[01:17] <smoser> oh neat, hallyn_
[01:30] <tohuw> Is there a simple ldapsearch command to enumerate all objects in the directory, with just their cn's and type?
[02:41] <mi2com> anyone in the Philadelphia, PA region looking for a job?
[02:41] <mi2com> in a new datacenter
[02:42] <mi2com> we dont really like ubuntu
[02:42] <mi2com> but
[02:42] <mi2com> ...
[02:42] <mi2com> +its okay :)
[03:43] <soy_el_pulpo> mi2com: I can move....
[06:21] <Bitbyte> Any of you guys use a geforce gt 220 on the server ?
[06:22] <Bitbyte> On Ubuntu 13 ?
[07:21] <sanderj_> Someone accidentely chown'ed all files in / to the www-data user.. do anyone know if there is a way of changing all permissions on a clean 12.04 os back?
[07:32] <sanderj_> is it possible to overwrite all permissions or even all files straigt from a ubuntu image?
[07:47] <freakynl> Hi, if I configure (statically in /etc/network/interfaces) 3 DNS servers for IPv4 and 3 for IPv6 only the 3 IPv4's are picked up
[07:47] <freakynl> apparently there's a limit of 3 ...
[07:47] <freakynl> I would need at least 2 on IPv4 and 2 on IPv6, is there any way to do that?
[07:52] <sanderj_> Would it work to do this to overwrite all files in a os?: debootstrap --arch i386 precise / http://archive.ubuntu.com/ubuntu
[07:53] <sanderj_> And how do I get the server version with debootstrap?
[09:05] <Senor> Is there wayes or tools I can monitor the state of a tcp server?
[09:22] <RoyK> Senor: what do you mean "state"? if it's listening or not?
[10:03] <Senor> RoyK : network and memory performence
[11:28] <swaT30> anyone from the Ubuntu OpenStack PPA team around?
[11:29] <koolhead17> swaT30: i wud assume yes
[11:29] <swaT30> just would like to see when https://bugs.launchpad.net/quantum/+bug/1169706 is going to be merged into the Grizzly PPA :D
[11:45]  * koolhead17 pokes zul 
[11:52] <zul> grizzly ppa?
[11:53] <zul> should already be there now via the cloud-archive
[12:54] <swaT30> zul: ok, thanks for the heads up! will see if an update fixes this. Do you happen to know what version it was compiled in?
[12:54] <zul> adam_g:  saucy-havana build deps when you get a chance http://people.canonical.com/~chucks/ca/
[12:54] <zul> swaT30:  not off the top of my head no
[12:55] <swaT30> zul: is there anything I could poke through to find out?
[12:55] <zul> swaT30:  other than the changelogs no
[12:57] <zul> yolanda:  ping https://code.launchpad.net/~zulcss/nova/2013.2.b1/+merge/166778
[12:57] <yolanda> hi zul
[12:57] <yolanda> ok
[12:59] <zul> yolanda:  just a heads up there is going to be like 7 merge proposals today
[13:00] <yolanda> fun!
[13:00] <zul> yeah..
[13:06] <yolanda> zul, i'm doing the dep-8 tests for nova. I see some packages as nova-volume, nova-ajax, etc... that are classified as oldlibs, and they don't have any content in the package. What are these?
[13:07] <zul> yolanda:  nova-volume is what cinder use to be called and nova-ajax is a webend front console, you can safely ignore them
[13:10] <senjin> Hi I am not able to passthrough a pci device in kvm and I keep getting errors in dmesg log "disallowing device assignment. Re-enble with "allow_unsafe_assigned_interrupts=1" module option." Does anyone have any ideas on whats wrong?
[13:12] <mardraum> did you google that option?
[13:13] <senjin> Yes and got a sugestion to add it under modprobe.d/ wich resultet in making it inpossible to start vm's at all
[13:14] <zul> yolanda:  https://code.launchpad.net/~zulcss/keystone/2012.2.b1/+merge/166790
[13:14] <mardraum> senjin: can you describe exactly what you did that didn't work?
