[00:01] Sigh. So if you have a 1GB server and unattended upgrades, and ubuntu kernels weigh even more than Debian ones (128MB each), you use up a quarter of your disk in a month :-/ === sixstringsg|noms is now known as sixstringsg === sixstringsg is now known as sixstringsg|afk [01:23] New bug: #906654 in nova (main) "nova should depend on openstackx and load appropriate osapi_extensions by default" [Undecided,New] https://launchpad.net/bugs/906654 [01:25] adam_g: you are evil [01:28] adam_g: while you are still here can you have a look at http://paste.ubuntu.com/775687/ [01:31] zul: one sec [01:39] zul: its been so long since ive messed with that stuff. isn't the userspace daemon required by the initiator utliities? [01:41] no [01:41] i dont think so [01:41] ah, i remember now.. iscsid is only used to establish preconfigured sessions [01:44] zul: that diff looks good as long as its been tested. im curious to know if the tgt off-by-one error you guys were looking at was caused by a split brain between nova and iscsiadm? [01:45] adam_g: yeah basically iscsadm wanting one thing while tgtadm wanting another [01:45] but yeah that patch has been tested [01:46] zul: has it been tested after removing iscsid? what i meant was, iscsid restablishing a connection, bumping the lun index before nova had chance [01:47] adam_g: ill test tomorrow first thing its getting late here [01:48] cool === bladernr_ is now known as bladernr_afk [04:53] Hi [04:53] is there a way to know list of modules loaded or enabled in apache2 using CLI [04:54] I can see it in /etc/apache2/mods-enabled/ [04:54] Server version: Apache/2.2.14 (Ubuntu) [04:55] on 10.04 LTS === kermit is now known as Guest53940 [05:11] kaushal: that's essentially it. or "apache2ctl -M" [05:12] shauno: thanks [05:13] shauno: what is static vs shared module ? [05:14] I believe static are built-in at compile-time, shared are loadable modules you'll find in mods-available / mods-enabled [05:18] oh ok [05:18] Thanks shauno [06:16] New bug: #906735 in puppet (main) "package installation should create /var/log/puppet" [Undecided,New] https://launchpad.net/bugs/906735 === sixstringsg|afk is now known as sixstringsg|away [08:39] anyone know how to troubleshoot a failing domain resolution? [08:39] $ host btcnearme.com [08:39] Host btcnearme.com not found: 1(FORMERR) [09:04] hi all [10:57] where would i look in my logs to see if someone was messing with my server? or trying to gain access to it [11:01] philipballew: /var/log/auth.log, in general many automated bots will be trying to gain access [11:02] fail2ban will ban them after a few trys if you have that installed. [11:02] alright, im looking now. Yeah. I might wanna change ssh from 22 sometime [11:02] philipballew: changing the port won't really add much security, disabling password based ssh and using key based instead and using fail2ban will provide much better security than changing the prot [11:03] alright. I need to just set up rsa keys again. [11:03] that isnt hard at all [11:03] indeed [11:04] Thank you for the help [11:23] hi [11:24] i'm trying to install server 10.04 from a usb stick [11:25] but it breaks because "can't mount cd drive" [11:25] what can i do? [11:29] angelete2: is this breaking in the BIOS or later? if in the bios is the boot order set to boot from USB before CD? [11:29] BIOS boot order is only from USB [11:29] it has no cd drive [11:30] installation begins, but it tries to mount cd, so it fails [11:30] angelete2, I've experience in the past that some USB sticks simply aren't bootable [11:30] i've installed on same server with same usbs stick ubuntu 11.04 [11:30] but i need to downgrade it [11:41] angelete2: down grade what ? [11:42] ubuntu server [11:55] angelete2, try installing from a cd === kermit is now known as Guest19518 === jonludlam_afk is now known as jonludlam_ [13:25] New bug: #906878 in tftp-hpa (main) "package tftpd-hpa 5.0-21ubuntu2 failed to install/upgrade: subprocess installed post-removal script returned error exit status 127" [Undecided,New] https://launchpad.net/bugs/906878 [13:52] good morning === bladernr_afk is now known as bladernr_ === dendro-afk is now known as dendrobates [14:57] hi guys, i'm trying to use ovh failover ip, actually putting ip in browser i see the sites on the server, i need to route that ip to a vm, how can i do that? [14:58] should i add before an interface alias? [15:23] hallyn: around? [15:24] zul: yup [15:24] (but not for long) [15:24] hallyn: is it cool to use lxc-attach with libvirt if you want to attach a block device to a lxc container [15:25] zul: I haven't looked at the blockdev support in libvirt-lxc at all [15:25] zul: but i shoudl think so [15:25] im thinking about getting euca-attach-volume to work [15:25] can i decrypt an lvm in a script? i want to put a drive in, have the UUID trigger a script that will load the vg, load the lvm, decrypt the lvm, mount, and start a program [15:25] hallyn: the attach device stuff is non-existant in libvirt-lxc [15:26] well you could always use qemu-nbd and copy the device nodes for it into the container [15:27] (and add a cgroup whitelist entry) [15:28] hallyn: ooh....thats a good idea [15:28] for persistent libvirt vms that'd be harder, but for one-time euca-attach... [15:31] hallyn: yeah [15:31] thanks! [15:32] np, cool, can't wait to see it now :) [15:33] dkn: you could probably do that with udev rules [15:33] ya? [15:34] decrypt the lvm in what way? [15:34] is it dmcrypted? [15:37] i don't recall the exact method, i actually cheat and use the GUI disk utility and just press "unlock volume" since it remembers the passphrase, then the device shows up in /dev/mapper/HERE === dendrobates is now known as dendro-afk [15:41] basically, right now i 1.vgimport -a 2.lvchange -a y /dev/mapper/HERE 3.unlock 4. mount /dev/mapper/udisks-luks-uuid-yada /mnt/HERE [15:42] once i'm done with the drive ~1hr i do the reverse to take it out and put in the next one, i'd just like to have some scripts to do it, idealy they'd run when i pop the drive in === dendro-afk is now known as dendrobates [15:49] like SpamapS said, you can do that from a udev script. Take a look under /lib/udev/rules.d and /etc/udev/rules.d for examples [15:51] :) thanks === utlemming_afk is now known as utlemming === sixstringsg|away is now known as sixstringsg [16:06] Hi, semi-noob here. I' seeing a lot info about Ubuntu thin clients booting trough PXE, but what I aiming for is to boot a Ubuntu server image from a remote pre configured image in a way similar to a preseed installation. [16:07] any way to do this [16:18] PedroGomes: works via pxe as well, think help.ubuntu.com contains some pages on it [16:20] TeTeT: can have, for example, a installed and configured image that as software X, Y and Z and you can boot it ate the same time on machine 1,2,3 and 4 ? [16:20] *can I have [16:24] PedroGomes: there's a suite of tools in Ubuntu 11.04 and 11.10 called "Ubuntu Orchestra" that help setting up netboot automated installations [16:25] PedroGomes: It centers a lot around 'cobbler', which has power control and pxe/dhcp configuration built in. [16:25] SpamapS: already have foreman doing that… [16:26] SpamapS: looking for a direct network boot, not a installation [16:31] btw, can someone find the link to download the tftp config files for Ubuntu, my google skills are not helping me.... [16:31] Hi folks, I have a litle question regarding vmbuilder - is this the right place to ask? [16:37] PedroGomes: OH [16:37] PedroGomes: you may want to look at LTSP [16:37] PedroGomes: https://help.ubuntu.com/community/UbuntuLTSP [16:42] SpamapS: mhh thanks, but isn't LTSP for thin clients only?, making every client depend on a single server. I'm looking for something where I can download the installed image from the server and boot it like it was a installed natively. [16:44] anyone want to do a quick merge proposal? [16:44] https://code.launchpad.net/~smoser/ubuntu/precise/rsyslog/lp905419/+merge/86429 [16:44] SpamapS: truth is that I feel like chasing a ghost, I remember seeing a similar systems years ago "Rembo", but the license fees… [16:45] PedroGomes: that sounds a lot like how cloud computing works [16:46] smoser: I'll take a look [16:46] SpamapS: yep I have a OpenStack installation also, but I was looking for a bare metal solution [16:51] smoser: +1'd [16:52] anyone here ever built an SFTP server on 10.10? I'm rather new to ubuntu server and need some documentation [17:03] PedroGomes: openstack is looking at integrating cobbler as a bare metal option [17:03] PedroGomes: though its more for things like ARM that have no good virtualization support [17:04] SpamapS: guided install and puppet was enough for me, but people are pressing me to find a way to deploy installed images [17:05] SpamapS: truth is, that systems existed, I just cant find something actual or open source [17:06] PedroGomes: whats so bad about sticking kvm between the metal and the installed image? [17:08] PedroGomes, sorry to jump in, but there is clonezilla and partition image that are (for lack of a better term) "ghost clones" [17:08] SpamapS: some people want the system to deploy images where they can test Xen appliances and such [17:09] SpamapS: its a cluster for distributed system experiments, some people want Vms other want bare metal…. [17:09] there is nothing in cobbler that will do "p2p". those solutions may be a faster than cobbler preseed installs, but cobbler could do what you want. [17:09] ie, you could create preseed files for each of your "images", then when you install that fully automated preseed the end result is the same. [17:12] smoser: I have that, not trough cobbler but foreman, but I have people that want more… thanks anyway [17:13] PedroGomes: so a pre-seeded install takes, what, 10 minutes at worst? I bet a pre-installed image laying down on disk will take 5. [17:14] PedroGomes: for test iteration, the openstack QA guys did something clever with LVM snapshots and kexec() that is pretty cool... you might want to talk to them. [17:14] preinstalled image you can multicast serve though ;) === jhobbs_ is now known as jhobbs [17:15] SpamapS: the question is not time, its the personal configuration that you can achieve easily with personal images, [17:15] SpamapS: ok I will ask there [17:15] jhobbs: ah? [17:18] PedroGomes: um, you can get that "personal" configuration easily with an install too. [17:19] PedroGomes: you may not have noticed, but cloud-init works with local seed data now. :) [17:19] PedroGomes: or you're saying they want golden-image support. [17:19] PedroGomes: another option is system-imager/sisuite [17:19] PedroGomes: though I don't know how well maintained that is. [17:19] SpamapS: som ething like that [17:20] SpamapS: another thing to check then [17:23] zul: thanks for driving the meeting, that was perfect. [17:23] Daviey: no worries [17:23] PedroGomes: I've used sisuite in the past, and it was really quite good actually. [17:23] We had a meeting? cripes. [17:24] * SpamapS wonders if its worth requesting it be moved back 1 hour for DST :-P [17:24] SpamapS: yeah, it's that weekly thing we do; whilst you slack off. :) [17:24] Daviey: I cherish every minute of slack off time you all provide me, thx. :) [17:25] heh [17:25] SpamapS: I wouldn't object to bumping it forward 1hr, during DST. [17:25] that creates a conflict for me [17:26] zul: conflict you say? I think we need to turn the conflict into a death match with SpamapS. [17:27] obviously i would win [17:27] winner gets the time of their choice. [17:27] loser gets to rest. [17:28] adam_g: How are the charms looking for openstack? [17:28] I'm fine with it being at 0800 PDT .. I just never seem to be at my keyboard at that time, so I miss the pop-up reminder and then get distracted otherwise. :p [17:29] Daviey: keystones working great, still cracking on horizon [17:33] bug 893926 [17:33] Launchpad bug 893926 in eucalyptus "Contains traces of UEC" [High,Triaged] https://launchpad.net/bugs/893926 [17:33] Is here someone that can answer my question in the bug comment? [17:34] Daviey: FYI bug #905286 is currently making things a bit awkward to get installed via orchestra without manually work [17:34] Launchpad bug 905286 in juju "orchestra: dict has no attribute