[00:01] New bug: #1048864 in juju "Latest python in quantal breaks juju test suite" [Critical,Triaged] https://launchpad.net/bugs/1048864 === dendro-afk is now known as dendrobates === n0ts_off is now known as n0ts === cpg is now known as cpg|away [01:34] hey i added "0 0 * * * * /sbin/shutdown -h now" to my crontab, but do i need to update the sudoers file? ...to what? [02:08] rtdos, no, but you have too many stars [02:08] if it's on your personal crontab, your user must have permission to do that [02:08] adding to sudoers will not allow that to work [02:09] you'd need to change the command to use sudo [02:28] so it should be "0 0 * * * sudo /sbin/shutdown -h now" (this is for the system to reboot when not used, whether or not someone is not logged in. [02:57] rtdos, that will work but only if you have passwordless sudo to run that command [02:57] you're probably better off using 'poweroff' instead because I don't think you can specify arguments in a command in sudoers [03:06] Anyone any idea what CIFS VFS: No writable handles for inode might mean? [03:06] Im assuming its something to do with a Samba share (I was writing to the network and now all activity has dropped dead and attempting to look at/umount the share locks up the terminal) [03:07] The network seems fine on every other machine on the network, just my ubuntu-server box seems to be having problems === n0ts is now known as n0ts_off === n0ts_off is now known as n0ts === dendrobates is now known as dendro-afk [04:06] I want to setup a box running ubuntu server with "Smoothwall" like capabilities but cannot figure out which packages to use. Please help [04:08] grizcreative: What are you actually trying to accomplish and what do you need it to do? [04:08] anyone have any experience with opennms and rancid on 12.04 x64? [04:09] their channel is dead (has been for 5 days...) [04:09] I want it to handle DCHP and firewall. Sit between my modem and a switch [04:10] grizcreative: Firewall's already installed - iptables. There are two or three good DHCP servers in the repos. [04:11] grizcreative: may i whisper you? [04:11] grizcreative: Looks like isc-dhcp-server is the recommended one [04:11] https://help.ubuntu.com/12.04/serverguide/dhcp.html [04:12] grizcreative: You can also use ufw for a simpler interface to iptables - https://help.ubuntu.com/12.04/serverguide/firewall.html [04:12] tonyyarusso: thank you [04:13] there are linux OSes out there specifically designed for firewall purposes and include DHCP, WiFI, DNS, and a bunch more [04:13] SteveThing: but i want to do more with it, thus why not just using smoothwall [04:14] oh, ok [04:14] try not to put all your eggs in one basket tho... can get messy [04:14] i think smoothwall uses iptables anyways :) [04:15] If only the wife gave me unlimited budget [04:15] hah! i know, right? women and their "needs"... psh [04:16] hmm, anyone know of a script to copy a file to file.original? I know the commands, just not how to script it [04:26] New bug: #1010345 in qemu-kvm (main) "Qemu doesn't support pSeries partitions" [Medium,Expired] https://launchpad.net/bugs/1010345 === cpg|away is now known as cpg [06:56] New bug: #1022901 in qemu-kvm (main) "Ubuntu 12.04 64bit Guest (qemu/kvm) dies randomly (dup-of: 1025244)" [High,Incomplete] https://launchpad.net/bugs/1022901 [07:08] hmmm, just had a drive fail, does this mean the drive is entirely bad or is there something i should check for it? [07:08] im 'rebuilding' the drive in the raid config bios, but im not sure if that will fix it === cpg is now known as cpg|away === cpg|away is now known as cpg [07:54] https://bugs.launchpad.net/ubuntu/+source/quantum/+bug/1045064 [07:54] Launchpad bug 1045064 in quantum "Impossible to run quantum-server from testing/archives packages" [High,Confirmed] [07:55] adam_g: reopened it . cool [08:01] New bug: #916576 in samba (main) "smbd crashed with SIGABRT in push_ucs2_talloc()/lp_servicename returns NULL" [Medium,Incomplete] https://launchpad.net/bugs/916576 === zz_KristianDK is now known as KristianDK === n0ts is now known as n0ts_off === KristianDK is now known as zz_KristianDK === mcclurmc_away is now known as mcclurmc === cpg is now known as cpg|away === zz_KristianDK is now known as KristianDK [10:41] Ursinha, are we doing automatic duping of install failures now? bug 1012058 for example? [10:41] Launchpad bug 1012058 in mysql-5.5 "package mysql-server-5.5 5.5.24-0ubuntu0.12.04.1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,Confirmed] https://launchpad.net/bugs/1012058 [10:51] iostat -N should show device names, but it doesnt on one of my servers. do i need aditional package/module in kernel installed to make device-mapper show proper LV's of disk... ? using ubuntu distro. [10:52] LVS names* === KristianDK is now known as zz_KristianDK === zz_KristianDK is now known as KristianDK [12:04] iostat -N should show LV names, but it doesnt on one of my servers. do i need aditional package/module in kernel installed to make device-mapper show proper LV's of disk... ? using ubuntu distro. [12:13] https://bugs.launchpad.net/ubuntu/+source/sysstat/+bug/358382 --> is this a bug or not ? [12:13] Launchpad bug 358382 in sysstat "iostat -N doesn't report device mapper names" [Undecided,Won't fix] === n0ts_off is now known as n0ts [12:36] New bug: #1049084 in bind9 (main) "package bind9 1:9.8.1.dfsg.P1-4ubuntu0.2 failed to install/upgrade: el subproceso instalado el script post-installation devolvió el código de salida de error 1" [Undecided,New] https://launchpad.net/bugs/1049084 [12:41] New bug: #1049087 in antlr3 (universe) "gunit missing in antlr3" [Undecided,New] https://launchpad.net/bugs/1049087 [12:43] morning [12:45] zul: maybe where you are. but greetings anyway [12:56] morning zul === cheez0r_ is now known as cheez0r [13:09] w00t - minimal virtual install looking much better for quantal - thanks cjwatson! === KristianDK is now known as zz_KristianDK [13:27] hi ... trying to setup a local apt-mirror. My first attempt is using apt-mirror - the result works pretty good so far. Question: do you know other ways of setting up an apt-mirror besides using apt-mirror? Maybe one using less bandwidth in general (difficult unless it doenst download all updates in the first place i know) === zz_KristianDK is now known as KristianDK [13:39] Hello all. Trying to hook postfix up to a postgresql backend. Is there a set of tables I need to create (and a schema somewhere so that I can recreate them), or does postfix create the needed tables automatically? === KristianDK is now known as zz_KristianDK === skaet_ is now known as skaet [13:46] stgraber: where the heck do i find the cached resolvconf info that dhclient got for an interface? [13:47] hallyn: /run/resolvconf/interface/ [13:47] ah /run :) thx [13:48] hm and that's correct. [13:48] but when i had resolvconf running, no wireless drivers, and eth0 brought up with dhclient, i couldn't resolv any hosts [13:48] rebooted several times, no help. killed resolvconf and edited /etc/resolv.conf, and all is well. [13:48] it's *probably* something i did, this system si a bit of a frankenstein, but... [14:00] hey guys if you were to set up a syslog server [14:00] what service would you use? [14:00] and why? [14:02] makezan: rsyslog, because it's already installed ;) [14:02] rsyslog comes default in ubuntu server? [14:03] it's part of the minimal task, and the ubuntu-minimal metapackage [14:03] wow okay thx tsimpson will look into it. I've been trying to set it up via centos without success. [14:08] hi [14:09] i tried to use palmiset to have a gui to lvm , however after much struggling i found how to add new pv, but i still can t see how to resize [14:10] anyone has an idea of a gui that allows extending lvm partitions? [14:11] orogor: what do you want to extend? Physical volume, Volume Group or Logical Volume? there is no common term 'lvm partition' [14:11] logical volume [14:11] does Volume Group have extra space? [14:11] yup [14:12] when i click it it says only create new LV [14:12] orogor: $ sudo lvresize --size 50G my-volume-group/my-logical-volume [14:12] this will make it 50G big. [14:12] orogor: after that you need to extend the filesystem as well [14:12] :/ [14:13] system config lvmm was sooooooo good :( [14:13] orogor: $ resize2fs /dev/my-volume-group/my-logical-volume [14:13] two intuitively named commands..... [14:13] --size +100 GB extend 100GB right ? liek from 200 to 300 ? [14:14] yes, that works as well. [14:14] orogor: can you give a link to 'lvmm'? I cannot google for such thing. [14:15] xnox, http://dailypackage.fedorabook.com/index.php?