[00:03] <sarnold> h0mer_: sure, it's just after the end of day for most of europe and america..
[00:04] <h0mer_> I understand that.  I just figured people would be able to help after as well.  I guess I'll try again tomorrow
[00:06] <sarnold> folks come and go, you might as well ask now, and see if anyone who's around now knows
[00:08] <h0mer_> understood
[00:09] <h0mer_> Alright so I'm trying to install the cannonical openstack/ubuntu distribution using the Maas/Landscape/openstack instructions on the Ubuntu website.  Problem is, I can't seem to get the networking to work correctly.  I have two networks (192.168.20.x and 192.168.30.x), the first one has DHCP enabled and the second one doesn't
[00:09] <h0mer_> Problem is when I install everything (landscape/openstack), it seems like the metadata service doesn't start up correctly and connect ping out to the internet
[00:09] <h0mer_> I'm not sure how to start debugging this.  So I was wondering if anyone had any tips?
[00:10] <h0mer_> And just as a side note, anytime I start up a VM instance in the Ubuntu Openstack dashboard, I get an issue on the VM startup where the networking config says something to the effect of "waiting 120 seconds for network config"  Could anyone point me in the right direction to start debugging this issue?
[00:12] <sarnold> h0mer_: that second network.. is that for second NICs in all the systems? or is that for guests?
[00:12] <h0mer_> for the second nics in all the systems.  (which would be for the guest vm's subsequently)
[00:12] <h0mer_> the first nic is connected to the .20 network and the second nic is connected to the .30 network
[00:13] <sarnold> aha
[00:13] <sarnold> does keystone catalog show you anything useful?
[00:13] <h0mer_> nope
[00:14] <h0mer_> juju debug shows nothing either
[00:15] <h0mer_> just out of curiousity, should the second netowkr (the .30 network on which all the vm's get their IP's from) have DHCP setup in MAAS?  Should Maas be managing the DHCP/DNS for that network as well?
[00:18] <sarnold> I don't think so, I think maas should only know about the physical machines
[00:21] <h0mer_> Right, but in the Maas network web page it shows both networks and I set up maas to only manage the first one (192.168.20.x) which seems to work fine as Openstack thinks it's the internal management network.  The second network is also listed in the Maas network's page, but is not managed by Maas, but is used as the ip allocations for the guest VM's in openstack
[00:22] <h0mer_> i can ping the VM's, but I cannot log into the machines using putty because the meta data service cannot seem to ping out to the internet.  I'm thinking it's because the JUJU instance on the machine where nova is installed doesn't have both ETH0 and ETH1 set up correctly?
[00:22] <sarnold> h0mer_: I _think_ guest addresses ought to be managed via neutron?
[00:23] <h0mer_> yea thats why I didn't let Maas manage the .30 network
[00:23] <h0mer_> figured it would be setup by neutron (DHCP)
[00:23] <h0mer_> and I set it to use the gateway 192.168.30.1 (which can ping out to the internet)
[00:23] <sarnold> .. in the sense that e.g. every individual tenant ought to be able to have their own e.g. 192.168.1.1 ...
[00:24] <sarnold> gah. just when I thought I was starting to get the hang of all these pieces. :/
[00:25] <h0mer_> problem is the first machine that the Landscape installer bootstraps JUJU onto has two NIC's, but if I juju ssh into that node and do a "ifconfig" I don't see the set up for the second NIC, only one network is setup
[00:25] <h0mer_> iface eth1 inet dhcp  # Primary interface (defining the default route) iface eth0 inet manual  # Bridge to use for LXC/KVM containers auto juju-br0 iface juju-br0 inet dhcp     bridge_ports eth0
[00:25] <h0mer_> crap that didn't come out the way I wanted it to
[00:25] <h0mer_> guess I can't just copy paste from the terminal
[00:35] <h0mer_> just let me know if you get any ideas.  I'll be chilling here trying to debug this.  Thanks for your help btw.
[00:36] <sarnold> h0mer_: there's also a #maas and a #juju that might be helpful, but I don't know which would be a better fit..
[00:37] <h0mer_> cool let me go check those out.  THanks.
[04:46] <jak2000> hi all, exist a ubuntu server 32 bits?
[04:49] <sarnold> jak2000: see e.g. http://www.ubuntu.com/download/alternative-downloads
[04:51] <jak2000> thanks
[06:32] <Abhijit> Hi.