[13:17] <senjin> Yes I created a new file in /etc/modprobe.d/ and added options kvm allow_unsafe_assigned_interrupts=1 to that file. And after a reboot virsh start vm would not start the vm
[13:19] <mardraum> I'm no passthru expert, but does this make sense? http://spica-and-roid.blogspot.com.au/2012/07/howto-kvm-passthrough.html
[13:20] <mardraum> looks like a few extra steps
[13:26] <zul> yolanda:  https://code.launchpad.net/~zulcss/glance/2012.2.b1/+merge/166794
[13:31] <hallyn_> jamespage: starting regression tests now with raring qemu with the rbd async flush patch.  could you email me the excerpt of the ping you got from the ceph folks?  (hopefully it includes justification that we can use in an sru)
[13:51] <senjin> mardraum: Now I cant start the VM again but still no passthrough but I will keep looking
[13:56] <zul> yolanda:  https://code.launchpad.net/~zulcss/horizon/2012.2.b1/+merge/166796
[13:58] <yolanda> Daviey, i'm targetting the openstack dep-8 tests to raring, because the openstack-ubuntu-testing is still on raring, am i right?
[14:05] <zul> yolanda:  half way done https://code.launchpad.net/~zulcss/quantum/2013.2.b1/+merge/166805
[14:06] <hallyn_> jamespage: http://people.canonical.com/~serge/qemu-rbd-async has the packages
[14:06] <hallyn_> not sure if you want to do some perf runs
[14:31] <Daviey> yolanda: Hmm, no - should be saucy
[14:31] <roaksoax> yolanda: everything that you do applies to the *current* development cycle
[14:31] <roaksoax> yolanda: this means that the dep-8 tests are for saucy, not for raring
[14:32] <roaksoax> yolanda: the BP was initially filled for raring, however, since it wasn't done there and you started working on it, you should have retargetted it to saucy
[14:36] <yolanda> oh i did all the tests for saucy, but i was testing the nova ones against raring
[14:36] <yolanda> anyway they are working with saucy
[14:40] <hxm> hello
[15:02] <yolanda> zul, working with the nova tests i found a bug in nova-spiceproxy. It depends on websockify but the dependency is not listed
[15:03] <zul> yolanda:  ok
[15:03] <yolanda> installing websockify by pip gives errors
[15:03] <yolanda> do we have package for it?
[15:08] <roaksoax> yolanda: you can check whether we have the packages: packages.ubuntu.com
[15:08] <roaksoax> yolanda: https://launchpad.net/ubuntu/+source/websockify
[15:11] <yolanda> i see, we have it, yes, nova-novncproxy is depending on it
[16:11] <p0wn3d> Anyone know of a good way to document current server settings compared to a default ubuntu server install
[16:14] <markthomas> p0wn3d: what kind of settings?  Installed packages?
[16:16] <hallyn_> stgraber: pushed those three patches to staging, thanks
[16:18] <p0wn3d> configuration settings for services
[16:19] <p0wn3d> "/etc" will have most of the files and settings but I want to make sure I get them all
[16:20] <markthomas> p0wn3d: You can check /etc/default, but as for the rest, there is no "good way".  There is no standard way in which services have to define their configurations.  When services use a conf.d directory and the service has been configured in a disciplined way, you can compare that way.  But it's not guaranteed to give you everything you need.
[16:21] <p0wn3d> thank you
[16:21] <markthomas> There is one good way: use something like puppet, chef, or (eventually) juju to deploy the configuration.
[16:21] <markthomas> Then, you can redeploy as needed.
[17:20] <pcn> Hi #ubuntu-server
[17:20] <pcn> Is there anyone here from canonical who can fix the s3 perms for the apt archive?
[17:20] <pcn> I'm having this issue in aws:
[17:20] <gstudent2> Hi! Hi need a new root server and I found a great offer that is virtualised via KVM. Now I want to use KVM in this machine. Is this a bad idea or is this commonly done?
[17:21] <pcn> https://gist.github.com/pcn/5686473
[17:21] <sarnold> gstudent2: at least the virtual-maas juju charm does exactly that for testing purposes. it might be slower than you like, but ought to work.
[17:22] <sarnold> pcn: note that natty went out of support six or seven months back: https://wiki.ubuntu.com/Releases
[17:22] <gstudent2> sarnold, thank you!