/archives/92-GUI-Thursday-System-config-lvm-LVM-GUI-tool.html [14:15] (i think thats what he means) [14:15] http://www.youtube.com/watch?v=IGScfs4cXOU [14:15] opps in japanese [14:16] orogor: $ sudo apt-get install system-config-lvm ? [14:16] https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Storage_Administration_Guide/s1-system-config-lvm.html [14:16] orogor: it's packaged in Ubuntu.... $ sudo apt-get install system-config-lvm [14:16] it s broken in last versions of ubuntu, signaled upstream ,tagged as won t fix [14:17] because all of system-config is broken? [14:17] * ogra_ bets systemd issue [14:17] orogor: do you have a link where you reported it upstream? [14:17] yes it wont start cause of gtk2 vs gtk3 issue or something [14:17] it s not me , i googled the error before [14:17] stgraber: ~. [14:17] orogor: generally bugs about ubuntu should be filed on launchpad with a link to upstream tracker. [14:18] well... there are a few crash bugs about it.... [14:18] let's try that again. [14:19] stgraber: rebooted, found /run/resolvconf/interface/lo.dhclient exists, with 127.0.0.1 [14:19] stgraber: when i rm that and resart resolvconf, all is well. [14:19] not sure which dhclient caused that - the lxc, libvirt, or system-wide one [14:22] ok , resizing the ext4 [14:22] ogra_: seems to work fine here on quantal. [14:23] ogra_: there is a bug open about python2.6 string exceptions.... [14:23] i thought it could resize in background and that was instant , or thats only at inode creation ? [14:23] well, wasnt me who complained :) [14:23] * ogra_ has yet to see a filesystem resize thats instant [14:23] hallyn: that's weird... why would you have dhclient running against .lo... [14:24] orogor: I am not familiar with that application, but depending on the size of the volume it can take quite a bit of time. [14:29] stgraber: i dunno. [14:29] stgraber: when ireboot again i'll look into it, but i can't reboot the thing now [14:36] jdstrand, we had discussed maas using isc-dhcp-server and needing to modify app armour. [14:37] woudl you be opposed to me adding a "#include /etc/maas/apparmor/usr.sbin.isc-dhcp-server" (or equivalent 'correct' path) [14:38] the reason i'd prefer this over explicitly writing a stanza in isc-dhcp-server's profile is that we can more easiliy then bring that back to precise. [14:38] our plan is on precise to just add '#include that-file-path' to the /etc/apparmor.d/local/usr.sbin.dhcpd [14:39] (realizing that we're not supposed to modify other packages config files, but trying to do so safely and considering it the least invasive path) [14:39] in jdstrand's absense or non-immediate reply perhaps mdeslaur could comment if this would be remotely acceptable. [14:40] (but i'm willing to just be patient too) [14:44] jdstrand: ^ [14:48] smoser: if jdstrand doesn't answer you, sbeattie may be able to when he gets in in an hour or so [14:48] k. [14:49] thanks === n0ts is now known as n0ts_off [14:53] smoser: so, the intention is fine, but it won't work if that file doesn't exist. if you modify the isc-dhcp to create the /etc/apparmor.d/isc-dhcpd.d directory (or similarly named) then '#include ', then you can just drop a file into /etc/apparmor.d/isc-dhcpd.d and it will all work the way you want [14:53] smoser: this is because the isc-dhcp package creates the directory you are including, so it is guaranteed to be there [14:54] smoser: and you can just drop a 'maas' file in there [14:54] and you're ok with that? [14:54] sure [14:54] smoser: the default install won't have the maas file so the profile is not actually changed [14:54] right. [14:54] smoser: so in that way, it is even better than editing the profile directly [14:55] so in precise, then, we will write '#include ' into the local file [14:55] and create the directory [14:55] it would be nice to clean out the euca bits to use this then, but obviously that doesn't need to happen right now [14:55] on installation of maas [14:55] smoser: I would recommend in precise doing an sru for isc-dhcp [14:55] jdstrand, ok. we can do that too. [14:56] just more painful, but you are correct more "correct" [14:56] smoser: it is effectively a no-op for the profile, and reduces the deltas [14:56] thanks. [14:56] np [15:01] New bug: #1049167 in antlr3 (universe) "package antlr3 3.2.is.3.2-7 failed to install/upgrade: trying to overwrite '/usr/share/java/antlr.jar', which is also in package libantlr-java 2.7.7+dfsg-4build1" [Undecided,New] https://launchpad.net/bugs/1049167 [15:10] Anyone here knows deb packages? I have