[06:33] <Abhijit> I have setup postfix only. I have not setup dovecot yet. will do that later. right now I can send email from my ubuntu 15.04 postfix to my real gmail and it works. however when I send email from my real gmail to this ubuntu postfox i do not get it in postfix. gmail do not give any send error. my domain is test.com its TX mail.test.com points to mailtest.cloupapp.net and in that i have this ubunt postfix
[06:33] <Abhijit> how to troubleshoot this?
[06:34] <sarnold> check your logs?
[06:35] <sarnold> make sure you can connect to your mail server from other machines off your network?
[06:35] <sarnold> can you send mail to your system by hand?
[06:36] <Abhijit> its not in my LAN. its on azure.
[06:36] <Abhijit> hence I used gmail to send email. which is not working.
[06:36] <sarnold> does azure have anything like amazon's security groups? do you need to open network ports?
[06:36] <Abhijit> I will check that.
[06:37] <Abhijit> sarnold, so on godaddy in my test.com dns settingsn i only changex mail @ record to mailtest.cloudapp.net. which is a azure ubuntu vps. thats all I need for it to work right?
[06:37] <Abhijit> dig shows that MX record for test.com is mailtest.cloudapp.net
[06:37] <Abhijit> priority as 0
[06:38] <sarnold> Abhijit: one thing at a time.. can you nc to your MX and send mail by hand?
[06:38] <Abhijit> I dont know what is nc. My postfix can send email to my gmail.
[06:39] <sarnold> Abhijit: nc is netcat; it's like telnet, but doesn't try to interpret the data the same way telnet does
[06:39] <sarnold> Abhijit: that's sending; you're trying to troubleshoot receiving, right?
[06:39] <Abhijit> right.
[06:39] <sarnold> Abhijit: so nc mailtest.cloupapp.net 25 and try to send an email by hand
[06:40] <Abhijit> trying...
[06:41] <Abhijit> NCat Connection Timed out.
[06:41] <Abhijit> sarnold, ^
[06:41] <sarnold> hey, there we go :) connection timed out usually means that there's a firewall DROPping packets, rather than REJECTing or ACCEPTing packets
[06:42] <sarnold> Abhijit: so, you've got to figure out if it's a firewall on the host, or a firewall run by azure, that's blocking your packets
[06:42] <Abhijit> ok. I will talk to my azure sysadmin regarding this.
[06:42] <Abhijit> sarnold, sure. Thanks for help.
[06:42] <sarnold> amazon calls them security groups (stupid name, I know..) -- hopefully azure's docs will help you find the similar thing with the name..
[06:43] <sarnold> have fun Abhijit :)
[06:43] <Abhijit> yeah.
[07:27] <neonixcoder> Hi team, I am trying to upgrade my machine with do-release-upgrade -f DistUpgradeViewNonInteractive, I am curious to know what other arguments I can pass with -f option?
[07:28] <neonixcoder> When search online some other day I see many arguments with -f, but unable to find it now..
[07:28] <neonixcoder> Any thoughts on this?
[08:14] <lordievader> Good morning.
[08:15] <ObrienDave> waves
[09:57] <killall> Hello my /etc/hosts freaks out when i insert some ips
[09:57] <killall> 17X.221.34.139 my.domain.com
[09:57] <killall> and
[09:57] <killall> #17x.221.34.171 my.domain.com does not work
[09:57] <ewook> 17X.221.34.139 isn't an ip.
[09:58] <killall> yes i now the X is a number ;)
[09:58] <ewook> and #something is a comment.
[09:58] <killall> but the weird is some ips work and others dont
[09:58] <ewook> what are you editing with?
[09:59] <killall> ewook, ok
[09:59] <killall> 176.221.34.171  www.domain.pt domain.pt
[09:59] <killall> this one does not work
[09:59] <ewook> IP_address canonical_hostname [aliases...]
[10:00] <ewook> 192.168.1.10    foo.mydomain.org       foo
[10:00] <ewook> do a man hosts
[10:00] <ewook> and look at the examples. your IP www.domain.pt domain.pt is not correct.
[10:01] <killall> ewook,  why does the  "176.221.34.171  www.domain.pt domain.pt" does not work and " 176.221.34.139  www.domain.pt domain.pt" works
[10:01] <ewook> first come first serve.
[10:01] <ewook> would be my guess.
