[00:58] <schmitt9531> I was wondering if you could help me out with the qemu-kvm setup. I'm trying to help my friend with his ubuntu hypervisor. He currently has libvirt running to implement qemu-system
[00:58] <schmitt9531> I connected to qemu:///session it did not work, when I did qemu+ssh I got access denied, added myself to libvirtd, it worked but I got qemu system access
[00:58] <schmitt9531> instead of qemu session
[00:59] <schmitt9531> and it gave me access to qemu system which has root privs
[00:59] <schmitt9531> how do I properly use qemu usermode?
[00:59] <schmitt9531> or is that not supported in ubuntu
[02:09] <bearly230> Hello, all I have what I hope is an easy fix question. I have a centos server with a group made, and users added to the group.  But my ubuntu 12.04 desktop clients can't access the folder over nfs. Any sugestions?
[02:18] <bearly230> Hello, all I have what I hope is an easy fix question. I have a centos server with a group made, and users added to the group.  But my ubuntu 12.04 desktop clients can't access the folder over nfs. Any sugestions?
[02:42] <patdk-lap> enable group usage on nfs
[09:09] <smoser> lifeless, dhcp
[10:34] <orogor> hi
[10:34] <orogor> may anyone helps me with checking the trip size chunk size and whatever for ext4 over lvm2 over an mdadm raid10 array?
[10:35] <hallyn> stgraber: so fwtw my problem turned out to be with upstart's mounted-dev.conf.  It calls /sbin/MAKEDEV console which recreates /dev/{console,ttyN}.  That needs to not happen in a container.
[10:59] <hallyn> oh, nope, guess that actually comes from mountall
[11:36] <koolhead17> lynxman, hey
[11:37] <lynxman> koolhead17: hello
[11:37] <koolhead17> lynxman, where are you?
[11:37] <lynxman> koolhead17: London
[11:38] <koolhead17> lynxman, not in UDS
[11:38] <lynxman> koolhead17: nope
[11:39] <roaksoax> SpamapS: any examples of a subordinate charm?
[12:56] <SpamapS> roaksoax: one example of a subordinate is nrpe
[14:34] <iggi_> Anyone here familiar with open-iscsi and mounting drives on boot? I have my startup to automatic yet my boot still hangs where it is waiting for the iscsi drive. I am able to go to the server and manually skip the waiting and it will boot where I can manually log into the NAS.
[14:34] <iggi_> I just want it to connect at startup
[14:35] <iggi_> and I would like to disable the waiting screen since I can't get into the system remotely
[15:01] <jpds> Anyone know why aftering upgrading my KVM host to 12.10, a 12.04 guest running cobbler can't give out DHCP addresses? http://paste.ubuntu.com/1315428/
[15:11] <fidel_> anyone used to 3ware raid controllers? i am using a 9650SE-12ML on a freshlyinstalled 12.04 - 64bit. i executed the installer from the code-set-iso and selected cli & webinterface. tw_cli works to aks the controller status - but the webinterface seems NOT to be accessible at all on http/https & default port 888. Any ideas? or any other place you would recommend to ask?
[15:13] <fidel_> btw: netstat shows nothing with port 888
[15:15] <ewook> fidel_: long time ago I fiddled with 3ware cards, but the web gui was broken back then as well ;).
[15:15] <fidel_> ewook: hehe
[15:15] <fidel_> ewook: it worked for me in 10.04 days
[15:16] <fidel_> so it should be possible to get it working ...at least ;)
[15:16] <fidel_> bloody 3ware/lsi stuff
[15:17] <RoyK> fidel_: I've used 3ware controllers with linux for a decade or so, but never tried that webinterface...
[15:17] <fidel_> RoyK: so you stick to tw_cli only or the bios as well?
[15:17] <RoyK> not yet tried with 12.04 either...
