> https://en.wikipedia.org/wiki/Cloud_computing#Issues
[01:19] <glaukommatos> So, I've set up an ad-hoc wireless access point with dnsmasq and iptables; I can access things on the same network as the machine that's doing the sharing, but is it possible for me to forward all of the traffic in such a way that it would be effectively the same as having the clients be directly on the network? (For example, I'd like things like iTunes sharing to work and zeroconf to be working.) [Forgive my rather inarticu
[01:22] <twb> glaukommatos: sure; use brctl as well as hostapd on the AP
[01:22] <twb> i.e. bridge instead of routing
[01:31] <glaukommatos> Does hostapd require that I be able to put my wireless device into master mode? I can only do ad-hoc afaik.
[01:31] <twb> AFAIK if you want an AP and you run linux, you use hostapd.
[01:32] <twb> I have only used it on managed networks but ISTR ad-hoc support
[01:32] <twb> What evidence do you have that managed mode is not supported?
[01:35] <glaukommatos> Oh, well I think managed mode does work- I had assumed that I needed it to run in master mode to run an ap (besides using ad-hoc); and when I tried to do master mode, it didn't seem to be supported by the driver (iwlagn).
[01:39] <twb> I don't know what master mode is
[01:39] <twb> hostapd should take care of putting it in the appropriate state
[01:56] <glaukommatos> Alright, I'll try this out tomorrow. I'd do it now, but I'm in the middle of a very long file transfer on the connection as it is set up currently. Thanks.
[03:10] <twb> sleepd vs. powernap -- anyone got an opinion?
[03:13] <SpamapS> twb: I've not used sleepd.. powernap is, IMO, a bit too aggressive in its default configuration for a server. But kirkland has shown me a few settings to back it off so its less dangerous.
[03:14] <twb> I'm just wondering why it exists since it looks like a new project from lp, when sleepd has existed for a while and AFAICT meets the same requirements
[03:18] <twb> I wish lp would buy the lp.net domain
[03:18] <twb> The same way sf.net -> sourceforge.net
[03:25] <twb> kirkland: why is powernap better than sleepd?
[04:12] <hallyn> zul: hey, tftpd-hpa upstart job has a type (needs 'start *on' runlevel and paren) - assume you've noticed?
[04:32] <tero> What is the most easy way to setup ubuntu server as a NAT router?
[04:34] <hallyn> tero: you can see the 'simple iptables example' at https://help.ubuntu.com/community/Internet/ConnectionSharing
[04:36] <hallyn> also, the /opt/bin/share-wlan-eth0 script at http://s3hh.wordpress.com/2011/12/15/simple-netboot-setup/ shows what I always do
[04:36] <hallyn> that's all i've got, i'm out
[05:09] <twb> tero: install it, raise both interfaces, add -o upstream -j MASQUERADE to *nat table via /etc/ufw or /etc/iptables/rules.v4
[05:10] <twb> tero: oh, and turn on ip_forward, of course
[06:36] <NullEntity> I have a quick postfix question and every remotely related channel is completely dead. Could someone here help me quick?
[06:45] <cwillu_at_work> NullEntity, ask a question, don't ask to ask
[06:45] <SpamapS> !ask
[06:49] <NullEntity> I'm trying to use Dovecot with Postfix. I can revceive with TLS fine, but I can't send. My client tells me "Authentication methods are not supported by server." I'm not sure if it's related, but ELOH fails the first time.  http://pastebin.com/QGaNq80Z
[06:57] <SpamapS> NullEntity: is there something in /var/log/mail.log at the same time?
[07:14] <chiggins> Hey guys. So I just installed bind from tasksel, now how do I add DNS entries to it? I'm new with DNS
[07:14] <chiggins> And bind
[07:15] <qman__> the configs are in /etc/bind/
[07:15] <qman__> be aware that it's very picky about syntax
[07:15] <qman__> you need to create zone files and then define them in the config
[07:15] <qman__> if you mess up, it'll show in /var/log/syslog
[07:16] <chiggins> Cool, thanks for that tip. Sorry that I'm not too dns experienced, but what's a zone?