[10:01] <killall> that is the only ip there
[10:02] <killall> no more ocurrencys of it
[10:02] <ewook> separate them into two lines.
[10:02] <ewook> IP www.domain.pt
[10:02] <ewook> IP domain.pt
[10:03] <lamont> killall: how does it "freak out"?
[10:03] <lamont> ewook: multiple names on the same line with one IP is just fine
[10:03] <killall> lamont, ** server can't find domain.pt: NXDOMAIN
[10:03] <ewook> lamont: second argument is supposed to be an alias.
[10:03] <ewook> lamont: so yes, it should work- but duplicate aliases?
[10:04] <lamont> NXDOMAIN never comes from /etc/hosts that I've ever seen, always from the DNS... what cmmand is telling you that?
[10:04] <killall>  nslookup cnt.calmetric.pt
[10:04] <killall> that is the domain i want to resolve
[10:04] <lamont> nslookup (on linux, at least, last I looked) says 'please ignore /etc/hosts and query the DNS, kthx"
[10:05] <lamont> and should never be confused with gethostbyname(3)
[10:05] <killall> when it has in hosts the ip ending in 171 it hgives nxdomain but when it has the 139 ending ip it works nice
[10:05] <lamont> strace the nslookup and see if it even opens /etc/hosts?
[10:05] <jamespage> rbasak, hey - can you point me at the dpdk stuff again
[10:06] <lamont> killall: that's... strange
[10:06] <killall> lamont,  yes it is
[10:06] <lamont> ** server can't find cnt.calmetric.pt: NXDOMAIN
[10:06] <rbasak> jamespage: you mean the work in progress? https://launchpad.net/~smb/+archive/ubuntu/dpdk
[10:06] <jamespage> rbasak, thanks
[10:06] <jamespage> that's what I was after
[10:07] <lamont> killall: nslookup on my 15.04 system never opens /etc/hosts
[10:08] <killall> lamont, on mine either
[10:08]  * lamont tends to use ping when he wants to see address resolution involving /etc/hosts
[10:08] <lordievader> How about 'host domain.pt'?
[10:08] <killall> but the strange is works with one ip and not with other, the same happens in firefox the one that nslookup resolvs opens ncie
[10:08] <lamont> if it was me, I'd be looking at the differences in strace in the working and non-working cases
[10:08] <killall> *nice
[10:09]  * lamont has to run
[10:09] <killall> Host cnt.calmetric.pt not found: 3(NXDOMAIN)
[10:09] <killall> the same i will change the ip
[10:10] <lordievader> cnt.calmetric.pt != domain.pt?
[10:11] <killall> yes the domain.pt was a test the one in hosts is cnt.calmetric.pt
[10:11] <killall> and should resolve but no :(
[10:14] <killall> i dont get it one ip works and other no
[10:15] <killall> thanks :)
[10:15] <killall> ill try to reboot pc and crack this down ;)
[10:22] <jamespage> rbasak, smb: urgh - all of the include header files for dpdk assume  they are in the root of /usr/include (not the subdirectory)
[10:29] <jamespage> rbasak, smb: ovs patched to deal with that
[10:44] <jamespage> rbasak, smb: are you targetting a specific CPU feature baseline for dpdk?
[10:47] <blob_> I'm trying to setup a kiosk PC based on Ubuntu server with nodm, openbox and network manager. I'm having trouble with network manager permissions to allow the logged in user to manage connections using nm-applet. I've been searching and I think the problem is caused by polkit or dbus permissions. If run nm-applet with sudo it works. With a normal user it does not work. If I just try to use nmcli from the terminal I get the foll
[10:50] <blob_> If I copy the contents of <policy at_console="true"> to <policy context="default"> in /etc/dbus-1/system.d/org.freedesktop.NetworkManager.conf I can run "nmcli nm", so maybe something is wrong with my session? Even with the modification I can't still configure any interfaces, but at least it may give someone an idea where I should try to look into next?
[10:52] <rbasak> jamespage: the plan is for SSE3 only although we want to talk to upstream about runtime detection
[10:52] <jamespage> rbasak, the ovs configuration assumes SSE3
[10:53] <rbasak> OK. AIUI we can't even build without SSE3
[10:54] <jamespage> rbasak, does that break any cpu compat baseline generally in the archive?
[10:54] <rbasak> jamespage: yes. There exist amd64 CPUs without SSE3 AIUI.