[15:17] <RoyK> yeah, it works
[15:18] <RoyK> tw_cli with nagios/icinga plugin is good
[15:18] <fidel_> sounds great ;)
[15:18] <RoyK> but then, I guess I'll stick to software raid on future servers
[15:18] <fidel_> but well - as it worked in the past i am pretty sure one of the team is "in need" for the web-ui ;)
[15:18] <udsslayer> I don't suppose anyone who is at UDS could help me setup a guest VM with a public IP?
[15:19] <fidel_> apart from that: i know the rest works - i would like to get the webinterface working as well
[15:19] <RoyK> fidel_: they don't need it unless you tell them there is one ;)
[15:19] <fidel_> RoyK: it worked in the past - they used it .... doomed ;)
[15:20] <RoyK> well, setup icinga for monitoring and they'll have an even fancier webinterface for monitoring far more than just the raid controller ;)
[15:21] <fidel_> thanks for the tips - but i am not asking for workarounds or other solutions. basically i would at least test for some more time if its possible to get the webinterface working
[15:21] <fidel_> sidenotwe: we are running nagios here ;)
[15:21] <fidel_> so - close to icinga
[15:22] <RoyK> I used to run nagios, but nothing much has happened to that codebase for, what, 5 years?
[15:22] <fidel_> pretty silent around nagios yep
[15:22] <RoyK> Ethan chose to move to Nagios IV, being commercial
[15:23] <RoyK> the Icinga people have rewritten quite a lot over the last 2-3 years
[15:23] <RoyK> nagios config files are still compatible, though
[16:11] <edve> anyon knows a good alternative to landscape ?
[16:13] <RoyK> edve: for what use? landscape has a few different uses (or a lot)
[16:13] <skipp3r> hi all, i want to send the output of a script under /etc/cron.daily/ to /dev/null, can i define the sdout inside the script? or how cna i send the sdout just of this script to /dev/null ?
[16:13] <edve> Monitoring , getting updated easily for some none linux user
[16:14] <RoyK> skipp3r: just create a wrapper script and run "yourscript > /dev/null"
[16:14] <RoyK> skipp3r: perhaps with "2>&1" if you want to get rid of stderr as well
[16:15] <RoyK> edve: for monitoring, there are several systems to choose from, for updating easily from outside the console or terminal, I don't know any, but it's easy to cron that part
[16:15] <RoyK> !monitoring
[16:15] <skipp3r> RoyK: the script is under /etc/cron.daily/ where are more scripts, i want to change the output just for a specific one
[16:15] <RoyK> !icinga
[16:16] <RoyK> !nagios
[16:16] <RoyK> stupid bot
[16:17] <RoyK> skipp3r: just move that script to /usr/local/sbin or something and create a wrapper script in /etc/cron.daily
[16:17] <RoyK> skipp3r: or modify the script to let the parts that create output redirect their output to /dev/null
[16:18] <skipp3r> RoyK: how can i modify it? can i somehow inside the script redirect the output to /dev/null ?
[16:19] <RoyK> skipp3r: editor /path/to/script
[16:19] <RoyK> it's just a script!
[16:20] <skipp3r> RoyK: yes, i mean i should just add >/dev/null inside the script? is that enough?
[16:20] <RoyK> just find the noisy parts and add >/dev/null to them
[16:21] <skipp3r> RoyK: ok i will try that, thanks!
[16:21] <recon_lap> bit stuck trying to use sendmail , keep getting this error Oct 29 16:18:37 xxx-Aspire-V3-571 sendmail[5892]: My unqualified host name (xxx-Aspire-V3-571) unknown; sleeping for retry, just what it for testing a web site account verification system on my local machine. anyone able to give a hint what i need to configure?