[07:19] <twb> https://en.wikipedia.org/wiki/Zonefile
[07:20] <twb> Personally I can strongly recommend NSD for hosting zonefiles, and unbound for a caching recursive resolver.  I'm not a fan of bind's "all in one" approach
[07:22] <chiggins> Right now I'm just looking to have a local dns setup so I don't have to remember all of my ip addresses
[07:22] <twb> In that case I recommend you use dnsmasq for DNS and dhcp
[07:23] <twb> Then it'll automatically know where everything is because it'll get it straight from the DHCP request
[07:23] <qman__> I use bind for that purpose, local zones and internet name resolution for the network
[07:24] <qman__> and yes, dnsmasq is superior in that feature
[07:24] <qman__> I used bind because that's what I know
[07:24] <twb> you can do it with isc but it's a PITA
[07:25] <chiggins> Well, I figured I would just use bind, considering that's usually the result I get when I look up linux dns
[07:25] <qman__> it's been around a long time
[07:26] <chiggins> Yeah, from my light research I noticed, heh.
[07:26] <twb> widespread ≠ goo
[07:26] <twb> widespread ≠ good
[07:27] <qman__> yeah, age and popularity have benefits, but that doesn't make it the best option
[07:27] <qman__> it's not a poor choice, though
[07:27] <chiggins> So would you two maybe recommend I look into dnsmasq instead?
[07:27] <twb> chiggins: for your needs, yes
[07:28] <qman__> you'll probably have a better experience with it
[07:29] <qman__> but bind isn't nearly as bad as sendmail
[07:29] <twb> I basically don't trust ISC anything anymore
[07:29] <qman__> still get nightmares about it
[07:29] <twb> It's a red flag for me
[07:29] <twb> Not necessarily enough to shitcan it, e.g. I still run ISC (vixie) cron
[07:29] <twb> And ntpd
[07:29] <chiggins> ISC?
[07:30] <twb> https://en.wikipedia.org/wiki/Internet_Systems_Consortium
[07:30] <twb> They make reference implementations of internet software
[07:30] <twb> (As in, software that the internet runs on, not the other way around)
[07:31] <chiggins> Ah I gotcha
[08:45] <jamespage> morning all
[08:49] <_ruben> g'day
[08:51] <ttx> jamespage: morning, and happy new year
[08:51] <jamespage> ttx: happy new year to you as well!
[08:51] <jamespage> morning _ruben
[09:54] <koolhead11> hi all
[09:54]  * koolhead11 wonders is people are finally back from holiday hangover :D
[09:54] <mhubig> Hello together, I need some advice: I want to restructure the IT in the small company I work for into a cloud. Now I have found a possible solution with two XenServer and three GlusterFS Servers. But I really like the idea of the Ubuntu Cloud Infrastructure. Now my question is if the Ubuntu Cloud Infrastructure is capable of running on "just" 5 servers?
[09:55] <koolhead11> mhubig: i would say yes!! :)
[10:00] <mhubig> koolhead11: But the docu says I need at least 5 servers for storage …
[10:01] <koolhead11> mhubig: i would simply suggest you to ask the question in mailing list
[10:01] <uksysadmin> happy new year y'all
[10:02] <koolhead11> hola uksysadmin
[10:03] <uksysadmin> how's koolhead11 ?
[10:03] <koolhead11> rocking!! :D
[10:03] <uksysadmin> awesome
[11:33] <jamespage> smb: around? seeing lots of kernel oops on the latest precise ec2 images
[11:37] <smb> jamespage, yes around. can you provide, point to more specific data?