[10:55] <ivoks> hey guys, do you know where is sentiment analysis charm?
[10:55] <rbasak> But currently the build requires SSE3 due to inline assembly I believe, so we have no choice (without taking on the porting)
[10:55] <jamespage> ivoks, ask samuel in #juju
[10:55] <ivoks> jamespage: thanks!
[10:57]  * jamespage ponders about how we enabled this effectively in ubuntu
[11:09] <jamespage> rbasak, well I have it building
[11:09]  * jamespage ships it
[11:09] <jamespage> lol
[11:09] <jamespage> anyway - a start
[11:10] <jamespage> https://launchpad.net/~james-page/+archive/ubuntu/ovs-dpdk/
[11:14] <jamespage> rbasak, do you think the runtime detection will happen in the near term?
[11:16] <rbasak> jamespage: seems unlikely
[11:32] <smb> jamespage, You can work around the headers issue by having a -I that points to your headers top. So they can be in /usr/include/dpdk. We build for default machine type which is sse3 only
[11:33] <smb> jamespage, are you using the lib from my package?
[11:44] <rbasak> jpds: do you have any plans to update strongswan?
[12:03] <teward> rbasak: ping, i have a question for you, any idea if the TB is meeting today or should I just lurk both meeting channels?
[12:04] <rbasak> teward: they're meeting today
[12:04] <teward> about the same time as the server team?
[12:04] <rbasak> teward: add the nginx item to https://wiki.ubuntu.com/TechnicalBoardAgenda I guess?
[12:04] <rbasak> teward: yeah, same time.
[12:04] <rbasak> teward: #ubuntu-meeting-2
[12:06] <teward> rbasak: i'll lurk the channel, my guess is they may bring it up, and if not i'll poke someone on the TB for next meeting - i don't like last-minute agenda items and i bet they don't either
[12:07] <teward> since they have a "scan the ML for things they missed" item
[12:07] <YamakasY_> I have a serious issue with puppet, anyone knows this ? https://tickets.puppetlabs.com/browse/PUP-3501
[12:18] <rbasak> YamakasY_: not seen it before. I don't think Puppet should be reading system configuration files assuming an encoding set by the locale defined in the environment. That doesn't make sense. It should use what encoding the file is defined to have.
[12:19] <rbasak> In the case of upstart jobs, it doesn't seem to be specified, but UTF-8 would be a reasonable assumption.
[12:20] <YamakasY_> rbasak: yes indeed, I have this bug already for months and everyone thought I was nuts
[12:21] <YamakasY_> rbasak: so, /etc/environment should be ok and for shell in the init script ?
[12:38] <jamespage> smb, I am yes
[12:39] <smb> jamespage, Ah, ok. I put some help for getting builds without their environment into README.debian though I guess that is too late now
[12:40] <YamakasY_> rbasak: ?
[12:44] <rbasak> YamakasY_: setting a UTF-8 locale in /etc/environment is a sensible thing to do anyway, and might work around that bug making it a low priority in terms of actual user impact.
[12:45] <rbasak> YamakasY_: but I think it's a bug that puppet assumes that system configuration files are encoded in the user's environment locale encoding. Instead it would use the encoding defined for the file its reading
[12:45] <rbasak> it's
[12:45] <YamakasY_> rbasak: ok, weird thing is, when I reboot the machine it gets the error, but when I restart it from shell with LANG in my init script it works again
[12:53] <YamakasY_> rbasak: so what would you set ? the puppet guys say it's always someone else his problem ;)
[13:02] <rbasak> YamakasY_: if it were me? Given it's puppet, I'd just modify the configuration file to be US-ASCII to work around the problem, and make the point in the bug that it is really a bug since the file encoding is not defined to be the same as the environment defined locale encoding.
[13:03] <YamakasY_> rbasak: I agree, bunch of morons they are :P
[13:03] <rbasak> The chef bug they refer to is a little different since it's a chef template so chef get to define the expected encoding.
[13:03] <YamakasY_> yap!
[13:03] <YamakasY_> but they don't say that ;)
[13:03] <YamakasY_> puppet guys always think they are right
[13:04] <rbasak> YamakasY_: mind your conduct please. http://www.ubuntu.com/about/about-ubuntu/conduct
[13:04] <YamakasY_> rbasak: no but this is known for months and they don't admit it
[13:04] <rbasak> I think they deserve the benefit of the doubt. At least put the argument to them clearly in the bug.
