[00:18] making a lampp server for a php based video sharing cms. need budget server for this and ffmpeg video conversion. I was thinking this server: http://www.pricegrabber.com/computers/servers++hp-proliant-dl380-g6-perf-xeon/m-730172506/#_product_details ::With this video card:: http://www.pricegrabber.com/computers/video-cards++isilence-radeon-hd-7750-h775p1gd-video-card/m-1018343980/?search=hd+7750#_overview [00:21] SunStar: what's your question? [00:21] wrong room sorry [00:22] but if some one wants to answer, i'm wondering if that is a good investment [00:50] anyone know how to get linux to be able to use more than 8 dvb tunrse? [00:50] i can see them all in /dev/dvb/adapterX [00:50] but when i start vdr, it only sees 8 [02:05] how do I use https://launchpad.net/ubuntu/quantal/i386/linux-source-3.5.0/3.5.0-23.35 [02:06] installing it results in a 83megs linux-source-3.5.0.tar.bz2 and 2 debian.* dirs. [02:07] ... [02:10] dongs: [ bunzip2 linux-blah-blah-blah.tar.bz2 && tar linux-blah-blah-blah.tar ] [02:11] SonikkuAmerica: im well aware of that. why is tehre debian dirs and is the .tar.bz2 the shipped kernel or does it need extra patches or some other crap. [02:13] dongs: It'll pull in binutils, coreutils and bzip2 (if you don't have them) and it recommends gcc, libc-dev and make [02:23] SonikkuAmerica: that doesnt answer my question [02:23] it appears the .tar.bz is not the shipped kernel. [02:23] but just generic 3.5.0 sources. [02:24] I wante whatever source tree thats used to bulid whatever is "3.5.0-23-generic #35~precise1-Ubuntu" [02:24] dongs: Hence the term linux-source [02:24] wrong [02:24] https://launchpad.net/ubuntu/quantal/i386/linux-source-3.5.0/3.5.0-23.35 -> linux-source-3.5.0_3.5.0-23.35_all.deb (81.4 MiB) [02:24] this implies the patches are included. [02:25] dongs: They won't be written as 3.5.0-23-generic #35, it they'll be written as 3.5.0 because that's the officially assigned number for them. [02:25] so how do I turn this into buildable 3.5.0-23-generic #35~precise1-Ubuntu kernel source tree. [02:25] -23-generic is patches. [02:26] which are not included in the .tar.bz2. [02:27] http://bcas.tv/paste/results/VxXbAU76.html I would have expected top of properly patched Makefile to be different [02:28] if this kernel tree was to be used to build 3.5.0-23.35 [02:28] so. [02:29] and there isn't even a .config [02:29] If you wish to learn how to compile a kernel from source, here: https://help.ubuntu.com/community/Kernel/Compile [02:30] i dont need to build a dep of new kernel. [02:30] all I want is 1) source tree matching precisely the kernel version dispalyed by uname, which I can build. thats all. [02:30] dep=deb [02:30] orw ahtever. [02:32] it seems there's bits and pieces of .config spread around debian.master/* dir [02:34] Well, I dunno anything beyond that. [02:34] i wish lunix would stop fucking with official releases of shit so I can just download it and use it [02:34] instead of having to figure out some idiotic distro specific way to do stuff. [02:35] !langauge | And FYI, dongs, it's the other way around [02:35] And FYI, dongs, it's the other way around: Please watch your language and topic to help keep this channel family-friendly, polite, and professional. [02:35] * 1000 for every non-irrelevant distro, and its quickly obvious why no serious manufacturer wants to do drivers for their device in lunix. [02:38] so back to my original question [02:38] what is the simplest way to get a buildable kernel source tree for a given ubuntu-patched kernel version. === thumper is now known as thumper-afk [05:01] Q: what is the simplest way to get a buildable kernel source tree for a given ubuntu-patched kernel version.. I don't want to build a .deb or anything. Just the patched source, which, if built, would result in the binary identical to currently running kernel. [05:22] dongs: apt-get source linux should do