[11:40]  * smb wonders whether he had been providing a change to make precise bootable at all already... The holiday season somewhat blurred the memory
[11:50] <smb> Ah, 3.2.0-7.13 already contains the revert from upstream...
 oops
[11:51]  * smb takes back the oops... that was for the wrong tab he wrote the line before... dammit cut and paste
[11:51] <jamespage> smb: https://jenkins.qa.ubuntu.com/view/Precise%20ISO%20Testing%20Dashboard/view/Daily/job/precise-server-ec2-daily/ARCH=i386,REGION=eu-west-1,STORAGE=ebs,label=ubuntu-server-ec2-testing/29/artifact/None/i386/m1.small/ebs/i-2c54ca65/uec2-20120103-0416-9326cb93108049-terminated.console.txt/*view*/
[11:52] <jamespage> that contains the actual oops message
[11:52] <jamespage> they all appear to die in the same place :-)
[11:53] <jamespage> https://jenkins.qa.ubuntu.com/view/Precise%20ISO%20Testing%20Dashboard/view/Daily/job/precise-server-ec2-daily/
[11:53] <jamespage> although some do boot (odd)
[11:53] <smb> jamespage, Ok, thanks. Hm, something bad with setting some timer...
[11:58] <smb> jamespage, Interestingly not all of them seem failed... Err, I seem to get lost trying to move from the dashboard bullets to actual dmesgs... is there a simple path to follow that I miss?
[11:59] <smb> oh... think I got it
[12:00] <smb> some build and then console output...
[12:04] <jamespage> smb: you have to look at the build artifacts for the failed tests
[12:05] <jamespage> so clicking a red icon; then selecting the last failed built from the LHS of the screen; then click 'Build Artifacts' in the center of the screen and drill down to the console.txt files
[12:05] <jamespage> (far to many clicks)
[12:14] <smb> jamespage, Ok, thanks. Got to it now. So at least I found one case where it happened with a 3.4.3 Xen as well. So there is a chance I should be able to get the same on a local test system.
[12:14] <jamespage> smb: great - do you want me to raise a bug report for this?
[12:15] <smb> Yes, please. Then we got a place to track things.
[12:15] <smb> At least it seems to be quite likely... Not like these after a day of doing work things...
[12:16] <Ursinha> good morning :)
[12:23] <jamespage> smb: bug 911204
[12:24] <jamespage> morning Ursinha - happy new year!
[12:24] <RoyK> happy new year, all :)
[12:29] <Ursinha> jamespage: happy new year! :)
[12:59] <zul> heylo
[13:51] <daff> I am trying to preseed an installation of Ubuntu 10.04.3 but the debian-installer keeps ignoring the pre-answered questions for mirror/country
[13:51] <daff> how can I debug this?
[13:52] <daff> after dhcp configuration is complete I keep getting asked to select a mirror
[13:56] <koolhead17> hola zul
[13:56]  * koolhead17 wishes every one happy new year!! :)
[13:57] <koolhead17> daff: are you sure your preseed file is correct
[13:57] <koolhead17> ?
[13:57] <koolhead17> !cobbler
[13:58] <koolhead17> daff: https://help.ubuntu.com/community/Cobbler/Preseed
[14:01] <Error404NotFound> How can i configure the from email address that cron would use to send any output as email? i know MAILTO, MAILFROM doesn't seem to work.
[14:04] <Error404NotFound> man 5 crontab only tells about MAILTO
[14:06] <daff> Error404NotFound: the From: address in the mail headers is based on the user whose crontab is run
[14:16] <daff> koolhead17: my preseed file was correct, but the server did not serve it correctly :)
[14:16] <daff> it was essentially empty, but appeared to have been retrieved correctly (HTTP 200 and all), so d-i used it
[14:16] <koolhead17> daff: you mean DHCP/cobbler was not correctly configured
[14:17] <daff> Foreman in my case, yes
[14:20] <koolhead17> daff: so was there some bug in Foreman?