[16:21] <RoyK> if they write to stderr, which is normal for errors, add 2>&1 after >/dev/null to redirect file 2 (stderr) to file 1 (stdout)
[16:21] <RoyK> recon_lap: erm - do people use sendmail anymore? ;)
[16:21]  * RoyK sticks to postfix
[16:22] <RoyK> some people like exim as well
[16:23] <recon_lap> RoyK: I'm using it through SMTP from a web server for email address verification. not sure how postfix/exim relate to this
[16:23] <RoyK> if you don't have a personal preference for sendmail, I'd recommend using postfix (or perhaps exim, but I don't know that too well)
[16:25] <recon_lap> RoyK: it's just a development system, I just want it to work so i can test some code. not going to be uses as a production mail server, but I'll have a look at postfix
[16:26] <RoyK> postfix is generally easier to configure for most stuff
[16:26] <RoyK> unless you're a sendmail wiz, I'd recommend switching
[16:26] <recon_lap> RoyK: lol, no whiz, it's just whats installed by default
[16:27] <RoyK> I don't think sendmail has been installed by default
[16:27] <RoyK> probably exim
[16:27] <RoyK> dpkg -l | grep sendmail
[16:34] <recon_lap> RoyK: thx, I'm looking into it :)
[16:35] <recon_lap> the bit I don't get is entering a FQDN(fully qualified domain name), I don't have one as far as i know, just using localhost.
[16:36] <IdleOne> RoyK: if you are inclined to write up a nice wiki howto page on wiki.ubuntu.com for icinga we can write a factoid pointing to it.
[16:36] <recon_lap> should I just use my host name?
[16:37] <RoyK> IdleOne: ok
[16:37] <IdleOne> Cool :)
[17:18] <bubu\a> hi guys - I want to use logrotate. Whats the difference in putting in an entry in /etc/logrotate.conf
[17:18] <bubu\a> or creating a specific conf in /etc/logrotate.d/?
[17:19] <bubu\a> i want to rotate hourly...
[17:58] <sean-lee> hi
[18:40] <xsl> hello all ... i have a LXC/APPARMOR/INITSCRIPTS problem
[18:41] <xsl> I want to use the funcionality of autostart a container wen its linked in /etc/lxc/auto but i get this error in syslog .. and the container does not start
[18:41] <xsl> Helena Soares 17:23:38
[18:41] <xsl> não será juntamente com o salário, mas será este mês	
[18:41] <xsl> NunoSeita 17:31:00
[18:41] <xsl> juntamente ker dizer o quê? não é no "recibo" ou não é no mesmo dia?
[18:41] <xsl> oops
[18:41] <xsl> wrong paste
[18:41] <xsl>  /shame
[18:42] <xsl> type=1400 audit(1351535408.081:10): apparmor="DENIED" operation="mount" info="failed type match" error=-13 parent=1384 profile="/usr/bin/lxc-start" name="/proc/" pid=1416 comm="lxc-start" fstype="proc" srcname="proc" flags="rw"
[18:43] <xsl> i tried uncomenting in /var/lib/lxc/<container_name>/config the string lxc.aa_profile = unconfined
[18:43] <xsl> but with no luck
[18:55] <Phibs> is there any way to change to 1000HZ vs 100 in the kernel ?
[20:00] <stiv2k> hello
[20:00] <stiv2k> i have a /dev/sdb mounted as my /home tree on my server and lately it keeps alerting me that /dev/sdb will be fsck'ed on next reboot. so I rebooted it but it still says /dev/sdb need to be fsck'ed. what do i do?
[20:11] <maxb> What's alerting you? the motd when you log in?
[20:13] <stiv2k> maxb yes
[20:14] <stiv2k> and in dmesg
[20:14] <stiv2k> [   16.744800] EXT4-fs (sdb): warning: checktime reached, running e2fsck is recommended
[20:15] <maxb> OK. Check at reboot is controlled by a field in /etc/fstab. You could either ensure that's not zero, and reboot, and see if it checks then, or you could unmount sdb and check it separately whilst the rest of the machine stays up
[20:15] <stiv2k> maxb, it has my /home tree, so i dont think i can unmount it right
[20:16] <maxb> You'd need to ensure no normal users were logged in before you could unmount it
[20:16] <stiv2k> lol
[20:16] <stiv2k> sounds like a catch 22
[20:16] <maxb> Though you could be logged in as root or as another user specifically created with a home outside /hoem/
[20:16] <stiv2k> ill probably just do the first method
[20:17] <maxb> Find the line of your /etc/fstab that refers to /dev/sdb and paste it here?