it. [05:22] except it doesnt [05:25] Right, because linux is also the name of a binary provided by linux-meta. [05:25] Install the ubuntu-dev-tools package and then pull-lp-source linux $RELEASENAME [05:25] That will get it, i just checked. [05:26] You could also use 'apt-get source --only-source linux' for a pure apt way of doing it [05:28] I want *exact* source used to build whatever uname -r reports. [05:29] wehn I follow sutff in https://wiki.ubuntu.com/Kernel/Dev/KernelGitGuide or https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel it ends up wiht Makefile that doesn't specify exact kernel version. [05:29] for example what i get from git step is http://bcas.tv/paste/results/VxXbAU76.html [05:29] that is NOT 3.5.0-23 or whatever. [05:34] Either what maxb said or what I did should do that. === _ffio_ is now known as ffio === smb` is now known as smb === thumper-afk is now known as thumper === thumper is now known as thumper-afk [09:01] jamespage : https://code.launchpad.net/~yolanda.robla/ubuntu/saucy/rrdtool/lua5.2/+merge/175474 [09:12] jamespage: golang stripping now removed in Debian as well. The maintainer also dropped golang-dbg as it is empty now, which I missed in Ubuntu. Something to do in the next upload, or are we going to sync at some point? [09:13] rbasak, if we can get infinity's elf header fixup into Debian as well then we can move back to syncs [09:13] rbasak, I'd not spend the buildd cycles just removing the -dbg package for the time being [09:14] rbasak, I'd forgotten todo that [09:14] rbasak, that should go upstream as well if possible [09:15] (might be in golang 1.1.2 which is due in august) [09:19] jamespage: yeah sure, not worth dropping -dbg in a separate upload, but perhaps if we need another one. Is infinity's patch upstream? [09:20] no sure [09:20] not sure rather [09:54] rbasak, i rebuilt the failed armhf build for juju-core - that now looks OK in saucy [09:56] Great! [09:58] rbasak, I wonder whether we should not be stripping that binary either [10:05] I'm writing a call-for-testing email for PHP 5.5. Do we have any documentation for users on how to test server things in the development release? Tools like LXC make this easy, so it'd be nice to point to a guide for doing this. [10:05] jamespage: AIUI, we shouldn't be stripping any go binaries at all. [10:05] rbasak, yeah [10:06] rbasak, almost like we heed dh-go [10:06] Maybe even have a lintian warning for that [10:06] todo all this stuff for us [10:06] That'd be even better :) [10:06] rbasak, I'll fixup the stripping in the next upload to saucy [10:06] (for juju-core) === thumper-afk is now known as thumper [10:27] adam_g, roaksoax: I pushed by openstack-dashboard redux branch to lp:~openstack-charmers/charms/precise/openstack-dashboard/python-redux [10:28] I think I have most things covered; unit testing of hook/contexts/utils is 100% [10:30] adam_g, roaksoax: a few comments - I pushed my unit tests into 'unit_tests' - tests is reserved for integration tests using amulet [10:31] I used '_hooks.py' instead of _relations as I thought this represented its content better [10:31] I also added setup.cfg and .coveragerc to tweak how testing and coverage reporting was done [10:59] adam_g, zul: we need to resync the havana trunk testing PPA from havana staging - there is lots missing now === hex20dec1 is now known as hex20dec [12:00] rbasak, https://launchpad.net/ubuntu/+source/juju-core/1.11.2-0ubuntu1/+build/4786692 [12:00] woot [12:00] \o/ [12:13] <\sh> tjaalton, ping how do we address the issue of raring/saucy and freeipa client 3.1.x not working with a 3.0.0 server (like on RHEL/CentOS 6) [12:14] zul: so conf.d no longer exists in apache2.4? [12:14] roaksoax: nope [12:14] zul: great! that breaks maas [12:14] roaksoax: there is a transition writeup on the debian wiki [12:14] zul: do you have a link [12:14] ? [12:15] yeah gimme a sec [12:16] roaksoax: http://wiki.debian.org/Apache/PackagingFor24 [12:16] zul: thanks! [12:59] hi [12:59] yesterday I was installing ruby and I created a file .bash_profile [13:05] jamespage: https://code.launchpad.net/~zulcss/ceilometer/ceilometer-tests/+merge/175553 <-- part 1 for enabling the tests in ceilometer [13:06] zul, btw I noticed this earlier [13:07] python-oslo.config: [13:07] Installed: 1:1.2.0a3-1ubuntu1 [13:07] zul, 1.2.0 will be challenging [13:07] :-) [13:07] (~) [13:09] jamespage: nnnngnnnnh [13:40] zul, https://code.launchpad.net/~james-page/python-cinderclient/1.0.4/+merge/175563 [13:41] jamespage: you did dch --release && debcommit --release correct? [13:42] zul, missed the last one - done now and pushed [13:42] jamespage: thanks [13:52] zul, and https://code.launchpad.net/~james-page/python-cinderclient/re-enable-trunk/+merge/175566 [13:53] yolanda: can you get python-swiftclient 1.5.0 ready for upload? [13:53] restores fixes back fro trunk. [13:53] jamespage: +1 [13:55] zul, no new upstream release of glanceclient - should I push the current snapshot in the changelog? [13:55] jamespage: yeah we should have the dep8 tests available for wider use [13:56] zul, oh, i did an MP for 1.4.0-2, has it changed to 1.5? [13:57] yolanda: yeah i rejected the MP for 1.4.0-2 we dont merge openstack stuff from debian, can you get the new tarball from pypi and get 1.5.0 MP ready [13:58] zul, ok [13:59] hi all [14:05] hey koolhead17 [14:05] howdy jamespage [14:07] hello zul :) [14:08] hi [14:18] zul, having some problems building, seems that isn't cleaning properly: http://paste.ubuntu.com/5887688/ [14:19] yolanda: can you add python-d2to1 an python-pbr to the build depends and try again [14:19] jamespage: am waiting for your signal to start testing havana pkgs :D === balloons_ is now known as balloons [14:20] zul, they are in the build-depends already [14:20] koolhead17, well havana1 is all in the Cloud Archive already [14:20] koolhead17, we are working towards havana2 right now [14:20] yolanda: wha? [14:20] next week for landing inthe cloud-archive [14:20] yolanda: *sigh* ill have a look [14:21] jamespage: cool. will start deploying & testing it per package then [14:21] zul, yolanda: just install those packages locally or build the source package with '-nc' [14:21] bzr bd -S -- -nc [14:22] let me try [14:22] works [14:22] i will have to be used to build with -nc [14:27] Can anyone point me to some instructions for setting up a PXE server to install 12.04 - I've found some instructions for doing it on 8 or 9 which I can't get to work and the few bits I've found on 12.04 don't work because the file structure on the ISOs uses aliases which seems to break the TFTP server. To clarify - it's not the setting up of the PXE server I need help with, it's the adding 12.04 Server to it. [14:31] 8 or 9? [14:32] stetho: http://archive.ubuntu.com/ubuntu/dists/precise/main/installer-amd64/current/images/netboot/ubuntu-installer/amd64/ for amd64 netboot kernel and initrd. Just get those loaded. Setting up PXE/TFTP isn't Ubuntu-specific at all. [14:32] There's a pxelinux.cfg example there too [14:32] Or if you don't want to do it by hand, use maas :) [14:32] zul, https://code.launchpad.net/~james-page/python-glanceclient/havana2/+merge/175590 [14:38] rbasak - that's the problem - I've used that but it doesn't work. I have a PXE server that installs Centos, VMWare and Ubuntu 10 but I can't get 12.04 to work. (Or 13 for that matter). From the ISOs I get file not found error which are evidently because the ISOs use symbolic links to directories like boot-screens. Getting the files from the apt repo that you've linked to gives me "violation errors" during the boot up even if everything is 777 and owned [14:38] tftpd. As I said - never had these problems with other versions. I'm missing some magical step somewhere. [14:40] stetho: try writing a minimal pxelinux.cfg directly, pointing