[14:20]  * koolhead17 is not aware of Foreman
[14:34] <daff> koolhead17: yes and no, a permission problem prevented the provisioning process from running correctly, but that wasn't obvious until I looked into the preseed file that was actually downloaded
[14:38] <ppetraki> hallyn, I plan to look at that libvirt/lvm bug later today, still catching up on email
[14:39] <koolhead17> daff: it be cool to file a bug then :)
[14:41] <hallyn> ppetraki: cool, thanks.
[15:50] <raubvogel> In nfsv4, how do I define the global root (using terminology from https://help.ubuntu.com/community/NFSv4Howto)?
[15:53] <pmatulis> raubvogel: it is user/administrator defined.  maybe i misunderstand you
[15:54] <raubvogel> pmatulis: from what I understand, v4 defines a root dir for NFS. So, if it is /export and you have a /export/moose, you mount it by saying mount fileserver:/moose.
[15:55] <raubvogel> How do you define /export as the nfs root or is it understood?
[15:55] <pmatulis> raubvogel: the latter.  you could also just try
[15:56] <raubvogel> pmatulis: in my case I am trying to mount the share in a windows box
[15:56] <raubvogel> can mount it as nfsv3 but am having issues mounting as nfsv4
[15:56] <raubvogel> and am trying to go down on the list of possible issues
[15:56] <pmatulis> raubvogel: i would first test with a linux client
[15:57] <pmatulis> raubvogel: i have never used (any version) nfs with 'doze
[15:57] <raubvogel> pmatulis: if you picture extreme suckyness, you are not even close ;)
[15:57] <zul> hallyn: you are running the meeting today arent you?
[15:57] <Daviey> utlemming: will you be chairing?
[15:58] <hallyn> zul: yes i am
[15:58] <hallyn> Daviey: utlemming did 2 meetings ago
[15:58] <pmatulis> raubvogel: obviously the 'doze client needs to support v4.  did you confirm that?
[15:58] <hallyn> i'll straighten out the order in the mtg page
[15:58] <utlemming> :)
[15:58] <Daviey> ahh, cool -thanks hallyn
[15:59] <raubvogel> Company claims it is the case; I am trying to validate it
[15:59] <pmatulis> raubvogel: you're using WSFU?
[15:59] <Daviey> hallyn: so is it SpamapS ?
[16:00] <raubvogel> nope. Hummingbird/opentext. Because it does kerberos and it seems the WSFU does not
[16:00] <pmatulis> Daviey: sent you an email
[16:00] <pmatulis> raubvogel: ah.  haven't used humingbird in donkey's years
[16:02] <raubvogel> Lucky you
[16:03] <raubvogel> One of the issues it seems is that they made it be more in the Solaris flavour of nfs than the linux one
[16:03] <pmatulis> raubvogel: can't you just, like, have a drone roam the office and obliterate the 'doze boxes one by one?
[16:04] <raubvogel> pmatulis: https://wiki.archlinux.org/index.php/NFSv4 seems to answer my root question
[16:04] <raubvogel> pmatulis: with a Lee Enfield?
[16:06] <pmatulis> raubvogel: a little raw, but yeah
[16:11] <pmatulis> raubvogel: fyi, i have never needed to mount the root
[16:12] <raubvogel> Me neither. The hummingbird people suggested me to setup the root in the solaris way to see if that will allow me to mount as a v4 instead of v3.
[16:16]  * pmatulis can't believe that any responsible software maker these days would cater to solaris at the expense of linux
[16:17] <raubvogel> I agree completely. And I do like Solaris. Oracle on the other hand...
[16:18] <pmatulis> 'xactly , and does oracle even support hp machinery anymore?
[16:19] <pmatulis> they dropped support for something i seem to recall
[16:19] <pmatulis> or maybe they just charge a whole lot more
[16:39] <rbasak> mdeslaur: ping
[16:40] <koolhead17> lynxman: sir
[16:43] <mdeslaur> rbasak: yes?
[16:44] <rbasak> mdeslaur: hey, regarding bug 858878 - it's a bit confusing - I prepared an update for -security but RoAkSoAx spoke to you about doing a PPA build before christmas?