[17:23] <sarnold> pcn: you can still get to pcakages via http://old-releases.ubuntu.com/releases
[17:23] <sarnold> pcn: .. but upgrading to e.g. 12.04 LTS would be worthwhile..
[17:25] <pcn> Yeah, we use that too, but for testing this particular package shouldn't return a 403
[17:26] <pcn> This just changed in the last day or two.  So is there anyone from canonical who can fix the permissions in s3 for this repo?
[17:29] <genii> pcn: When end-of-life is reached, the repositories are normally frozen and archived to the old-releases URL that was mentioned.
[17:32] <pcn> How's the supposed to work for ec2-clients where that dns points outside of s3?
[17:33] <adam_g> zul,  i think this is totally cheating http://people.canonical.com/~chucks/ca/python-pbr/python-pbr_0.5.11-0ubuntu2~cloud0_source.changes
[17:33] <adam_g> zul, but those look good otherwise
[17:33] <sarnold> pcn: being looked at :)
[17:34] <zul> adam_g:  yeah totally cheating
[17:36] <zul> adam_g:  so ok to upload them to the ppa?
[17:36] <adam_g> zul, ya
[17:37] <zul> adam_g:  cool thanks
[17:38] <pcn> sarnold: thanks
[17:38] <sarnold> pcn: happy to middleman :)
[17:53] <axisys> is there a image that will let me install server OS over the remote serial console ?
[17:54] <axisys> I tried to boot from usb precise amd64 iso while connected from a remote serial console and I do see in the screen.. probably defaulted to vga
[17:55] <axisys> I do not see*
[17:55] <axisys> is there an image that defaulted to ttyS0 , so I can install from remote console
[17:56] <axisys> PXE is not working .. keep getting "PXE-E11: ARP timeout" after DHCP and the answer here saying not to enable option #60 in dhcp http://www.bootix.com/support/problems_solutions/pxe_e11_arp_timeout.html
[17:56] <axisys> I do not think I have that enabled
[17:57] <zul> SpamapS:  ping whats going on with mariadb and percona in debian btw?
[17:58] <SpamapS> zul: MariaDB packaging is ready for review. Nobody has time.
[17:58] <SpamapS> zul: Percona I haven't seen anything going on.
[17:59] <axisys> looks like alternate installer ISO offer to install from remote console
[17:59] <axisys> let me try that
[18:45] <hxm> hello
[18:45] <hxm> can i control a process to use only a % of the cpu?
[18:46] <sarnold> hxm: you can make a process take relatively less of the CPU by setting its nice level; see the nice(1) and renice(1) command. they can't give you percent-of-cpu, but can make hogs more tolerable..
[18:47] <RoyK> hxm: no, but you can renice it
[18:47] <sarnold> hxm: I think I heard groups can give you percent control like that, but I've nevre gone to the effort of figuring it out myself
[18:47] <sarnold> hxm: s/groups/cgroups/
[18:48] <RoyK> hxm: only cpu family I know that allows suballocation of power is IBM POWER
[18:49] <hxm> sobad, i have intel
[18:50] <RoyK> hxm: you can use ulimit, but it it'll just limit usage
[18:50] <RoyK> hxm: better renice those processes
[18:52] <hxm> ok, there i go
[18:53] <RoyK> hxm: what sort of processess are these?
[18:53] <hallyn_> sarnold: pretty much yes,
[18:53] <hallyn_> except if nooen else is competing for cpu, you'll get all the cpu anyway
[18:54] <hxm> bitcoind, when downloading the blockchain (10Gb slowly)
[18:54] <hallyn_> the % is only a guarntee that the rest is available for others, not that you won't get more
[18:54] <hxm> the developer says is only when indexing the database
[18:54] <sarnold> hallyn_: ah! thanks.
[19:13] <adam_g> zul, any idea why this is waiting on subunit + testrepository? https://launchpadlibrarian.net/141174494/buildlog_ubuntu-saucy-i386.quantum_1%3A2013.2~b1-0ubuntu1_MANUALDEPWAIT.txt.gz
[19:15] <zul> adam_g:  looks like the subunit binary is in universe
[19:15] <adam_g> zul, does it actually require subunit or just python-subunit?