a single executable file that I want placed in /usr/local/bin. What is my path to least resistance here? [15:21] jdstrand, http://paste.ubuntu.com/1198843/ does htat look remotely correct? [15:22] i used usr.sbin.dhcpd.d rather than 'isc-dhcp.d' just as that seemed like more consistent naming. [15:22] (so i'm asking for input there) [15:23] smoser: for bug 1049177 can you just attach a patch or commit to ubuntu:isc-dhcp but without uploading? I'm still planning on doing some isc-dhcp work today/tomorrow so unless it's urgent, I'd prefer to limit the number of uploads [15:23] Launchpad bug 1049177 in isc-dhcp "isc-dhcp-server apparmor profile should have include ".d" " [Medium,In progress] https://launchpad.net/bugs/1049177 [15:24] stgraber, sure. is a mp ok? [15:25] smoser: sure [15:26] xnox, just finished extending [15:27] orogor: well done =) [15:29] Anyone know the right way to set the dns timeout via /etc/network/interfaces ? [15:34] dax_roc: I'd suggest adding the options line to /etc/resolvconf/resolv.conf.d/head or /etc/resolvconf/resolv.conf.d/tail [15:34] stgraber: just found it thanks. appreciated. [15:37] stgraber: is that the same for 11.04 ? [15:37] no [15:37] resolvconf was only introduced in 12.04 [15:37] before that, /etc/resolv.conf was a fairly standard file you could edit directly (unless you're using dhclient or some other tool that modifies it) [15:38] No, I'll add it directly, Thought network interface scripts might overwrite it [15:39] *upgrading soon anyway, time permiting. [15:46] stgraber: ha, I had actually found your article, thanks agian Stéphane :D [15:46] np :) [15:46] stgraber: but the resolvconf package was available pre-12.04. And for example I was using resolvconf package for a long time =) [15:46] xnox: yeah but we don't support nor recommend it before 12.04 as it was quite broken on Ubuntu [15:47] stgraber: worked for me. I did take time to configure it though. But the days are gone when I was using highly weird university networks. [15:49] stgraber: oh no, i only just noticed that when you use the ecryptfs backed containers, 'ecryptfs_root' shows up as though it's a container. i need to change the target mount dir. [15:49] xnox: yeah, and you probably had a system where /etc was read/write and /usr wasn't a separate partition [15:51] =( it was a macbook as well [15:59] is there any way to force the auto update process to send an e-mail (basically I just want to test it's email functionality - set up heirloom (prev. nail)) [16:00] went for heirloom as it's just 1 package opposed to bsd-mail+ssmtp for example. Seems simple enough - mailx sends stuff out nicely [16:02] apt-get install mailx ? [16:04] seems the same [16:09] mailx is a virtual package, but how does that force apt's auto update process to send a test mail? [16:24] stgraber: I've added options rotate, options timeout:1 to the base file but it's not rotating ? any suggestions. Any way I can see the full resolv.conf ? [16:25] dax_roc: did these options get into your /run/resolvconf/resolv.conf? [16:26] I don't have one [16:27] stgraber: just interface, postponed-update in /run/resolvconf === hazmat is now known as kapilt [16:32] roaksoax: you'll work with zul and adam_g for "[andreserl] Track GlanceAPI, Glance Registry, Keystone RA's, and ensure they hit the archive.: TODO" ? [16:35] Daviey: yeah, well debian maintainers has made a debian packaging available on github but since I haven't been able to catch up with him I don't yet know if his plans are to upload that to debian or what [16:35] Daviey: cause initial he said that those resource agents would be part of upstream (openstack) [16:36] initially* [16:41] New bug: #917824 in qemu-kvm "qemu loops/hangs on extending qcow2-diskspace" [High,New] https://launchpad.net/bugs/917824 [16:44] hey i added "0 0 * * * /sbin/shutdown -h now" to my crontab, but do i need to update the sudoers file? ...to what? or should that be "/sbin/poweroff" instead? help? [16:45] rtdos: Which user runs that command? [16:48] rtdos: sudoers won't enter into it. You can supply a password to a cronjob and it's not a good idea to let a user use shutdown without a password. If you're going to use that cron at all, it should be root's crontab. Root, of course, doesn't need permissions from sudo to power down the computer. [16:48] *can't supply a password to a cronjob. [16:49] e_t_: you can