[13:04] <YamakasY_> it's annoying as all kinds of fixe's don't fix it
[13:05] <YamakasY_> no matter what I do, every reboot gives the same issue
[13:05] <YamakasY_> I need to restart puppet manually to get UTF-8
[13:17] <FunnyLookinHat> Are any of you familiar with uvt-kvm?  I can't figure out how to restart a kvm that I shutdown.
[13:17] <YamakasY_> rbasak: it seems undoable
[13:19] <rbasak> FunnyLookinHat: virsh start <name>
[13:19] <FunnyLookinHat> rbasak, thank you!
[13:19] <FunnyLookinHat> Was trying to find a uvt-kvm command and none seemed to exist  :)
[13:23] <jamespage> smb, meh - it was pretty easy to figure out - tiny patch to ovs to make it work
[13:23] <jamespage> smb, I'll probably work on some sort of auto-detection as well
[13:24] <jamespage> so --with-dpdk just works
[13:25] <smb> jamespage, Yeah, ideally there would be a pkg-config info file... though not (yet) from upstream
[13:27] <YamakasY_> rbasak: any other options ?
[13:29] <rbasak> YamakasY_: can you not just modify the file to be US-ASCII?
[13:31] <YamakasY_> rsalveti: which file, the pp ?
[13:31] <rbasak> YamakasY_: /etc/init/php-fpm.conf or whatever file is causing the issue
[13:32] <YamakasY_> rbasak: the php5-fpm restart
[13:32] <YamakasY_> rbasak: would not be nice
[14:00] <teward> rbasak: sarnold: in other news, *actually usable debug data in nginx bugs now!*  https://bugs.launchpad.net/ubuntu/+source/nginx/+bug/1476656, https://launchpadlibrarian.net/212180277/SystemctlStatusFull_Nginx.txt.txt, E:NotABug
[14:01] <rbasak> teward: \o/ good job!
[14:03] <ws2k3> i have setup an ldap server. but when my ldap server is offline i cannot login in my ubuntu servers anymore how can i resolve this issue?
[14:04] <teward> rbasak: I'm going to Invalid the -5ubuntu3 ones, and say to file a new bug if an upgrade or install to -5ubuntu3.1 triggers the issue
[14:05] <teward> rbasak: but we NOW KNOW that 1476656 is user/server config error - look at that bug's DpkgHistoryLog
[14:05] <teward> rbasak: they installed Apache first, and probably didn't shut it off
[14:05] <teward> but seriously, USABLE DEBUG DATA!  *dances a happy dance, before deciding he needs coffee and goes off to find some*
[14:11] <teward> rbasak: pretty certain we as the server team may need to issue a notice in Release Notes that the default Apache and default Nginx configurations can't be run side-by-side because they are designed to 'just work' out of the box without any additional configuration.
[15:08] <thebwt> howdy guys, I do a small repo of custom deb packages, and I'm having a rough time getting unattended upgrades to update our repo. Anyone got a good article on repos, their metadata andhow that interacts with everything? we've poured over man pages but I was hoping there was a good flat article out there that someone may know of.
[15:08] <iliv> Does do-release-upgrade select the right mode automatically?
[15:59] <teward> so, general question, anyone got any insights on https://bugs.launchpad.net/nginx/+bug/1476296 per chance and what init file should be in place?
[17:00] <Daviey> teward: Nicely done, RE: TB /Nginx
[17:01] <teward> Daviey: thank you!
[17:01] <teward> rbasak had a hand in it, i lagged briefly and he was able to chime in as well
[17:01] <Daviey> :)
[17:01] <teward> Daviey: initially looked like they were "Why is this here" but I did refer to your suggestion referring it to TB
[17:02] <teward> I honestly didn't expect a decision today, i just wanted it to be looked at as soon as possible because of the Wily impact
[17:03] <teward> given the vote and the conditions set forth i'm fairly comfortable doing a merge of 1.9.x to Wily, following through up to Feature Freeze, and leaving it, merging again the moment X is open.
[17:03] <teward> provided that's still the general agreed-upon track
[17:04] <teward> which it should be based on rbasak, jcastro, yourself, and sarnold
[17:04] <teward> and me :)
[17:04] <Daviey> teward: If you need a hand with sponsorship, let me know.
