[02:35] <hallyn_> SpamapS: is there a good reason not to have the host mount the iscsi target, and then bind-mount that into the contaienr?
[03:21] <Guest39238> samba question can I use ".nogroup" as a group with users in it or should I make a new group then add users to it?
[03:31] <ubuntutis> anybody know what programming languages i should use to make a chatbox
[03:56] <raub> In 10.04, dovecot in ubuntu had two separate config files, /etc/dovecot/auth.d/01-dovecot-postfix.auth and /etc/dovecot/conf.d/01-dovecot-postfix.conf
[03:57] <raub> But now in 12.04 it has only the conf.d directory with a ton of files in it. How do they relate?
[06:28] <bombev> hello
[06:28] <bombev> I am running Ubuntu Linux 12.04.1 is there some manual to show me how to configure two ISP and to switch it via ssh?
[06:35] <warzauwynn> bombev: can you be more specific about your setup?  how are these ISPs connected to your ubuntu machine?  how do you intend these two networks to work together?
[07:34] <lpapp> hi, is there anyone using wsgi and ubuntu here? my wsgi file does not seem to load through apache (mod-wsgi). I see the raw content rather than getting it interpreted through the wsgi protocol. This is a ubuntu 10.04 server, and almost the same vhost config file works on another Ubuntu server which is version 12.04.
[07:37] <lpapp> here you can find my vhost site config file, http://paste.kde.org/~lpapp/pb2efb5f7/
[07:38] <rbasak> Are you missing a WSGIScriptAlias directive?
[07:38]  * rbasak hasn't read the paste
[07:38] <lpapp> At least there is one there.
[07:42] <lpapp> rbasak: I have this, WSGIScriptAlias / /var/www/foo/foo.wsgi
[07:43] <rbasak> I don't know then. Sorry.
[07:44] <lpapp> rbasak: ok, np. Hopefully someone has already done something like this or/and have a clue. :-)
[08:35] <rbasak> roaksoax: have you seen bug 1227252?
[08:57] <lpapp> anyone having any ideas?
[10:45] <jamespage> roaksoax, was there a specific reason the squid-deb-proxy-client-(udeb) packages only reside in universe?
[11:35] <freakynl> Hi, the auto updates are nice et all, but is there an auto garbage cleaner as well?
[11:36] <freakynl> Many boxes fall over due to the sick amount of kernel images (or more notably - initrd's) that are kept around. Just had another machine with 6GB storage that only had 2.5GB in use after first install fall over cuz it was full (it only runs powerdns, which stores a little data in mysql but that's about it)
[11:44] <greppy> freakynl: there is an apt-get autoremove, which may help.
[12:24] <freakynl> greppy: hmm last time I used that it didn't remove any kernels at all, now it removed some (left 5 lying around)
[12:25] <freakynl> Just running: 'dpkg -l 'linux-*' | sed '/^ii/!d;/'"$(uname -r | sed "s/\(.*\)-\([^0-9]\+\)/\1/")"'/d;s/^[^ ]* [^ ]* \([^ ]*\).*/\1/;/[0-9]/!d' | xargs sudo apt-get -y purge' - reduced consumed diskspace (grew it first as dpkg obviously crashed configuring kernel w/o diskspace) from 76% to 22.6% on a 5GB (now 7) disk - removing nothing but kernels, initrd's and kernel headers... I can't believe how long they ignore this issue. Auto update shou
[13:48] <roaksoax> jamespage: i didn't know s-d-p-client packages were in universe.. we werent the ones who filed a MIR for it
[13:48] <roaksoax> afaikl
[13:49] <jamespage> roaksoax, oh
[13:49] <roaksoax> jamespage: oh nevermind I guess I did
[13:49] <roaksoax> jamespage: https://bugs.launchpad.net/ubuntu/+source/squid-deb-proxy/+bug/961726
[13:49] <jamespage> roaksoax, I thought so
[13:50] <roaksoax> jamespage: so I'm guessing it's because they weren't seeded ?
[13:50] <jamespage> roaksoax, probably - I just wanted to check to see if there where any other reasons why not
[13:50] <roaksoax> jamespage: same has happened for maas-dns/maas-dhcp
[13:50] <jamespage> roaksoax, reference is bug 1183326
[13:51] <jamespage> folks are trying to use the udeb in d-i installs but it borkes
[13:51] <roaksoax> jamespage: right, but d-i with internet access should still obtain the package, shouildn't it?