configure sudo to let a user use certain commands without a password [16:49] e_t_: or you can run the cronjob as root [16:50] RoyK: that's almost exactly what I said: "it's not a good idea to let a user use shutdown without a password" and "If you're going to use that cron at all, it should be root's crontab" [16:51] e_t_: if it's a reboot, with a controlled script, it's ok for a certain user if that user only runs cronjobs [16:52] e_t_: and if the command allowed is only the reboot thing [16:52] e_t_: allowing a user full sudo access, even if it's a cron user, isn't very wise [16:53] Does anyone here, have any good guides on setting up "permissions" espically nfsv4 permissions for Ubuntu [16:53] This is *baffling* me, how your supposed to implement access control in linux as is [16:54] Daviey: can you review websockify for me? [17:01] New bug: #1049240 in vsftpd (main) "nologin shells don't allow sftp access" [Undecided,New] https://launchpad.net/bugs/1049240 === mcclurmc is now known as mcclurmc_away [17:10] zul: NEW review? [17:10] Daviey: dah === matsubara is now known as matsubara-lunch [17:27] hi, can some of you recommend a simple hosting panel, just to manage apache, ftp, mysql? [17:28] I was just told I need to install ubuntu-server here at work. this OS worth it?? [17:28] grefter: depends on what you need it 4? [17:29] monitoring software [17:29] thats it [17:29] zabbix ? [17:29] opennms [17:30] or something else? [17:31] you can use ubuntu for that [17:31] using it to monitor our datacenter servers and controllers [17:31] alright, sounds good [17:31] you can use ubuntu for all your linux/unix based software [17:31] lol [17:31] easy to administer [17:31] in comes the sales pitch ;) [17:32] ohh no.. I am not a salesman [17:32] heh [17:32] like debian is I hear? [17:32] acctually Im technitian as you [17:34] putting it on an r520 .. debian missed the beat [17:34] centos purss like a gurkin, but they won't let me use it, cause opennms was build around .deb files...bloody developers :) [17:36] then just use ubuntu [17:36] I've just upgraded 10 of our webservers lastweekend to latest relase and some of the servers were running 8.04 without any problems [17:38] my point is good documentation === zz_Kiall is now known as Kiall [17:59] Anyone have mollify setup with sqlite? I am apparently doing it wrong [18:06] well, guess we'll never know until we try it out :D [18:10] how do i use root's cron job? [18:10] rtdos: do you mean, how do you make a cron job that runs as root? [18:12] yes. what i am trying to do is power off the system unconditionally. [18:14] pull the cord ? [18:15] use ipmi :) [18:23] jamespage: stgraber: I'm torn. right now the lxc testsuite is one big py-unittest. To convert it to utah, should i be splitting it up with one test per dir? [18:24] it'll mean everyone will *have* to use utah to neatly run the tests... [18:24] whereas right now they can just branch my tree and run it all [18:25] ipmi ? [18:26] hallyn: what's the URL of your test branch again? want to check if there's no easy way to get something roughly at the middle where we can still easily run the tests without utah but can also have utah call the individual tests [18:34] hallyn, I'd just stick it in a one test [18:42] stgraber: lp:~serge-hallyn/+junk/lxc-test [18:42] I'm going to put it all in one test under lp:~serge-hallyn/+junk/utah-lxc for now [18:42] jamespage: ok, thanks. will do [18:43] hallyn, I think we should ask that utah know how to deal with tests which are actually test suites/multiple tests [18:43] ah, actually it let me use lp:~serge-hallyn/utah/utah-lxc [18:43] jamespage: +1 [18:44] hallyn: ok, I think you may eventually interested by refactoring the tests to look like what we're doing in ubiquity and a few other package. Basically a bunch of .py files each containing a unittest.TestCase class and using a "run" script that calls them all but also allows you to only run a subset [18:47] hallyn: lp:~ubuntu-qa-website/ubuntu-qa-website/python-qatracker is I believe a reasonable example of this (look at the tests directory) [18:47] stgraber: that's similar to what the utah structure woudl be [18:47] ok, thanks. that would probably be nicer long term [18:52] not a good start for ubuntu, get to the install screen, hit enter to install ubuntu-server, get a black screen.. [18:52] fuckin server takes off like a rocket and