[16:47] <mdeslaur> rbasak: I've subscribed ubuntu-security-sponsors. tyhicks is on community duty this week, so he'll look at it and upload it. Thanks!
[16:47] <rbasak> mdeslaur: aha, thanks. I was unaware of that team.
[16:48] <mdeslaur> rbasak: np, thanks!
[16:52] <RoAkSoAx> adam_g: pinn
[16:52] <RoAkSoAx> adam_g: bug #908895
[16:53] <RoAkSoAx> adam_g: are you manually setting the hostname to precise-x86_64 or something? or is it being done automatically?
[16:54] <adam_g> RoAkSoAx: if im installing a non-juju profile, during installation the hostname 'precise-x86_64' is automatically assigned to the node during installation, or at least attempted. d-i fails as its an invalid hostname
[16:56] <RoAkSoAx> adam_g: right, though, orchestra doesn't really set the hostname automatically, so there's maybe something done differently in cobbler to automatically assign a hostname
[16:57] <adam_g> RoAkSoAx: from what i could tell, it was inherited from the profile name. renaming the to something like precise-x86-64 fixed the issue
[16:58] <RoAkSoAx> adam_g: yeah, that's why I'
[16:59] <RoAkSoAx> adam_g: right, but the preseed doesn't really set the hostname
[16:59] <RoAkSoAx> adam_g: but it's set in the kernel parameters
[17:00] <adam_g> RoAkSoAx: *shrug* im not sure how thats set, but using "_" wherever it is set (during iso import / profile creation) needs to change
[17:00] <Daviey> RoAkSoAx: hmm
[17:01] <Daviey> We ARE looking for automatic hostname generation as a fallback
[17:01] <RoAkSoAx> adam_g: the _ is set by cobbler automatically for the profiles
[17:01] <RoAkSoAx> adam_g: what should change should be "inheriting" the hostname from the profile name
[17:01] <adam_g> Daviey: for enlisted machines, i was attempting to boot a new machine that hasn't been enlisted, and choose a profile manually from the boot menu
[17:01] <Daviey> i wold say that should be done in enlistment/
[17:01] <RoAkSoAx> adam_g: cause, as far as I can remember, that wasn't happening before, so it might be a new feature
[17:02] <Daviey> ahhh
[17:05] <RoAkSoAx> adam_g: or simply replace the _ with - unless, it used to happen that way and now it's broken
[18:00] <zul> rbasak: ping are you still worknig on those horizon bugs?
[18:01] <rbasak> zul: yes - I've done most of the lintian fixes but have been slowed down by the embedded jquery one. While you're asking - what's would be the quickest/easiest way to test the jquery functionality?
[18:02] <zul> rbasak: because im cleaning things up getting ready for the MIR
[18:03] <zul> and it was on my list of cleanup
[18:03] <zul> what jquery functionality?
[18:08] <rbasak> zul: I can send you what I have so far if you like? jquery functionality as in http://lintian.debian.org/tags/embedded-javascript-library.html - I've depended on libjs-jquery{,-ui,etc} and had just changed the path in the template
[18:08] <zul> k sure if you want to do that
[18:09] <rbasak> Is a patch series out of git OK?