[20:17] <stiv2k> ok
[20:17] <stiv2k> maxb, i made it really simple
[20:17] <stiv2k> UUID=66aa7d2b-2805-47a3-9f18-2b8d2ace60a2 /home         ext4
[20:18] <maxb> hm is there no more? there should be more fields after that
[20:18] <stiv2k> i never added any
[20:19] <stiv2k> it just worked so i left it like that :S
[20:21] <maxb> stiv2k: After the ext4 add                          defaults   0  2
[20:22] <stiv2k> whats the 2 for?
[20:22] <maxb> Not having a number in the sixth field probably is why it didn't get checked
[20:22] <maxb> 2 is for check this, but do it after the root file system (which should have number 1 there)
[20:23] <maxb> 'man fstab' for more in-depth discussion
[20:23] <stiv2k> ah ok
[20:23] <stiv2k> now just reboot?
[20:24] <maxb> Just to double-check, the entire fstab line should now look like:
[20:24] <maxb> UUID=66aa7d2b-2805-47a3-9f18-2b8d2ace60a2 /home         ext4   defaults   0  2
[20:24] <maxb> At which point, yes, rebooting ought to cause the drive to be checked
[20:24] <stiv2k> sweet
[20:29] <orogor> hi
[20:34] <briggz> Hey guys, anyone here a server admin professionally?
[20:35] <briggz> I'm looking for final project ideas for my intro to linux class.
[20:35] <orogor> <=
[20:35] <briggz> server specific area to study that would be useful for future admins.
[20:36] <orogor> san ?
[20:36] <orogor> if you have access to the hardware
[20:36] <briggz> what's san?
[20:36] <orogor> storage area network or something like that
[20:36] <orogor> storage attached n...thing
[20:37] <orogor> zonning fiber  san switch , stuff like that
[20:38] <bbhoss> Anyone have issues with runit not sending signals to the correct process when you use chpst? It seems like it is sending the signals to the chpst process instead of the process chpst runs
[20:38] <briggz> oh, looking more for software server side.
[20:38] <orogor> never had the courses at univeristy, you need to have like 10k€ just to make a single setup that not everyone will play with
[20:38] <briggz> like we're doing samba stuff now.
[20:39] <orogor> you want software only course?
[20:39] <briggz> it's not a hardware focused course.
[20:39] <briggz> it's linux operating systems
[20:40] <orogor> software raid + lvm  is also usefull , and they need to grasp the lvm concept
[20:40] <briggz> like I installed unbuntu-server
[20:40] <orogor> and you can do the course with commodity hardware
[20:42] <briggz> Im one of the 'they' I'm a student looking for ideas on my final project that I get to pick. I asked my teacher for ideas, he said look at a list of things that might benefit a career.
[20:42] <orogor> how long the projet is?
[20:45] <briggz> I've got 8 weeks to do it
[20:45] <briggz> we need to select a topic and start this week.
[20:45] <briggz> plus it needs to be approved.
[20:51] <orogor> sorry had to reboot to fix boot sequence :)
[20:51] <briggz> that's ok
[20:52] <orogor> so , how long?
[20:52] <briggz> I have 8 weeks
[20:53] <briggz>  to compete
[20:53] <orogor> gotta be networked or be usefull in any way ?
[20:54] <orogor> like setup nfs/samba shares for roaming profiles for students would be nice maybe
[20:54] <briggz> it's got to be something that has to focus on server and/networks and more specifically Linux
[20:54] <briggz> wer
[20:54] <orogor> no ?
[20:54] <briggz> we're already doing samba this week
[20:54] <orogor> like wherever you login your desktops and my document folder is always the same
[20:55] <briggz> well, we're setting it up and playing with it
[20:56] <marsje> my friend just upgraded from ubuntu server 10.04 to 12.04 and now it seems his dns settings are wrong... how can I see what is the current dns-server? how can I change it?
[20:56] <orogor> boot a computer and get it automagically installed
[20:56] <briggz> no specific tasks in it, something might come up in the coming weeks. but as of right now, we're making windows and our servers see each other, share, etc..