[13:51] <roaksoax> jamespage: and yeah, I have no idea why they were left out, but judging for the MIR bug, they shouldn't have been
[13:52] <roaksoax> jamespage: i guess we need to seed them
[13:57] <roaksoax> jamespage: are you officially back btw?
[13:57] <jamespage> roaksoax, yep
[13:58] <roaksoax> cool
[14:11] <zul> jamespage:  since you are here https://code.launchpad.net/~zulcss/ceilometer/deps/+merge/186553
[14:13] <jamespage> zul, do those need to be applied to the binary package depends as well?
[14:13] <zul> jamespage:  *sigh* yeah probably
[14:14] <jamespage> zul, :-)
[14:20] <zul> jamespage:  im thinking we should be explicit in the binary deps as well and not rely on pydist
[14:23] <jamespage> zul, I think that makes sense
[14:29] <zul> jamespage:  done
[14:30] <jamespage> zul, does dh_python not do the versioned depends on the binary package automatically?
[14:30] <zul> jamespage:  i dont think so
[14:30] <zul> better safer than sorry though
[14:32] <jamespage> zul, headache tho
[14:32] <zul> jamespage:  true we need a way to automate this kind of stuff
[15:22] <jamespage> zul, SQLAlchemy>=0.7.8,<0.8
[15:23] <zul> jamespage:  ?
[15:23] <jamespage> zul, thats in the requires of ceilometer
[15:23] <jamespage> we have 0.8.2
[15:23] <zul> jamespage:  yeah its patched
[15:23] <jamespage> so entry points bork
[15:23] <jamespage> no its not
[15:23] <zul> isnt it?
[15:24] <jamespage> zul, no - you dropped it
[15:24] <jamespage>   * debian/patches/fix-setup-requirements.patch: Dropped no longer
[15:24] <jamespage>     needed.
[15:24] <zul> why the hell did i do that?
[15:24] <zul> ill patch it
[15:25] <jamespage> zul, no idea
[15:26] <jamespage> adam_g, where are we up to with charm redux testing and stuff? I need to be reminded
[15:26] <zul> jamespage:  alright review branch for that one after
[15:28] <jkyle> When pxe booting an ubuntu server, is the file located at /preseed.cfg during installation the one being used?
[15:28] <jamespage> zul, I like your versioning:
[15:28] <jamespage> https://launchpad.net/ubuntu/saucy/+source/python-wsme/0.5b5-1ubuntu0
[15:29] <zul> jamespage: thanks! :P
[15:29] <jamespage> zul, you need to drink more wakeup juice in the mornings :-)
[15:30] <zul> jamespage:  meh
[15:33] <jamespage> zul, do you want to push the version fixup to that deps branch? I'd like to test it out and I'm blocked right now
[15:33] <zul> jamespage:  sure gimme a sec
[15:36] <jkyle> I'm getting the following error during install of ubuntu
[15:36] <jkyle> "The following packages have unmet dependencies language-selector-common depends accountsservice"
[15:36] <zul> jamespage:  done
[15:36] <jkyle> 12.04.2
[15:37] <miketeevee> hi guys, i'm having a fork issue, can someone help me?  10.04 LTS works, 12.04.3 with the same code forks way too much memory
[15:37] <jkyle> I'm aware of the fix for these types of errors  in a running os, apt-get install -f, but how do you resolve this when the installer itself is throwing the error?
[15:37] <zul> jamespage:  we have alot of extra crap in nova deps that we dont need anymore
[15:37] <miketeevee> i have my findings in a gist
[15:37] <miketeevee> https://gist.github.com/michael-trelinski
[15:37] <miketeevee> i feel like it's a kernel flag i need to set, but i can't find it
[15:39] <D`Fam> !scans
[15:44] <jgornick> Hey guys, is there something out there that presents a user with a configuration screen or a status screen after an Ubuntu virtual appliance is booted? In my VM window, I would like to see a summary of system information, primarily network information.
[15:46] <Rory> jgornick: Would it be acceptable to show this information in the Message of the Day that appears on a sucesfull console login?