[18:10] <zul> rbasak: yep
[18:10] <rbasak> zul: http://paste.ubuntu.com/791935/
[18:11] <rbasak> zul: lintian is bug 899427 and apache reload 905527
[18:11] <zul> rbasak: cool thanks
[18:12] <rbasak> zul: also I have a suspicion that debclean doesn't work due to s/override_dh_auto-clean/override_dh_auto_clean/ in debian/rules maybe but I hadn't looked at it yet
[18:12] <rbasak> I've not put the bugs in the changelog message yet
[18:12] <zul> rbasak: k ill take a look
[18:13] <rbasak> zul: shall I hand it back to you and unassign myself for now? I don't want to hold up your MIR
[18:13] <zul> rbasak: nah when ill upload it will close the bugs
[18:14] <rbasak> zul: ok but just to check you don't expect me to be doing anything more with these bugs then? I'm happy to, just want to make sure that we don't both expect the other to be doing it :)
[18:15] <zul> rbasak: right i dont expect you to be anything more with those bugs
[18:15] <rbasak> zul: cool, thanks
[18:17] <zul> rbasak: the apache stuff isnt in that patch
[18:17] <rbasak> zul: no I hadn't got that far
[18:17] <zul> k
[18:18]  * rbasak spent a while today reading debian policy to work out what to do about the jquery thing
[18:51] <SpamapS> rbasak: which jquery thing?
[18:52] <SpamapS> rbasak: I ran into a pretty nasty problem with Wordpress and jquery being out of sync on what API version to expect.
[18:52] <rbasak> SpamapS: horizon made lintian complain about http://lintian.debian.org/tags/embedded-javascript-library.html because upstream embeds jquery and related rather than using libjs-jquery et al
[18:52] <rbasak> SpamapS: ah. I was hoping it would be OK
[18:53] <SpamapS> rbasak: you're allowed to embed it if the API is different.
[18:53] <rbasak> SpamapS: well that would make things easier. Just a lintian override instead of all this messing about :)
[18:53] <SpamapS> rbasak: otherwise you need to replace the embedded file with a symlink to the jquery on disk.
[18:53] <SpamapS> rbasak: err .. s/on disk/in the package/
[18:54] <rbasak> ah. I was going to change the template to point to a different URL, but a symlink would be easier and less invasive, thanks :)
[18:54] <rbasak> although zul is doing it now
[18:55] <SpamapS> just make sure they're the same API version.
[18:55] <SpamapS> Its *hell* if they're not.. javascript doesn't exactly fail at compile time. ;)
[19:02] <zul> js is fun
[19:03] <zul> RoAkSoAx: ping have you actually powered on systems with cobbler?
[19:42] <RoAkSoAx> zul: yes
[19:43] <RoAkSoAx> zul: we have the openstack qa lab working that way
[19:43] <zul> RoakSoAx: ok sweet
[19:43] <RoAkSoAx> zul: I had to write a fence-agent for the power device there though
[19:45] <zul> RoakSoAx: how do you configure it?
[19:46] <RoAkSoAx> zul: its already configured isnt it?
[19:47] <zul> RoakSoAx: no i want to know how you configure cobbler to do it
[19:47] <RoAkSoAx> zul: sudo cobbler system edit --name <system name> --power-user=<user> --power-pass=<pass> --power-id=<id on power device> --power-address=<power device address> --power-type=sentryswitch_cdu
[19:47] <RoAkSoAx> fence_cdu -a <IP or host> -n <id on power device> -l <user> -p <pass> -o <action: on|off|status>
[19:48] <zul> RoakSoAx: cool thanks
[21:47] <mgw> anybody here familiar with isc dhcp?
[21:48] <mgw> especialy dhcrelay
[21:48] <guntbert> !anyone | mgw applies here too
[21:49] <mgw> sorry...
[21:50] <guntbert> mgw: no need to apology, just ask your question and prepare for patience :)
[21:51] <mgw> I'm having trouble getting dhcrelay to work…. as a test, i have dhcrelay listening on two vlans, relaying to an ip on one of them
[21:51] <mgw> and dhcpd is listening on that single interface
[21:51] <mgw> dhcrelay logs that it is relaying the packets, but tcpdump does not see anything
[21:51] <mgw> only the original inbound request
[21:51] <mgw> (discover)
[21:52] <mgw> tick tock… tick tock…. patience engaging….
[22:41] <RoAkSoAx> kirkland: ping?
[22:41] <kirkland> RoAkSoAx: pong
[22:41] <RoAkSoAx> kirkland: duuude how' it going
[22:42] <RoAkSoAx> kirkland: hey did you happen to find a bug in byobu where it goes nuts when running byobu under another byobu session?