[20:56] <orogor> cat /etc/resolv.conf
[20:58] <orogor> virtualisation is also good to play with if you never tried it
[20:58] <briggz> done that. =)
[20:58] <briggz> I have my actual ubuntu server, my windows machine, and 4 virtual linux boxes on it.
[20:58] <xsl> guys i need help with LXC/Apparmor ( get this error message: type=1400 audit(1351535408.081:10): apparmor="DENIED" operation="mount" info="failed type match" error=-13 parent=1384 profile="/usr/bin/lxc-s
[20:58] <xsl> tart" name="/proc/" pid=1416 comm="lxc-start" fstype="proc" srcname="proc" flags="rw")
[20:59] <briggz> they're even seprate networks.
[20:59] <briggz> is there something specific in security that you might think would be good to know more in depth?
[20:59] <xsl> i tryed commenting out the line lxc.aa_profile in /var/lib/lxc/container_a/config with no luck
[21:01] <orogor> roles base access c..thing  rbac
[21:01] <orogor> see apparmor
[21:01] <orogor> nobody ever use configure this kinds of stuff however
[21:02] <xsl> i didnt wanted to disable the profile, i been searching but i dont quite get it how to configure that "mount"
[21:02] <orogor> but it s  the only real way i ve found to get a secure system
[21:02] <xsl> they just disable ? ... lack of security feature imo ... right?
[21:03] <orogor> basically you start apparmor in profiling mode , run the system and makes it do all thge stuff it usually does
[21:03] <orogor> then put the profile in enforce mode , and after that it can only do the things it could do while in profile mode
[21:04] <xsl> and then i commit the changes
[21:04] <xsl> yeah i see where your going to
[21:04] <orogor> xsl,  not speaking to you :)
[21:04] <xsl> its easyer than writing myself the profile
[21:04] <orogor> but you might learn as well
[21:04] <xsl> your not? ... but seems like the solution to my problem
[21:04] <orogor> xsl,  yhea  it maybe it
[21:05] <orogor> xsl, in fact afterward you slightly edit the profile and put * in some places , like for a cookie folder
[21:06] <orogor> else it wouldn t be able to create cookie-zer3zer41ze5rzer4.txt  if this wasn t created during the profiling step
[21:06] <orogor> but then that s only if you re creating your own profiles
[21:07] <orogor> else it may means you re encountering a bug of some kind that you should report , or need an update cause it s already fixed
[21:08] <orogor> xsl, ?
[21:08] <orogor> :msg nickserv hi
[21:08] <xsl> i have just a plain install of precise 12.04
[21:09] <xsl> created my lxc container of a lvm
[21:09] <xsl> at it gives me that error above wen i reboot the server and the container should start automaticaly
[21:10] <xsl> im trying to boot the container manually in debug mode ... takes alot of time imho
[21:11] <orogor> humm bug is marked as solved, odd
[21:11] <orogor> stupid question , why not try another virualisation solution ?
[21:12] <xsl> my goal is not rly virtualization
[21:12] <xsl> its more like the jails in freebsd feeling i want
[21:12] <orogor> chroot ?
[21:12] <xsl> chroot not that good as lxc imo
[21:12] <xsl> cannot define mem limits and such
[21:12] <orogor> you want really a loot of differt process in their own containers?
[21:13] <orogor> a lot of
[21:13] <xsl> 1 container just for mysql 1 container just for php 1 container just nginx
[21:13] <xsl> something like that
[21:13] <orogor> would be me it d all go ina  single vm
[21:14] <xsl> call me paranoid :D
[21:14] <orogor> and not a chroot or jail , but a  full vm
[21:14] <orogor> so your desktop upgrade won t break your carefully crafed vm config for php and nginx
[21:15] <orogor> and you can snapshoot and rollback dureing devellopement
[21:15] <xsl> its a server
[21:15] <xsl> and i can snapshot LVM
[21:15] <orogor> not the same thing
[21:18] <orogor> well anyhow if you know lxc use that, and fix the apparmor issue
[21:18] <orogor> if it s single process completely isolated it s  not that bad to turn off apparmor i guess
[21:19] <orogor> else report the bug to ubuntu or fix it yourself by doing the profile and enforcing it
[21:20] <xsl> i think its an initscripts problem
[21:20] <xsl> i can start the container manually
[21:20] <xsl> only the startup of the server fails
[21:33] <briggz> orogor:  I think I have a idea of my topic.