then nothing happens :D [18:53] !language [18:53] Please watch your language and topic to help keep this channel family-friendly, polite, and professional. [18:53] i hear families swear all the time.. [18:54] unless they are mormons [18:54] grefter: just don't do it in here please [18:54] heh [18:54] regardless, ubuntu doesn't work on this server either it appears. [18:55] dell r520, anyone got one? [18:55] grefter: can you switch virtual terminals? [18:56] New bug: #1049309 in keystone (main) "keystone package: logrotate should have compress option" [Undecided,New] https://launchpad.net/bugs/1049309 === mchamplain_alt is now known as mchamplain [18:58] guntbert: no, just a black screen, monitor appears to be in hibernation [18:59] grefter: sorry, I have to ask: do you know how to switch VTs? [18:59] if I reboot the server, i get a display, only drops when I try to install [19:00] guntbert: yes I know, that doesn't appear to be the issue [19:00] grefter: another idea: did you check if the iso was ok before burning it? === troy_ is now known as Troy^ [19:02] no, actually I didn't. does that commonly happen with ubuntu? [19:03] heh? [19:03] grefter: not at all, but a bad d/l makes a bad CD :-) [19:03] that has nothing to do with ubuntu, but your internet connection when you downloaded :) [19:03] !md5sum | grefter [19:03] grefter: To verify your Ubuntu ISO image (or other files for which an MD5 checksum is provided), see http://help.ubuntu.com/community/HowToMD5SUM or http://www.linuxquestions.org/linux/answers/LQ_ISO/Checking_the_md5sum_in_Windows [19:08] appers to be the server, have a second one and it works fine. [19:09] nothing like an 8k paper weight [19:10] dell? [19:10] yer [19:10] my friend has the strangest issue with ubuntu cd's and dell cdrom's [19:11] try using the alt iso image [19:11] or you could attempt a mini-iso/net boot [19:11] yea, good idea [19:11] for some reason, his dell cdroms refuse to boot ubuntu :( [19:11] but it's weird, both server are the same..but the disk works in 1 not the other :D [19:12] I dunno if I have that issue with my dells, as I always netboot install [19:13] i'll try a net on the 2nd box [19:13] patdk-wk: thats fine after you get them to wake-on-lan without windows [19:14] heh? [19:14] why would I bother with wake-on-lan with a server? [19:14] but I have a crapload of dells at home, they all wake-on-lan just fine [19:15] go in bios, turn on wake-on-lan, disable ultra-low-power-savings, done [19:17] patdk-wk: (we are off topic here, so..): I guess you are right with a dell*server*, the consumer machines behave somewhat different it seems [19:18] actually, I was talking about dell workstation machines that I do that with, sx270, sx280, gx620, optiplex 745, optiplex 755, optiplex 780 optiplex 790, optiplex 960 [19:19] all of those wake-on-lan the same exact way, works perfect with ubuntu server [19:19] I have used most all of those as mythtv frontend machines using wake-on-lan [19:20] patdk-wk: glad to hear it - my experience was different - but lets drop that here :=) [19:25] New bug: #1049321 in irqbalance (main) "package irqbalance 1.0.3-1ubuntu2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/1049321 === Ursinha_ is now known as Ursinha === cpg|away is now known as cpg === matsubara-lunch is now known as matsubara [20:14] mdeslaur, smoser: does that look right to you? http://paste.ubuntu.com/1199407/ [20:15] mdeslaur, smoser: the branch history is in a pretty bad state so I had to go through debdiffs and as the introduction of that code was pre-natty there was no good way of extracing a clean diff... [20:16] stgraber: you're missing the argument I believe [20:16] mdeslaur is right. wait_for_root takes an argument. [20:16] oops [20:17] hmm, you sure? last version I see with wait_for_rw doesn't pass an argument [20:18] * stgraber finds a 12.04 machine, will be much easier this way ;) [20:18] indeed it does and my code is wrong too, guess I used an older version [20:18] let me update that [20:20] mdeslaur, smoser: that should be better: http://paste.ubuntu.com/1199418/ [20:23] stgraber, http://paste.ubuntu.com/1199428/ is the content pre merge [20:23] key things there are 'readlink -f' [20:23] to get the correct directory of the file if its a link [20:24] stgraber: looks like you're using new_resolv_conf before you assign it [20:24] mdeslaur: indeed... fixed [20:26] smoser: right, so that's something like this then: http://paste.ubuntu.com/1199437/ [20:27] * stgraber really wishes we'd just consider non-resolvconf setups as unsupported and stop having to workaround these cases... [20:28] stgraber, i'm surprised... but i was actually coming to that conclusion the other day. [20:28] or rather this morning [20:28] as i was thinking with how to correctly update /etc/resolv.conf [20:28] stgraber: the rm -f $new_resolv_conf will fail if the partition isn't r/w yet [20:29] and was inventing some scheme that would allow multiple parties to collaborate in its maintenance [20:29] and then realized that we had such a thing, and it was already the default [20:30] :) [20:30] mdeslaur: indeed... moving the call one line up to fix that one ;) [20:30] jamespage: there is something i'm missing... when i do sudo utah -r ./runlists/print-server.run i get a failure [20:30] i gather i can't use the .run file as a runfile? [20:31] oh it's probably prety late for jamespage [20:34] stgraber: you're also going to need to use ${resolv_conf} instead of "/etc/resolv.conf" everywhere lower down do you replace the symlink target instead of the symlink [20:38] mdeslaur: yep, noticed that when going through the script again... [20:38] mdeslaur: http://paste.ubuntu.com/1199462/ anything I missed? [20:40] I'll make sure this one gets forwarded to Debian because it's going to be a pain to merge [20:40] stgraber: looks good to me === gary_poster is now known as gary_poster|away [20:46] nice [20:46] got it going :) [20:46] this will save me a few bucks [20:46] ahhh wrogn chan === cpg is now known as cpg|away [20:55] stgraber, if you have not committed yet, could you put a comment on the ': >>' line ? [20:56] referencing bug 857524 [20:56] Launchpad bug 857524 in isc-dhcp "dhclient-script will not work with read-only /etc" [Undecided,Fix released] https://launchpad.net/bugs/857524 [20:56] just because at one point that was 'test -w' but that was insufficient in reality [20:57] smoser: already commited and forwarded to Debian. I'll send a comment to the Debian bug though. [20:58] wait. [20:58] but htat bug is bad. [20:58] stgraber: is there a way to increase the time for dnsmasq to wait for an answer from the NS? [20:58] just a minute [20:58] smoser: ok [20:58] bug 856984 is right [20:58] Launchpad bug 856984 in isc-dhcp "dhclient-script attempts to write /etc/resolv.conf before it is writable" [Undecided,Fix released] https://launchpad.net/bugs/856984 [20:59] smoser: thanks, commented in the Debian bug [21:06] what would cause the dns-nameservers directive not to be loaded when I perform a networking restart? I've configured my IP statically and am attempting to assign DNS servers, but when I restart networking, my resolv.conf is over written with my nameservers as 127.0.0.1, however, my dns-search directive is being read correctly. [21:07] honestly, I could just remove the symlink, add my name servers, then make /etc/resolv.conf immutable, but that is just the wrong way to fix the problem. [21:11] New bug: #1048677 in horizon "UncompressableFileError: 'horizon/js/horizon.js' isn't accesible via COMPRESS_URL" [Undecided,New] https://launchpad.net/bugs/1048677 [21:12] so... can I just purge resolvconf? `apt-get purge resolvconf` so it doesn't overwrite my "custom" /etc/resolv.conf (since I removed the link then created an immutable file /etc/reslov.conf), and frankly the errors "resolvconf: Error: /etc/resolv.conf isn't a symlink, not doing anything." are simply annoying since my system seems to have it out for me today. === cpg|away is now known as cpg [21:13] crap, when I try to remove resolvconf it wants to remove ubuntu-minimal... ubuntu-minimal is a meta package right? === kapilt is now known as hazmat === tblamer_ is now known as tblamer === cpg is now known as cpg|away [22:28] zul: when is cinder supposed to initialize its sqlite db? its doing it in both cinder-scheduler and cinder-common postinsts === leojrfs is now known as leonas [22:41] cinder-common [22:42] adam_g: i could have sworn i put it in there [22:48] zul: it is there [22:48] zul: but its also in cinder-scheduler [22:48] ill remove it [22:48] (from scheduler) [22:48] yeah i didnt put it in cinder-scheduler [22:49] zul: either way, neither of them actually work [22:50] adam_g: bah === cpg|away is now known as cpg [23:35] what is initrd preseeding