[15:46] <jgornick> Rory: Might be acceptable :)
[15:46] <jgornick> Rory: Looking for something like this: https://github.com/turnkeylinux/confconsole
[15:48] <Rory> jgornick: If you put shell scripts in the /etc/update-motd.d/ folder they are executed in order on each login
[16:00] <rbasak> jgornick: you could replace the login on vt1 with a status screen if you like. The job runs from /etc/init/tty1.conf.
[16:00] <rbasak> jgornick: (by running getty with an alternate program)
[16:01] <rbasak> Be careful to make sure that the user can't interrupt it to get a root shell or something though
[16:03] <plm> Hi all
[16:03] <Rory> Hi plm
[16:03] <plm> people I have a process running in my server but I don't know who/how it is calling. It is rsyn. I ps aux | grep -i rsync show this: # ps ax | grep -i rsync 7683 ?        Ss     8:59 rsync --server --sender -vlogDtprz . /wts/
[16:03] <plm> I'm not use rsync
[16:04] <plm> how I know what script are calling that?
[16:05] <Rory> plm: if you run the command "ps l" and find the process, look at the PPID column to find the PID of the process which spawned it
[16:05] <Rory> PPID is Parent process ID I think
[16:17] <jamespage> zul, ceilometer needs seeding for main inclusion
[16:18] <jamespage> zul, I pushed your branch - thanks
[16:19] <jamespage> zul, its not a fault of that change by ceilometer need python-babel as well now
[16:19] <jamespage> ceilometer-common fails in a clean chroot
[16:20] <plm> Rory: ps l | grep -i rsync not show about rsync...
[16:21] <Rory> plm: You need to do it when it's actually running, as the user that rsync is running as, or with sudo
[16:21] <plm> Rory: it are running
[16:21] <plm> # ps ux | grep -i rsync
[16:21] <plm> root     18090 47.7  4.2  46516 43252 ?        Rs   13:08   7:19 rsync --server --sender -vlogDtprz . /CSP/
[16:21] <plm> Rory:
[16:21] <plm> # ps a | grep -i rsync
[16:21] <plm> 19093 pts/1    S+     0:00 grep -i rsync
[16:22] <Rory> OK plm so now do "ps l | grep 18090"
[16:22] <plm> Rory: # ps l | grep 18090
[16:22] <plm> 0     0 19148  9019  15   0   2880   796 pipe_w S+   pts/1      0:00 grep 18090
[16:23] <Rory> plm: basically you need to catch the process when it's running. Find out it's PID and then use ps l to find its parent
[16:23] <plm> Rory: it are running..
[16:24] <plm> are always running
[16:24] <Rory> plm: sudo ps l | grep 18[0]90
[16:25] <Rory> The 4th column will be the parent process of PID 18090
[16:25] <miketeevee> i summarized my question into a stack overflow thread, much help would be appreciated with this fork issue http://stackoverflow.com/questions/18900127/how-do-i-control-the-vmem-when-forking-having-a-same-code-same-java-issue-on-di
[16:25] <plm> Rory: I think I find: other machine are doing a sync for this server...
[16:26] <zul> jamespage:  argh ok
[16:26] <plm> Rory: but rsync start in other machine, I not knew that in client rsync show as process too
[16:30] <adam_g> jamespage, https://code.launchpad.net/~openstack-charmers
[16:33] <jamespage> adam_g, right - I see we still need to push in the fast exit stuff and clearing of relations for non-leaders
[16:33] <jamespage> right?
[16:35] <adam_g> jamespage, yeah, what is the fast exit stuff?
[16:35] <jamespage> adam_g, in contexts - when the first complete context found gets returned
[16:36] <jamespage> right now the contexts in helpers iterate all relations and then check the last one
[16:36] <jamespage> relations/units rather
[16:36] <adam_g> oh right
[16:36] <jamespage> actually relations and units
[16:36] <jamespage> although two would be odd
[16:36] <jamespage> (relations that is)
[16:37] <adam_g> jamespage, there is a charm-helpers branch there as well, that is the sync source for all those charms
[16:37] <jamespage> adam_g, right
[16:37] <jamespage> adam_g, I can find some time to work throught that tomorrow PM
[16:37] <jamespage> (know you are busy)
[16:37] <adam_g> jamespage, cool.