[17:04] <teward> Daviey: i have upload rights, so sponsorship isn't a huge concern anymore
[17:04] <Daviey> (I'm sure rbasak will be happy to aswell)
[17:04] <jcastro> o/ awesome work fellas
[17:04] <teward> Daviey: i'd be glad for a sanity check though
[17:04] <Daviey> Oh! I forgot that.. nice
[17:04] <teward> if you're offering :)
[17:05] <teward> jcastro: thank you!
[17:05] <jcastro> did we send you an ubuntu server team shirt or something yet?
[17:05] <rbasak> Daviey: we'll still need review from the release team before executing the FFe during X development if you'll be around to help with that please :)
[17:05] <teward> ^ that
[17:05] <teward> jcastro: not yet, i never translated the US shirt size I wear to UK shirt sizes for that to be sent my way
[17:06] <teward> got a tad busy with school, work, and other things :)
[17:06] <jcastro> ok, lmk. It's easy to translate, just go up one size or two.
[17:06] <jcastro> though it can seem that a UK Small is a US XXXXXL.
[17:07] <teward> heheheheh
[17:07]  * rbasak still has a XXL Heroku shirt or something that doesn't fit him
[17:07] <rbasak> I never really understood what's going on there
[17:07] <teward> heh
[17:08] <teward> jcastro: if you want to do the conversion, I wear US sized XL due to being taller than average so large shirts don't fit me xD
[17:08] <jcastro> ok, just mail me your shipping adress and I'll sort it for you, jorge@ubuntu.com
[17:09] <teward> will do, that'll be coming in now
[17:12] <teward> jcastro: the confusing part will be the estimated date of arrival - if you estimate it to arrive after August 7 I have to have you mail to one address, or another if before August 7
[17:12] <teward> :/
[17:13] <teward> i hate complex schedules
[17:13] <teward> shoudl be in your inbox now
[17:13] <teward> I'm going to get something to eat, then rebase the merge diff
[17:17] <sarnold> teward: yes! that's awesome :)
[17:20] <teward> sarnold: indeed!  This is definitely a plus, and thanks to the TB.  The apport hooks in there WORKING for postinstallation failed bugs though is ALSO a plus
[17:20] <teward> because now we know that some people fail at configurations
[17:20] <teward> or fail to understand that both Apache and nginx have 'just works' defaults
[17:20] <teward> rather than
[17:20] <teward> 'designed to run side by side and not cause a multitude of infinite other problems'
[17:20]  * teward shrugs
[17:28] <sarnold> teward: now you just need to keep that text handy for easy copy-n-paste :)
[17:41] <teward> sarnold: heh
[20:29] <med_> smoser, your ears should be burning
[20:29] <med_> how does LXD get config drive data?
[20:29] <med_> and does cloud-init handle a "pre-mounted" filesystem
[20:30] <med_> ^ utlemming
[20:30] <smoser> well the second bit is not reallycloud-init. its systemd or upstart. and that is just part of booting in lxc (or lxd). so thats worked for quite a while.
[20:30] <med_> nod thanks
[20:30] <smoser> the first part is new, and is in the works.
[20:30] <smoser> currently, lxd will seed the nocloudnet datasource
[20:30] <med_> nova would like to NOT support XFS mounted drive
[20:31] <med_> (that parallels pushed)
[20:31] <smoser> (this is how lxc-clone does in old lxc also)
[20:31] <smoser> in the hopefully not distant future, cloud-init will read data from /dev/lxd
[20:31] <med_> nod.
[20:31] <smoser> parllels pushed to nova-what ?
[20:32] <med_> apparently s/xfs/ext43
[20:32] <med_> sorry, using ext4 instead of iso9660
[20:32] <med_> for cloud-init data
[20:32] <smoser> as they want the libvirt driver to do it ?
[20:32] <med_> bindmount is being proposed
[20:32] <smoser> ah. i see.
[20:32] <smoser> for config drive.
[20:32] <smoser> right. ok.
[20:33] <med_> not sure libvirt even comes into it.... and very sorry I conflated xfs/ext4...
[20:33] <smoser> so yeah, we talked about this for lxc also
[20:33] <med_> yes, config drive
[20:33] <smoser> er.. for lxd.