read" [Critical,New] https://launchpad.net/bugs/905286 [17:37] mainerror: hey [17:37] adam_g: Good to hear, have you dicussed the issue with the juju team? [17:38] ah, it's a cobbler bug? [17:38] Daviey: yes [17:39] Hello Daviey. [17:39] mainerror: basically the packages are "branded" as "Ubuntu Enterprise Cloud" or "UEC" .. and they should just be called "Eucalyptus" now. [17:39] mainerror: replied [17:40] Awesome, thanks. [17:40] That should be an easy fix. [17:40] mainerror: Something you fancy driving? === nkv_ is now known as nkv [17:41] hoi all === kermit is now known as Guest26514 [17:41] Daviey: I'm sorry? [17:42] mainerror: If it's something you want to work on.. There is a uec branding patch which can be dropped (should remove the css)... and there are other goodies in debian/* which can be dropped. If you want any pointers, please feel free to ask! [17:44] Well I have already fetched what is in the repository and greped for "UEC" and "Ubuntu Enterprise Cloud" so find out which files I should look at but if you have more accurate information for me it would be great. :) [17:44] SpamapS: have you seen 905286? [17:45] mainerror: Have you used quilt before? [17:46] mainerror: If you have used quilt and bzr before, if you wanted to propose a branch with a first scratch - we can walk through what is left to go? [17:47] I have used bzr before, although not to fix a bug- I have not heard of quilt yet. [17:52] mainerror: Maybe this is of help, http://developer.ubuntu.com/packaging/html/patches-to-packages.html [17:52] mainerror: quilt and bzr make for really messy changesets, but ho hum. :) [17:52] mainerror: If you do, /something/ - we'll work through it to get it sorted :) [17:54] Right. I'll read through that and start fixing the files up then I'll get back to you. :) [17:55] Should I mark the bug as "In Progress" and assign it to me? [17:58] mainerror: sounds great :) [17:59] mainerror: in debian/* there are some files which refer to avahi, they can also go. :) [18:14] Daviey: There are source files that contain "uec" in their name, like "uec_component_listener.c" for example. Do we want to change the file names too? === cjwatson_ is now known as cjwatson [18:34] Daviey: yeah I thought fwereade_ was workign on it [18:36] SpamapS, Daviey: I've only had a chance to give it a short look I'm afraid, I've been trying to get the groundwork for the machine constraints feature down [18:37] I suspect a trivial workaround might be in order. :P [18:37] SpamapS, Daviey: and I don't really know anything yet: *something* in between CobblerClient and cobbler itself is doing something different with xmlrpc [18:38] SpamapS, I think it might be a little trickier... if the format we get nested dicts in has changed, it seems reasonable to suspect that the send format has changed too [18:38] SpamapS, which then means we need to be able to characterise the critical difference between environments it shows up in and ones it didn't [18:39] SpamapS, I can imagine a dirty workaround, but I'm not sure I could get one merged :/ === kermit is now known as Guest23736 [19:01] how do I update repos from shell? [19:02] nvm apt-get update did it === sixstringsg is now known as six|afk === bitmonk_ is now known as bitmonk === Guest23736 is now known as kermit === bitmonk_ is now known as bitmonk [20:31] Daviey: There is quite a lot of avahi stuff going on in debian/* === dendrobates is now known as dendro-afk [21:27] I am having a problem with a PHP project which connects to mysql database, the site gives me error that the database is down. While it worked flawlessly on windows machine running xamp. Any ideas? === dendro-afk is now known as dendrobates [21:27] Is the database down? [21:28] However normal PHP files, which doesn't use mysql databases work fine. [21:29] mysql-server is running. And I created the database through phpmyadmin. When I tried using mysql-admin -h host_ip I got an error that mysql couldn't read hostname. [21:30] mustafa: where is the mysql-server instance? on a windows machine or a linux machine? [21:30] ... [21:30] i know your current issue [21:31] Linux, Ubuntu. [21:31] EvilResistance: No, it was already setup. I did change the bind-address to localhost. I'll check the skip-networking now. [21:31] if its only binding to localhost it'll only accept connections from itself [21:31] On the windows machine it was already setup. But I'm trying now to configure it on my Ubuntu machine. [21:32] mustafa: you need to make sure skip-networking is commented out or removed in the conf file [21:32] Uh? Did I miss that line somehow? [21:32] mainerror: he's violating the guidelines... [21:32] getting help in #Ubuntu and here [21:32] i'm moving my support for him to here ;P [21:32] Oh! [21:33] EvilResistance: there is a comment "Instead of skip-networking the default is now to listen only on localhost" [21:33] then tell it to not listen on localhost [21:33] rather listen on your IP [21:34] whatever your system's IP might be [21:34] EvilResistance: no uncommented #skip-networking [21:34] mustafa: read my last 3 messages [21:34] EvilResistance: which is 127.0.0.1 right? [21:34] * EvilResistance facedesks [21:34] mustafa: no [21:34] mustafa: 127.0.0.1 is the *localhost* [21:34] if you're on a router, it'd be something like 192.168.1.1 or something [21:35] or if its directly internet facing, the public IP (which you should know if its internet facing) [21:35] EvilResistance: Ah 192.168.1.20 [21:35] use that for now then [21:35] note that you might run into some occasional breakage if you eventually want this to be useable on the internet [21:36] EvilResistance: OK, I have changed the bind-address. Then restarted mysql and apache and still the error is there. [21:38] I installed the apache2, php5, mysql-cleint then mysql-server and finally the phpmyadmin. All using apt-get. I'll really appreciate your help, I have a discussion tomorrow at college on this project using my laptop. [21:43] EvilResistance: Are you still there? [21:44] nope, kinda busy atm [21:44] dealing with some abuse@ emails for my group [21:44] Take your time, I'll be waitting :) [21:50] mustafa: so if you try to access mysql from the command line on the box which php is running on, can you? === six|afk is now known as sixstringsg [21:56] drt24: How can I know? [21:58] run the "mysql" command with the relevant options and see if it works, if not change things until it does, then the configuration specified by the opitions to mysql should also work from php using the relevant php function [22:01] drt24: I am little bit lost. The server is running, but when I use mysql-admin -h host_ip I get that error "Could not connect to host '192.168.1.20'. [22:01] MySQL Error Nr. 1130 [22:01] And that the host isn't allowed to connect to mysql-server [22:02] so firewalls or similar? [22:02] Not that I know of, I'm running Ubuntu 11.10 Desktop edition. [22:02] or the mysql-server is not listening correctly? [22:03] sorry, going to sleep now. [22:04] OK, if you could wait little more 5 min. I'll appreciate it I a project discussion tomorrow at college and I have been searching and asking for 6 hours now with no luck. [22:04] If not I'll understand.. [22:07] mustafa: so you have 2 pc's? [22:08] or is everything on one pc? [22:12] alex_joni: One PC. I tried it earlier today on a friends PC. Just moved the site files to the apropriate folder and imported the database from phpmyadmin [22:14] alex_joni: I am using a custom path for enabled sites if that helps. === EvilResistance is now known as TheEvilPhoenix [22:52] in /etc/mysqlmy.cnf bind-address = 0.0.0.0 would have solved it [22:52] [22:53] true, but that would have bound it to * basically, no? === bitmonk_ is now known as bitmonk [22:56] well as far as i understand this enables the server accept connection from any external host === nande is now known as nandemonai === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates === TheEvilPhoenix is now known as EvilResistance [23:48] hello [23:48] i use ubuntu server 10.04 and i have route configured on ip command, how to save this and restore on boot? [23:49] i havent in 10.04 ip route save/restore === guampa|2 is now known as guampa