[16:38] <adam_g> jamespage, are you back?
[16:38] <jamespage> adam_g, yes
[16:38] <adam_g> w00t
[16:38] <jamespage> (althought not around my time tomorrow am)
[16:44] <zul> jamespage/adam_g: http://people.canonical.com/~chucks/ca/ (for bug #1227266)
[16:49] <jamespage> zul, ooo
[16:49] <jamespage> zul, tricksy
[16:50] <zul> yep
[16:50] <zul> its apparently apart of the chef cookbooks
[16:56] <jamespage> zul, +1
[16:56] <zul> jamespage:  cool thanks
[17:04] <miketeevee> where's the best place to post my ubuntu server problem to?
[17:04] <miketeevee> messageboard-wise
[17:04] <miketeevee> stackoverflow? ubuntuforums (hacked?)?
[17:05] <rbasak> If you can phrase it as a general question that can apply to others, askubuntu is good.
[17:06] <miketeevee> irc channel?
[17:06] <miketeevee> ah i see
[17:10] <zul> adam_g:  https://code.launchpad.net/~zulcss/nova/house-cleaning/+merge/186588
[17:13] <adam_g> zul, ack. might take me a bit to get through that one
[17:14] <zul> adam_g:  yeah im going through everything with a fine tooth comb today
[17:14] <adam_g> zul, awesome
[17:33] <raub> I know that due to resolvconf I can have dns-nameservers in /etc/networks/interfaces
[17:34] <raub> But, can I also have the "options" or osmething equivalent in the same file, so it is loaded into /etc/resolv.conf?
[17:34] <raub> Or should I put it in  /etc/resolvconf/resolv.conf.d/tail?
[17:36] <sarnold> raub: resolvconf(8) suggests the ../tail answer is the right answer
[17:55] <tonyyarusso> raub: I think you can have dns-options lines in the interfaces file too
[17:59] <adam_g> zul, added one comment to that merge but looks good otherwise i think
[18:00] <zul> adam_g: cool
[18:30] <GarrettKajmowicz> Greetings! How can I go about resuming the boot process in 12.04 after I've been dumped to the Busybox shell and manually mounted the root filesystem? Alternatively, how can I figure out why I'm being dumped to busybox and fix the issue so my server can boot normally?
[18:30] <ikonia> GarrettKajmowicz: error message on screen normally gives a clue
[18:30] <GarrettKajmowicz> I'm "running" 12.04 LTS. This problem started right after I upgraded from 10.04 LTS. The old kernel image still boots correctly, but the new ones don't.
[18:33] <sarnold> GarrettKajmowicz: at least in the old days, ^D would exit a rescue shell and continue booting, dunno if that's still true, but I hope it is. :)
[18:33] <ikonia> it's going to busy box
[18:33] <ikonia> ctrl +d won't do anything
[18:35] <GarrettKajmowicz> ikonia: I don't recall a particular error message. I'm just dumped to the shell. The root filesystem isn't mounted, and manually typing "mount /dev/md0 /root -o ro" works just fine.
[18:35] <ikonia> GarrettKajmowicz: so you're using a raid device
[18:35] <ikonia> GarrettKajmowicz: it's possible it's not able to mount the root file system at boot time,
[18:35] <GarrettKajmowicz> ikonia: Yes - RAID1.
[18:35] <ikonia> check the initramfs
[18:36] <GarrettKajmowicz> ikonia: How can it not be possible to mount the root fs at boot time if I've done so from the busybox shell? Isn't that what the initramfs is for?
[18:37] <GarrettKajmowicz> ikonia: What would you like me to check in the initramfs?
[18:37] <ikonia> GarrettKajmowicz: could be a delay in assembling the array
[18:37] <ikonia> GarrettKajmowicz: or your mdadm.conf could be wrong so it tries to assemble it wrong
[18:37] <ikonia> (just easy / common issues with it)
[18:37] <GarrettKajmowicz> I've added rootdelay=30 to the kernel boot parameters and it boots and dumps to the shell within 5 seconds. It's 2 disks, 1 of which is a SSD.