directly to the TFTP paths for the kernel and initrd files you've downloaded. Then there can't be any symlink issues. [14:41] jamespage: https://code.launchpad.net/~zulcss/ceilometer/ceilometer-fix-tests/+merge/175596 [14:41] zul: https://code.launchpad.net/~yolanda.robla/ubuntu/saucy/python-swiftclient/1.5.0/+merge/175592 [14:43] yolanda: you didnt propose against the lp:~ubutnu-server-dev/python-swiftclient/havana branches [14:43] oh [14:44] you are right [14:44] -1 for me :( [14:47] but zul, then 1.5.0 version is already merged [14:47] latest changelog: python-swiftclient (1:1.5.0.2.gc460ebf-0ubuntu1) UNRELEASED; urgency=low [14:47] i just push to saucy? [14:48] yolanda: yes change the 1:1.5.0.2.gc460ebf-0ubuntu1 to 1:1.5.0-0ubuntu1 and then change the UNRELEASED to saucy [14:59] zul, shorter MP: https://code.launchpad.net/~yolanda.robla/python-swiftclient/havana_1.5.0/+merge/175599 [15:15] when I ssh to my ubuntu box, I get ^[[A^[[B^[[C^[[D for arrow keys. how can I fix it? [15:18] hadifarnoud: on the console? [15:18] yes RoyK. using OSX terminal [15:18] try "echo $TERM" [15:19] I use OS X and ti works well [15:19] xterm-256color [15:19] I guess it's my server [15:19] what shell are you using? [15:19] works fine on any other server [15:19] zsh [15:19] on local [15:19] bash on server [15:20] hm... works for me [15:20] RoyK: correction [15:20] I'm not using bash on server [15:21] when I typed "bash" [15:21] everything's fine [15:21] which shell are you using on the server? [15:21] try ps $$ [15:22] 13195 pts/1 Ss 0:00 -sh [15:22] if it shows dash or -sh, then you're using dash, which is a minimal shell enabled by default for some rather bizarre reason [15:22] so chsh -s /bin/bash (or zsh if you like) [15:23] jamespage: https://code.launchpad.net/~zulcss/ceilometer/ceilometer-fix-tests/+merge/175596 [15:23] thanks RoyK. after a relogin it worked [15:25] hadifarnoud: you can change the default in /etc/adduser.conf with the DSHELL setting [15:25] or variable, that is [15:25] ok [15:25] dash is useless IMHO [15:29] yeah. that's why I use zsh locally [15:38] anyone has experience with git deploy? it's kind of painful to ssh into server and git pull. [15:39] ssh into server and git pull? Why not just use git push? I use a tiny post-receive hook to keep the working directory at the far end synced with master. [15:44] yolanda: i pushed swiftclient 1.5.0....thanks [15:44] zul, easy one === NomadJim_ is now known as NomadJim [15:54] Patrickdk, jemalloc is amazing [15:55] went from slowly creeping up above 1gb per server, to a solid 300mb without growth [15:55] i put it in production on every machine [15:55] nice [15:56] now to see if it magically fixes my other problem [16:09] are there any quirks I should be aware of before setting up Striped LVM logical volume ? [16:12] or...has anyone ever st this up ? [16:12] *set [16:22] zul: https://code.launchpad.net/~james-page/python-neutronclient/havana2/+merge/175629 [16:23] jamespage: if python-neutronclient is a new binary then the epoch is not needed isnt it? [16:23] zul, I was pondering that myself [16:24] jamespage: i dont think it does [16:24] zul, no - it is because we provide a transitional package for python-quantumclient [16:24] jamespage: ok [16:24] otherwise the binary upload would be rejected post build [16:24] gotcha [16:25] jamespage: +1 [16:28] wh00t ceilometer down to 17 failures [16:34] are the instructions here intended to work in raring: https://juju.ubuntu.com/docs/getting-started.html#test? === matsubara is now known as matsubara-lunch [16:43] yep [16:43] brendand: which part isn't working for you? [16:44] jcastro, in the version of juju i got, generate-config doesn't exist === StathisA_ is now known as StathisA [16:46] can you ensure that `juju-core` is installed and not `juju`? [16:47] jcastro, aha [16:47] \o/ that should sort you out [16:48] jcastro, i gave in to the temptation