[19:16] <zul> i think python-subunit
[19:24] <slicslak> i have a 10.10 server i'm trying to upgrade.  when i apt-get upgrade i get 404 errors: Err http://security.ubuntu.com maverick-security/main Sources 404  Not Found [IP: 91.189.91.13 80]
[19:25] <sarnold> slicslak: maverick has been unsupported for a little over a year; it's about time to upgrade :)  https://wiki.ubuntu.com/Releases
[19:25] <Pici> !eolupgrades
[19:25] <sarnold> cool, thanks Pici :)
[19:26] <slicslak> perfect, thx
[19:36] <pcn> Hi, it looks like sarnold is away at the moment.  Can anyone from canonical comment on how fixing the us-east-1 s3 repo is going for 11.04?
[19:38] <pcn> let me ask: is there a policy/preference from canonical's PoV about clients mirroring the s3 repo via s3 calls?
[19:55] <RoyK> pcn: 11.04 isn't supported anymore AFAIK
[19:56] <Pici> Aye.
[19:57] <pcn> Royk: the distro is not supported, but the repo shouldn't return arbitrary 403s
[19:58] <RoyK> pcn: upgrade
[19:58] <pcn> Are you tring to be helpful?  I covered this earlier: we are upgrading.
[19:59] <pcn> However, not being able to re-launch doesn't actually facilitate the upgrade process.
[19:59] <pcn> How do I e.g. check to see if something is broken in the 12.04 install if I can't confirm it on an 11.04 install?
[19:59] <RoyK> not sure
[20:00] <soy_el_pulpo> pcn: check your logs
[20:00] <soy_el_pulpo> pcn: look for erros
[20:00] <shauno> you'll have to repoint your sources.list to archive.ubuntu.com/ubuntu/dists/natty-security/
[20:00] <soy_el_pulpo> errors
[20:00] <pcn> https://gist.github.com/pcn/5686473
[20:00] <pcn> I'm in s3
[20:00] <pcn> ec2.
[20:01] <pcn> Unless the images provided by ubuntu fix this automagically, this problem will not let anyone install snmpd in ec2 with an 11.04 ami
[20:01] <pcn> Since I'm using an image based on a ubuntu image, i don't think they're fixed automagically, so I'm asking for help
[20:02] <pcn> In fact, everything in that repo is 403'ing.
[20:02] <pcn> Jeez
[20:03] <Pici> You'd need to use http://old-releases.ubuntu.com/
[20:03] <RoyK> pcn: not to be rude or anything, but you should have upgraded earlier
[20:04] <pcn> Are you trying to help?
[20:04] <RoyK> I think Pici is
[20:05] <pcn> I understand, but we covered this ground earlier
[20:05] <pcn> pici not with you, I'm not trying to dismiss your advice
[20:06] <pcn> the issue is that the repo still appears to be there, and was working yesterday and is returning 403s today
[20:06] <pcn> If there is an official "this repo is not working" message somewhere, then I can work with that.
[20:07] <pcn> My understanding is that this is not intentional, so I'm asking about an ETA.
[20:08] <shauno> it looks quite intentional to me.  http://security.ubuntu.com/ubuntu/dists/  doesn't have a natty-security folder
[20:08] <Pici> pcn: you could try poking someone in #canonical-sysadmin , but my frist guess would be that its not supposed to be there and they finally got around to removing it.
[20:09] <pcn> shauno: that's expected - there are no more security releases, and that's fine.
[20:09] <Pici> s/frist/first/
[20:09] <pcn> pici I'll check there, thanks
[20:09] <jdstrand> shauno: use old-releases.ubuntu.com/ubuntu/dists/
[20:10] <shauno> yeah, Pici's covered that.  I'm trying to explain why it's not where he is looking
[20:10] <jdstrand> I see, sorry
[20:11] <jdstrand> it is intentionally not there
[20:11] <jdstrand> no need to ask anyone
[20:13] <pcn> Hmm...