[20:33] <med_> that's kind of why I poked you... we heard lxd was interested in the same ext4 approach
[20:33] <med_> s/we/openstack nova mid cycle/
[20:33] <smoser> whatnova should do is jus create the data and prior to start put it in a well known path
[20:33] <med_> perfect, that's what's proposed now
[20:34] <med_> mikal ^
[20:34] <smoser> it could attach a disk, and then cloud-init would actually try to work
[20:34] <smoser> but the kernel wouul (hopefully) NACK its attempt to mount the /dev/XXX device
[20:35] <smoser> ubuntu is looking to make mounting of some filesystems in a user namespace actually safe
[20:35] <med_> ah.
[20:35] <med_> that's good too.
[20:35] <smoser> that would allow you to actually attach a block device and safely mount it
[20:35] <smoser> from inside the container (cinder)
[20:35] <smoser> as without that, you're stuck inside the container with fuse.
[20:35] <smoser> and iirc thats even only on ubuntu kernel
[20:36] <med_> nod
[20:47] <mikal> Hi, coming to this late because I was being ragey in person
[20:47] <mikal> So, parallels merged support for a loop mounted ext4 config drive into the libvirt driver
[20:47] <mikal> Which we didn't love
[20:47] <mikal> So we're reverting it
[20:48] <mikal> The new plan is to allow the ocnfig drive code to generate to a specified path, which can then be added into the contianer
[20:48] <mikal> i.e. no loop mount required
[20:50] <Luke> is there an ubuntu dev channel?
[21:08] <TurBoss> Hi
[21:09] <TurBoss> I have a headless server running 14.04.01
[21:09] <TurBoss> today it refuses to boot network so i put a livecd and chrooted into
[21:09] <TurBoss> can't ping can't ssh
[21:10] <TurBoss> server is on remote place
[21:10] <TurBoss> i can only use a livecd to troubleshot
[21:11] <TurBoss> /etc/netowork is fine /etc/hosts is fine /etc/hostname etc...
[21:11] <TurBoss> i can ping the internet on the chrooted enviroment but when i boot from hdd ir refuses to give netowerk acces
[21:11] <TurBoss> what i can do to solve my problem
[21:12] <TurBoss> so i'm on chroot via ssh
[21:12] <TurBoss> that works
[21:12] <TurBoss> :)
[21:59] <TechIsCool_> I have a question about net-snmp-config I can't seem to find it on ubuntu
[21:59] <TechIsCool_> where is it located
[22:02] <tarpman> TechIsCool_: the search tool on packages.ubuntu.com suggests it's in libsnmp-dev
[22:03] <TechIsCool_> tarpman: Awseome thank you
[22:05] <teward> Luke: if you're still looking for the Ubuntu Devel channel, visit #ubuntu-devel
[22:51] <teward> because i'm tired, would anyone like to sanity-check an nginx merge debdiff for me before I bother uploading it, please?  Sanity checks are nice...
[22:51] <teward> might ask in -devel if no response here :)
[22:52] <Daviey> teward: Would rather look at it with fresh eyes tomorrow...
[22:52] <teward> Daviey: works for me
[22:52] <teward> i don't plan on uploading tonight anyways :0
[22:52] <teward> Daviey: For the record, https://bugs.launchpad.net/ubuntu/+source/nginx/+bug/1476811 is the bug
[22:52] <teward> there's two debdiffs there - one's the actual merge, one's the 'compare to debian' sanity-check :P
[22:54] <teward> which looked sane to me, but eh
[22:54] <teward> we're both tired so meh :P
[22:54] <teward> rather have it be sane than fail
[22:54] <teward> i've done that before too, fail with uploads
[22:57] <Daviey> teward: Yeah, having both makes sense... as a reviewer, i tend to prefer debian->ubuntu or ubuntu->ubuntu if it is smaller.
[22:57] <teward> Daviey: i always include both when it's a merge from Debian
[22:57] <Daviey> Had a quick look, source_nginx.py seems missing from declared remaining changes?
[22:57] <teward> Daviey: already in?
[22:58] <teward> changelog 1.6.3-5ubuntu4
[22:58] <teward> already uploaded to wily :)
[22:58] <teward> don't forget - it's not in Debian
[22:58] <teward> oo shit right i forgot in remaining chagnes
[22:58] <teward> i'm tired
[22:58] <teward> :)
[22:58]  * teward respins for the changelog
[22:58] <teward> i keep copies of my merge dir lol
[22:59] <teward> Daviey: thanks for that, i almost forogt o.O
[22:59] <Daviey> teward: The rest of it looks tidy, but i'd rather look with non-sleepy eyes
[23:00] <Daviey> And on that note, nn.