[21:33] <briggz> thanks for the help
[22:13] <Zer0Glitch> Good day, folks
[22:16] <Zer0Glitch> Can anyone recommend ideal RAID build for small apache/sql server?
[22:16] <Zer0Glitch> I'm thinking of using 6x 120GB SSDs
[22:16] <Zer0Glitch> (128)
[22:20] <Zer0Glitch> Anyone registering this?
[22:22] <Zer0Glitch> Jesus Christ, people. If you're going to occupy space in a help and discussion-focused chat room quit running bots/AFK and respond to inquiries or logout
[22:23] <jeremydei> well, i was gonna say.. 6xssd's sounds pretty screamin fast .. that's all i got sry :)
[22:26] <Zer0Glitch> Looks better in this: http://www.thermaltakeusa.com/Product.aspx?C=1156&ID=2031
[22:26] <Zer0Glitch> Running in one of these: http://www.antec.com/product.php?id=2149&pid=8
[22:29] <bbhoss> Zer0Glitch: sounds like overkill. Also, nginx > apache
[22:31] <Zer0Glitch> Hoss > Sounds like overkill, but I'm going to be running high IOPS because I'll be hosting a high-visitation site off of it. Combination of Wordpress & custom CRM. I'll have a larger fibre-channel rackmount for the media storage.
[22:32] <Zer0Glitch> Low latency, high uptime
[22:32] <Zer0Glitch> I want to minimize my maintenance and chances for hardware failure, so SSDs make sense
[22:34] <Zer0Glitch> So again, anyone have any suggestions on primary raid setup for server? I was thinking Raid5
[22:35] <patdk-lap> hmm? raid5 is low iops and high latency
[22:35] <Zer0Glitch> First time raiding a server. Hence the question
[22:35] <Zer0Glitch> <-- clearly needs assistance
[22:36] <Zer0Glitch> I've been running off my home server and run into problems. Now making move to upgrade and need recommendations.
[22:36] <patdk-lap> http://bytepile.com/raid_class.php
[22:37] <patdk-lap> generally, you use raid10 for everything, except if it's for streaming workloads, then raid6
[22:37] <patdk-lap> but high iops is normally always going be a raid10 config
[22:38] <Zer0Glitch> Why Raid6 for streaming? A big part of the content is podcasts, but I'm moving it to Torrented distribution - first download originating at my site. Not streaming.
[22:38] <Zer0Glitch> Secondary distribution P2P
[22:39] <patdk-lap> as it sounds like podcasts and torrents should be streaming, they defently are not
[22:39] <patdk-lap> atleast when your user count is >4
[22:39] <patdk-lap> torrent is inherently designed to be random
[22:40] <Zer0Glitch> I'm getting 10K hits/mo
[22:40] <patdk-lap> raid6 is only better for streaming, cause you gain extra space really
[22:41] <patdk-lap> where the performance of raid6 and raid10 could be close, just raid6 assumes your moving large amount of data per transaction
[22:41] <patdk-lap> and raid10 doesn't make that assumption
[22:41] <patdk-lap> raid5 is the same, just it's inherently unsafe, expecially the larger disks get
[23:04] <jbrowne> https://cloud-images.ubuntu.com/query2 appears to be throwing 404s.  I know the existing format is to be deprecated, but the new format (from the announcement) is supposed to still live at that URL (per the examples).  Did I miss another announcement, or has it just broken as a part of releasing the latest refreshed Precise cloud images from last week?
[23:20] <blendedbychris> anyone decent at troubleshooting snmpd?
[23:40] <pmatulis> !ask | blendedbychris