[20:13] <pcn> OK, so for our use, we sould mirror from old-releases.ubuntu.com
[20:14] <pcn> Is there some kind of announcement that the repo is now officially dead?
[20:16] <pcn> Everything here seems to be heresay-ish "you should" and "look here" but not "here's the EOL announcement for that repo"
[20:18] <yofel> pcn: 11.04? https://lists.ubuntu.com/archives/ubuntu-announce/2012-October/000165.html
[20:18] <Pici> pcn: http://ubottu.com/y/natty/
[20:20] <jdstrand> pcn: the EOL announcements go out to ubuntu-announce@ and ubuntu-security@. there is a warning that it will go EOL and then the EOL
[20:21] <jdstrand> pcn: there is not a separate announcement for when the EOL release moves from archive.ubuntu.com to old-releases.ubuntu.com
[20:22] <pcn> jdstrand OK.  The hard part is that for ec2 users, we're not using archive.ubnutu.com, we're using us-east-1.ec2.archive.ubuntu.com which doesn't feel the same
[20:22] <pcn> So that's on me
[20:23] <Pici> The sysadmins don't usually move the releases to old-releases immediately, it can take months.
[20:23] <pcn> Yeah, the technical part of that seems like the really important one to me
[20:23] <maxb> It's the same principle as <countrycode>.archive.ubuntu.com
[20:23] <pcn> maxb: understood, that's on me
[20:23] <pcn> I would really appreciate having the transition be announced a bit more though
[20:24] <Pici> How?
[20:24] <shauno> if it was working for you yesterday, it's taken 10 months to leave your mirror.  that's quite a grace period
[20:26] <pcn> Indeed, I'm not complaining about the grace period.  I'm complaining about the transition not being in google.
[20:27] <pcn> pici: how would be anything that turns up high in a google search, I guess.
[20:27] <pcn> I know I'm sounding whiny but I would like to be able to contribute that this took me by surprise because it appears to be an arbitrary date.
[20:27] <maxb> I suppose the current attitude is that once it's EOLed, there's no obligation to announce further
[20:29] <yofel> pcn: the date was end of October 2012, after that it may stop working on any arbitrary date
[20:29] <shauno> that's how it strikes me too.  you're told your support path from that point is to upgrade to $next, and then moving to old-releases (so mirroor providers don't need to find a home for 19 releases) isn't a support issue, it's a housekeeping issue
[20:31] <pcn> Sure, it's defensible.  However, it's poor busines practice
[20:31] <pcn> Specifically, it's bad communications.
[20:31] <pcn> Hopefully I haven't said that I expect anything more than "if it's gone, tell me it's gone somewhere, somehow"
[20:32] <yofel> pcn: https://wiki.ubuntu.com/Releases shows end of live dates, please upgrade before then
[20:32] <yofel> *end of life
[20:39] <pcn> yofel: again I understand this is completely within the description of the service that canonical is providing.
[20:40] <pcn> yofel: I'm saying that communicating that it's going away on a date is a really good idea, too.
[20:41] <chilicuil> do you know of any solution to test if a network is fast enough to install voip telephony?, I'm looking at http://myspeed.visualware.com/indexvoip.php
[20:44] <pcn> chilicuil: iperf can be used to test performance between two points that you control, but throughput across the internet can be iffy
[20:45] <chilicuil> pcn: I only control 1 side, the datacenters from where we offer the sip service, so I doubt I'll be able to use iperf in the windows machine clients, I'll check iffy out, thanks for the suggestion =)!
[20:45] <hallyn_> zul: btw, debian Bug#710537: libvirt-bin: /etc/libvirt/qemu.conf shouldn't be world-readable.  we might also care.
[20:47] <pcn> chilicuil: iffy isn't a tool, sorry I mean it may be variable/no guarantees unless you get a provider that e.g. provides tunneling (mpls for ex) but that doesn't sound like what you're doing
[20:49] <chilicuil> pcn: ohh!, got it! =), then I'll look further
[20:50] <pcn> chilicuil you may want to try mtr to get a feel for what the path between you and the other end looks like, and watch to see how variable it is
[20:51] <chilicuil> pcn: yep, I'll probably build a web interface for common network utilities if I don't find anything