[18:38] <zul> adam_g:  https://code.launchpad.net/~zulcss/cinder/housecleaning/+merge/186612
[18:38] <ikonia> GarrettKajmowicz: you're doing software mirroring on SSD's ??
[18:38] <GarrettKajmowicz> ikonia: If it was wrong, why would mounting it work correctly with no additional interventions?
[18:38] <ikonia> GarrettKajmowicz: they won't last long
[18:39] <GarrettKajmowicz> ikonia: The math I've done said it shouldn't be a problem. In any case, that isn't the immediate concern.
[18:40] <ikonia> understood
[18:41] <GarrettKajmowicz> ikonia: Where can I go from here to attempt to diagnose this issue? I've been trying to address this since July.  :-(
[18:41] <ikonia> since july !
[18:41] <ikonia> GarrettKajmowicz: first thing is look at the grub config - check it's valid, check it's pointing at the right locations, and that the files are there and valid
[18:42] <GarrettKajmowicz> ikonia: Extra data: http://askubuntu.com/questions/307509/upgrade-to-12-04lts-dumps-to-busybox-on-boot
[18:43] <ikonia> GarrettKajmowicz: are you sure it's not upgraded grub and should be using grub.conf now ?
[18:45] <GarrettKajmowicz> Grub seems to boot fine. It gets new kernel images as they are added.
[18:45] <GarrettKajmowicz> Grub is version 0.97. I don't think grub2 started being used until a newer Ubuntu version. (I'm running 12.04 LTS).
[18:45] <ikonia> so you're confident it's still reading from menu.lst not grub.conf
[18:46] <ikonia> 12.04 uses grub2 by default, but that doesn't mean it's upgraded yours
[18:46] <ikonia> GarrettKajmowicz: for arguments sake, break the kernel line - point it at a wrong filename see if it errors and what it errors
[18:47] <GarrettKajmowicz> ikonia: I can't do that at the moment (I'm at work, server's at home). I will do so once I get home tonight, however.
[18:48] <ikonia> GarrettKajmowicz: ping me when you get home, maybe interesting to work it through
[18:49] <GarrettKajmowicz> ikonia: Absolutely. Will you be here on in about 3-4 hours?
[18:49] <ikonia> I suspect so
[18:50] <GarrettKajmowicz> ikonia: Okay - I'll come and harass you then. I'll be typing via smartphone instead of keyboard, so that'll be interesting.
[18:51] <ikonia> we can work with that
[19:12] <mgw> I'm trying to figure out the best way to calculate cpu usage on an lxc container
[19:12] <mgw> is there something in cgroups that will give me that, or do I need to use ps and then compare cpu% against the cpushares?
[19:40] <adam_g> zul,  https://code.launchpad.net/~zulcss/nova/house-cleaning/+merge/186588 <- you merged before approved?
[19:42] <zul> adam_g:  yeah did the wrong alias i have locally
[19:42] <zul> sorry about that wont happen again
[19:42] <adam_g> zul, its okay
[19:43] <adam_g> zul, im gonna push a fix to put the python-six version specification in d/control
[19:43] <zul> adam_g:  ack
[19:43]  * zul disapears for a while
[19:45] <adam_g> zul, also.. i know i mentioned before but please get out of the habbit of pushing directly to the branches.
[21:08] <adam_g> zul, https://bugs.launchpad.net/ubuntu/+source/python-cliff/+bug/1227884  no idea how that didn't FTBFS in the archive.. but trying to backport to precise is failing
[21:38] <maxped> hey guys, dumb question. Why does one computer, my ubuntu-server ver 12.04, get an unusally high ip number from my router? everything else, iphones, androids, other laptops, get 192.168.0.100-109. the ubuntu server gets *.119. when i try to set static and match setting on router reservation list i can never acess from outside the network. any input? TIA!
[21:39] <RoyK> maxped: the dhcp server is doing the job ;)
[21:42] <maxped> RoyK: thanks for the reply, i thought the reservation on the router was above anything else. i obviously need to learn more, but is this problem more wifi-router or ubuntu settings? so i know where to start to look. i think router.
[21:43] <bjf> on a maas server what does "node added but never seen" mean?
[21:43] <bjf> i'm trying to enroll a single system
[21:45] <maxped> RoyK: and why so high an ip? why not just next available?