to just apt-get install juju without reading the installation section :/ [16:48] yeah, that's a common problem [16:48] unfortunate package rename === VD is now known as Guest87755 [17:19] jamespage: shit...apache2 2.4 transition for horizon (just remembered) === vidplace7_ is now known as vidplace7 [17:27] zul: What is the current status on h-2 in saucy? [17:27] Daviey: we should be ready for tomorrow [17:28] zul: super. Please can you give me a shout if it looks like it won't happen? [17:28] Daviey: yep [17:28] ta [17:29] jamespage/adam_g: https://code.launchpad.net/~zulcss/horizon/apache-transition/+merge/175641 === Jikan is now known as Jikai [17:39] w0ahh one of the first mails in list from zul :) === Jikai is now known as Jikan === matsubara-lunch is now known as matsubara [18:31] i have a server thats toppping out on its amount of space, in the past. i expanded the space and then booted the server in livecd and used gparted to add more space... is this a good method? [18:32] resno, that's a great method. also consider formatting the drive for more space [18:33] formatting or an erasing the content i need? [18:33] as in* [18:33] its in a vm, so the host server has the space [18:48] I need help with crontab [18:50] Vc pode me ajudar com o crontab? [18:51] lindomar: you can ask here (best in english) -- or try in #ubuntu-pt if you want.. [19:00] resno: if you're using LVM you can resize it while running [19:01] jsonperl: i am not using crontab [19:16] Hi guys, I have been looking all over for this answer, and it is driving me crazy. I have cloud-init on 12.04.2 aws and the ulimit from /limits.conf is not being applied on cloud-final script... It's really driving me crazy, is pam_limits not loaded in cloud-init, where should i increase nofiles limits for cloud-init scripts? [19:17] breed111: if they're shell scripts, can you just add the appropriate 'ulimit' command in place? [19:17] breed111: if they are upstart scripts, you can manage limits 'natively': http://upstart.ubuntu.com/cookbook/#limit [19:18] @sarnold, i tried that ulimit in shell script, but it is launching a twistd daemon from the shell, and I don't think the twistd daemon is inheriting the limits [19:19] sarnold: I'll try that again [19:21] breed111: does the script run as root? a process needs CAP_SYS_RESOURCE to raise its hard limits [19:22] sarnold: the script runs as sudo -i -u ubuntu [19:23] hmm [19:23] breed111: aha, check /etc/pam.d/sudo, make sure the pam_limits is in the call stack somewhere.. [19:25] sarnold: this is what the pam.d/sudo looks like - http://pastebin.com/BHKHkLAD [19:26] sarnold: this method / script used to do find in 11.04, but we upgraded server to 12.04 and now does not [19:27] ok i am putting in session required pam_limits [19:27] what can I do with ubuntu server? [19:27] breed111: hrm, I don't see any pam_limits.so in my /etc/pam.d/common-* files -- I think it's just not there. [19:28] breed111: yeah, sounds like a good idea. [19:28] sarnold: thanks [19:28] breed111: (have a root shell open while testing!) [19:32] sarnold: that worked thanks - how would a root shell open help tho? [19:32] breed111: so that if you break something you have a command line available to fix it. :) [19:33] :_) [19:33] BTDT, now I recommend it to folks fiddling with PAM and AppArmor policies on login services, fiddling with firewall rules, etc. :) [20:11] when setting up a authoritative name server, do you self host or pay a company? [20:11] Unless you know what you're doing, pay someone. [20:12] who do you suggest? [20:12] many registrars offer it as a service [20:12] we presently use netwokr soltuions, but i kinda distrust doing that [20:13] i read they were ddos recently [20:13] resno: see also e.g. http://www.gandibar.net/post/2012/03/02/DNSSEC-at-Gandi [20:43] smoser: hey, just to make sure that what i'm seeing makes sense: if I start an ubuntu-cloud container