[22:42] <kirkland> RoAkSoAx: hmm, i haven't;  but i don't do that often
[22:42] <kirkland> RoAkSoAx: is this tmux or screen, or both?
[22:43] <Troy_> quick question for home server and i want to have the ability to launch a gui such as xfce. should i install ubuntu-server on the machine and then install xubuntu-desktop or install xubuntu and add server software/packages?
[22:43] <RoAkSoAx> kirkland: i was first running byobu screen, then upgraded and defaulted to tmux. Then by mistake con win 0 i run byobu and it just went completely nuts
[22:43] <Patrickdk> troy, same diff
[22:44] <kirkland> RoAkSoAx_: ugh, yeah, i just reproduced that
[22:44] <RoAkSoAx_> kirkland: yeah it happens when running byobu tmux under a byobu screen
[22:44] <kirkland> RoAkSoAx_: can you open a high bug?
[22:45] <RoAkSoAx_> kirkland: sure ;)
[22:45] <RoAkSoAx_> kirkland: did you reproduce this in precise?
[22:45] <RoAkSoAx_> kirkland: i had it on lucid
[22:45] <kirkland> RoAkSoAx_: yep
[22:45] <kirkland> RoAkSoAx_: i just need to guard against an infinite loop there
[22:47] <RoAkSoAx_> kirkland: happy new year btw!!
[22:48] <kirkland> RoAkSoAx: you too, buddy!
[22:48] <kirkland> RoAkSoAx: good to hear from you
[22:49] <RoAkSoAx> kirkland: indeed!! how's downtown austin treating ya?
[22:49] <kirkland> RoAkSoAx: loving it, actually
[22:50] <kirkland> kirkland: it's fun being downtown
[22:51] <RoAkSoAx> kirkland: i bet! I actually was looking into moving to austin but I think I'm gonna stay here for a little longer
[22:51] <kirkland> RoAkSoAx: aw, come on over man
[22:51] <kirkland> RoAkSoAx: you'd have a blast over here :-)
[22:53] <osmosis> to use python pip or not to use python pip
[22:54] <RoAkSoAx> kirkland yeah i'd love too but if i start studying again its gonna be more cost efficient fpr me here
[23:03] <hallyn> kirkland: i'll need to check out your hq one day!
[23:07] <kirkland> hallyn: yeah, come on down sometime
[23:07] <kirkland> hallyn: do you find yourself downtown ever?
[23:08] <hallyn> kirkland: not too much lately, unless we need somethig from that whole foods or feel an urge to eat at chez nous
[23:08] <hallyn> but maybe i'll find a reason in february
[23:08] <kirkland> hallyn: well drop by, we're right next to that whole foods
[23:09] <hallyn> will do :)
[23:09] <mgw> roaksoax, spamaps, et al… you always have good answers… do you have any ideas about the question i posed earlier regarding dhcrelay?
[23:09] <hallyn> kirkland: oh hey, has anyone ever signed up to write the userspace ecryptfs file reader?
[23:09] <kirkland> hallyn: you!
[23:10] <kirkland> hallyn: would love to see that
[23:10] <hallyn> maybe.  would like it on my phone.
[23:11] <hallyn> but, i also have a compiz plugin or two to write :)
[23:12] <hallyn> (shiny)
[23:20] <pythonirc101> Anyone knows why my scp is getting stuck --  http://pbin.be/show/344/ -- It gets stuck at 0%?
[23:44] <aarcane_> is there a "proper" way to use vmbuilder on natty to build an oneiric VM ?
[23:46] <SpamapS> aarcane_: are you seeing a failure when trying to do so?
[23:47] <aarcane_> SpamapS, I saw several failures, and managed to work around them.
[23:47] <SpamapS> aarcane_: note that the cloud images should boot fine w/ kvm.. https://cloud-images.ubuntu.com/