[23:00] <teward> good night, i'll update.
[23:14] <jak2000> why i cant install ubuntu64 on a vmware ? i gave to this pc: 40gb hdd and 2 gb ram,  and say:  http://postimg.org/image/miqvfxcoz/
[23:15] <teward> jak2000: you have to turn on virtualization in the host's BIOS
[23:15] <teward> you should read the numbers it specifically says there
[23:15] <sarnold> jak2000: reboot your computer; smack f1 and del keys to try to get into the bios. look around for "Virtualization extensions" or "VT" support, and turn that to "Enabled", save, and reboot
[23:15] <teward> ^ that
[23:15] <jak2000> need upgrade the vios of the pc?
[23:15] <teward> jak2000: not usually
[23:15] <sarnold> probably not
[23:16] <teward> jak2000: usually you just have to "Enable" a feature that's present already but 'disabled' in the bios
[23:16] <teward> it's a configuration value
[23:16] <teward> not an update/upgrade/bios reflash
[23:16] <jak2000> ups i am remotely
[23:16] <teward> then you're stuck
[23:16] <bekks> jak2000: So use the remote interface of your server.
[23:16] <teward> ^ that though is an option :P
[23:16] <jak2000> its a normal pc
[23:17] <jak2000> with 8gb ram
[23:17] <sarnold> you can probably change it with lights-out or ipmi or whatever
[23:17] <teward> except if it's a plain old desktop
[23:17] <teward> (might not be present)
[23:17] <sarnold> oh. call your hoster?
[23:17] <jak2000> no no old desktop it have win8
[23:17] <sarnold> they may have a "remote hands" service
[23:17] <jak2000> :)
[23:17] <jak2000> mmm no way
[23:17] <jak2000> not know about pcs,
[23:17] <teward> jak2000: 'plain old' is a phrase - it means it's a standard desktop that most end users have
[23:17] <jak2000> thanks
[23:17] <teward> rather than 'old desktop'
[23:17] <teward> it's not an actual indicator of the age
[23:18] <jak2000> :)
[23:18] <teward> now, me saying "This old-as-sin POFS server is evil.", well...
[23:18] <sarnold> jak2000: "turn on VT" ought to be within their skills, regardless of hardware..
[23:18] <jak2000> ok
[23:18] <jak2000> i am on  my laptop
[23:19] <jak2000> restarting
[23:19] <bekks> jak2000: How would that help you? :)
[23:19] <sarnold> wait..
[23:19] <sarnold> is your vm host remote or local?
[23:19] <teward> jak2000: we don't mena on your laptop we mean on your desktop
[23:20] <teward> jak2000: you need to make the change at the DESKTOP that's running VMware Player
[23:20] <teward> NOT on your Laptop
[23:20] <teward> ... i bet he quit
[23:20] <teward> and we'll have to wait
[23:20] <jak2000> teward yes i understand
[23:20] <jak2000> but for test, and see the option
[23:20] <teward> jak2000: so which syste are you restarting
[23:20] <sarnold> different bios vendors will put it in different locations
[23:20] <teward> ^ that
[23:20] <teward> and different manufacturers do so to
[23:20] <teward> and can rename it to something similar but different
[23:21] <teward> (Dell called it "Enable virtualization extensions" in the Latitude E6500's BIOS, it's different in the latest BIOS versions and such)
[23:35] <jak2000> i have a laptop and see this option disabled:
[23:36] <bekks> The option on your laptop is irrelevant.
[23:36] <teward> ^ that
[23:36] <teward> jak2000: it's IRRELEVANT what it's set to on your laptop
[23:36] <bekks> The option on your desktop pc is important.
[23:36] <jak2000> Intel Virtual Technologiy
[23:36] <teward> jak2000: you need to look at the BIOS on the DESKTOP
[23:36] <jak2000> ok
[23:36] <sarnold> jak2000: it's usually disabled by default: https://en.wikipedia.org/wiki/Blue_Pill_(software)
[23:36] <teward> and it's USUALLY disabled by default
[23:36] <teward> (most end users aren't doing virtualization, and the ones that are know how to power-user the BIOS to make it work)
[23:37] <jak2000> oki