[21:45] <RoyK> no idea
[21:46] <maxped> RoyK: hmmm...thanks.
[22:09] <jkyle> so, I've discovered the current 12.04.3 version of precise has a broken openvswitch module  out the box
[22:09] <jkyle> fix is to roll back the kernel or build your own module with a current ovs version. Is there a backports that might have the current ovs versions for precise?
[22:14] <jkyle> ah, thanks for the blog post Mr. Page => http://javacruft.wordpress.com/2013/08/21/openvswitch-for-ubuntu-12-04-3-lts/
[22:21] <GarrettKajmowicz> ikonia_: I'm baaaack!
[22:21] <bananapie> is there a way I can execute a program and see all the files it accesses or tries to access ?
[22:22] <warzauwynn> bananapie: strace cmd
[22:22] <bananapie> cool :D
[22:22] <GarrettKajmowicz> bananapie: strace
[22:25] <bananapie> YES! This is exactly what I needed. Thanks warzauwynn and GarrettKajmowicz. I see hundreds of 'no such file or directory' errors :D
[22:26] <GarrettKajmowicz> ikonia_: I tried getting grub to boot non-existant kernels or initramfs and grub properly detected that the file didn't exist.
[22:26] <sarnold> bananapie: skip past the usual ld stuff at the beginning, there's always hundreds or thousands of those..
[22:28] <bananapie> thanks
[22:36] <GarrettKajmowicz> ikonia: I managed to save dmsg output from a failed boot: pastebin.ca/2455008
[22:40] <sarnold> GarrettKajmowicz: is this bit intentional? Kernel command line: root=/dev/md0 ro debug
[22:43] <GarrettKajmowicz> sarnold: I added 'debug' for this run only. The other flags have been part of the boot params forever.
[22:43] <sarnold> GarrettKajmowicz: aha :)
[22:44] <sarnold> .. sure enough, my /proc/cmdline has an 'ro' in it too. sheesh. I need to look around more. :)
[22:46] <GarrettKajmowicz> I think the idea is to mount the rootfs ro until it is confident of being able to actually boot the system  so as to avoid powering down with a dirty fs.
[22:52] <GarrettKajmowicz> I'm back. Webchat + phone browset + phone multitasking = frustration
[22:55] <sarnold> GarrettKajmowicz: :/ you missed nothing while you were gone though
[22:55] <GarrettKajmowicz> Any other stuff you'd suggest I look at?
[22:55] <shauno> it's quite normal to boot readonly and remount rw mid-boot.  I think it remounts right after it decides if fsck is to be run (because fsck on a rw partition is mindboggling)
[22:57] <sarnold> hrm, /var/log/boot.log is remarkably .. short of information
[22:58] <sarnold> can you poke around in your logs to find corresponding userspace messages from the failed boots?
[23:01] <GarrettKajmowicz> sarnold: Can you clarify? In failed boots userspace doesn't get a chance to run, and the filesystem is never mounted rw to be able to save anything anyways.
[23:02] <sarnold> GarrettKajmowicz: oh right :( It it is just that dmesg is only one part of the problem, and nothing there particularly stands out as troublesome. (the days-since-fsck isn't great but also shouldn't explain the busybox shell.)
[23:07] <GarrettKajmowicz> Right. Busybox shell is repeatable. Server booted up with an older kernel and successfully ran fsck.
[23:10] <GarrettKajmowicz> sarnold: What else can I do to investigate?
[23:27] <GarrettKajmowicz> I'm somewhat frustrated at this point. I need to be running the new kernels to address a munor networking issue, yet there don't seem to be any diagnostics for the boot subsystem.
[23:54] <gartral> how can I make a folder have group read/write perms and make that inherent to all *future* directories and files in that folder?
[23:55] <gartral> /s/folder/directories
[23:58] <sarnold> gartral: there is nothing you can do to force specific _permissions_ on files and subdirectories. the only things you can change are (a) set the setgid bit on the directory to force all future subdirectories and files to have the same group _ownership_ (b) set the sticky bit to prevent users from deleting files and directories they do not own.
[23:59] <gartral> sarnold: that first one seems like what I want
[23:59] <sarnold> gartral: you could try hacking something together with e.g. incron, but it'd be easy to write a security problem that way :)