with user-data setting password for user ubuntu, and that container does not have networking, then cloud-init will never set the password for user ubuntu, is that right? [20:43] o. [20:43] i.e. if no network, then user-data does not get executed? [20:43] it depends. [20:43] it could. [20:44] just a minute [20:44] k [20:44] in /usr/share/lxc/templates/lxc-ubuntu-cloud [20:44] seed_d=$rootfs/var/lib/cloud/seed/nocloud-net [20:44] if you make that /var/lib/cloud/seed/nocloud (dropping '-net') [20:45] then it will be found as a local datasource [20:45] and will not depend on netowrking at all. [20:45] * hallyn tries [20:46] but in the images as they boot is going to block waiting on a eth0 to come up [20:46] (due to /etc/network/interfaces having eth0) and portions of boot needing static-networking or the like [20:47] smoser: thanks - the parcelling of network bits is the only thing not yet done for unprivileged contaienr create/start - I just wanted to make sure I wasn't unable to log in due to something else I effed up [20:47] since you brought this up.. [20:47] ubuntu-cloudimg-query $release $stream $arch --format "%{url}\n" [20:47] (waiting 120s for upstart :) [20:48] you're waiting for static-networking-emitted [20:48] yes [20:48] just saying, after that i can confirm whether i can log in :) [20:48] yeah. you could actually add a boothook in user-data to cloud-init that (i think).... fixed that. [20:48] maybe [20:48] so that eth0 wouldn't show up [20:48] but dont know for sure. [20:49] i coudl probably just emit the event :) [20:49] smoser: but what where you going to say about 'ubuntu-cloudimg-query $release $stream $arch --format "%{url}\n"' ? [20:49] well. i was going to say that we could change that to use sstream-sync [20:49] but probably the better thing woudl be to make ubuntu-cloudimg-query use that [20:49] (should have said sstream-query, not sstream-sync) [20:50] smoser: good point, if that's doable [20:50] as for changes to te template itself, so far I'm still letting utlemming be the owner... [20:50] waiting for him to tell me he has no time [20:50] until then i don't want to step on toes [20:51] whered we get with cirros? [20:51] smoser: it's upstream [20:51] git://github.com/lxc/lxc has templates/lxc-cirros.in [20:52] I actually haven't tested it lately with unprivileged - I"ll test that [20:52] woohoo, i'm logged in [20:52] completely created/started without privilege [20:53] nice work hallyn. [20:53] oh no, but lxc-cirros is brokewn [20:53] $ sstream-query --max=1 http://download.cirros-cloud.net/ ftype=lxc.tar.gz "arch~(x86_64|amd64)" --output-format="%(item_url)s" [20:53] http://download.cirros-cloud.net/0.3.1/cirros-0.3.1-x86_64-lxc.tar.gz [20:53] http://download.cirros-cloud.net/0.3.2~pre1/cirros-0.3.2~pre1-x86_64-lxc.tar.gz [20:53] how did that happen. [20:53] oO [20:54] that's my eyes glazing over [20:54] smoser: do you want to send a patch for lxc to switch to that? [20:54] I need to patch it to recognize '--rootfs' [20:54] not rigth now. but, yeah, thats the kind of thing that we want to do. [20:54] --rootfs ? [20:56] yeah - to specify where the backing store is mounted when the template runs. [20:56] (it's not necessarily the same as lxc.rootfs in the configuration file) [20:56] lxc-create always passes it to the templates now. cirros breaks when it sees it [21:08] smoser: you're such a perfectionist, i figure you'll have som einput on this: http://paste.ubuntu.com/5888828/ [21:11] looks good i think. just reaading it. [21:11] make sure you got indentation right [21:11] (tabs or spaces, whatever i did) [21:14] yup i'm using 4 spaces [22:42] say i've got a device running on an ubuntu server ... this interface allows people to interact with a device ... and I'd like to only allow one device interaction at a time [22:43] where should this be done? === _thumper_ is now known as thumper === Aww is now known as [[Aww]]