[00:31] <sudormrf> hey guys
[00:31] <sudormrf> anyone around for a very quick question?
[00:32] <sudormrf> not asking to ask, just checking to see if anyone is around :D
[00:32] <sarnold> sudormrf: 420 people in the channel. surely one or two are bots..
[00:32] <sudormrf> heh
[00:34] <sudormrf> ok, so here it goes.  I think I know the answer already, but I just want to verify.  I would like to mount all the shares from one of my servers on another server.  But instead of mounting each share individually, I would like to mount it at the top level, so I can just rsync all of the subdirectories rather than doing it share by share.  Is this possible?
[00:34] <sarnold> sudormrf: I don't think it's possible with NFS / SMB / CIFS
[00:35] <sarnold> sudormrf: no idea about afs or other network filesystems..
[00:39] <wiredfool> just upgraded a server from lucid -> precise -> trusty. Somewhere in that, one of the kvm guests is now failing to boot because app armor is denying it access to its disks
[00:39] <wiredfool> It looks like the profile_load operation is happening after the operation open in the logs
[00:40] <wiredfool> I've checked the /etc/apparmor/libvirt-[uuid].files  and it appears correct
[00:40] <wiredfool> other kvm guests on the same machine are working properly.
[00:40] <sudormrf> sarnold, didn't think so.  I suppose instead of doing a pull I could do a push
[00:40] <sudormrf> that would have probably been the smarter thing to do
[00:40] <wiredfool> ideas?
[00:40] <sarnold> wiredfool: can you pastebin your DENIED messages?
[00:41] <wiredfool> http://pastebin.com/qEY8ni6U
[00:42] <sudormrf> sarnold, question: it "appears" possible when using a GUI.  meaning, in a gui you will see the network object under "network" and clicking on that brings you to the top level.  not sure how that is actually working.
[00:43] <sudormrf> notice the quotes around appears.  all of my experience has been that it isn't possible.
[00:43] <sarnold> wiredfool: try adding "/var/lib/vm/** r," to your /etc/apparmor.d/usr.lib.libvirt.virt-aa-helper
[00:43] <sarnold> sudormrf: I presume those gui things aren't actually making mounts so much as using gvfs or kdeioslaves to do the networking
[00:44] <sudormrf> sarnold, ah! now that makes sense.  I figured it wasn't acutally making the mount (and that is backed up by when you actually open one of the shares it then mounts that share).  interesting to know that is how it is working :D
[00:45] <wiredfool> sarnold: I bet it's actually the final lines that are the difference, where there's an enumeration of extensions for disks, and .raw isn't one of them.
[00:45] <sarnold> wiredfool: oh! I just assumed it didn't have the storage that you were using. heh. :)
[00:46] <wiredfool> well, the other siblings are working, but they'er all qcow2, since they're not high IO images
[00:48] <wiredfool> hmm. it's failing differently now. progress I guess
[05:59] <dtscodefish> hey guys... how would i fix the perl encoding stuff? https://bpaste.net/show/c81da4763b47
[06:59] <dtscodefish> hey guys... how would i fix the perl encoding stuff? https://bpaste.net/show/c81da4763b47
[07:01] <Seveas> dtscodefish: set LANG to a locale you installed, or install the missing locale :)
[07:01] <dtscodefish> how would i do that?
[07:01] <Seveas> for the first: unset LANG
[07:02] <dtscodefish> k
[07:02] <Seveas> for the second: add en_US.UTF-8 to /var/lib/locales/supported.l/local and run sudo locale-gen
[07:02] <Seveas> the text to add is: en_US.UTF-8 UTF-8
[07:04] <dtscodefish> would it matter if its .d instead of .l?
[07:07] <dtscodefish> thanks Seveas :D
[07:08] <Seveas> dtscodefish: .l was a typo, .d is the right path :)
[07:08] <dtscodefish> oh well. assumed it was. errors went away. thanks man :D
[08:05] <lordievader> Good morning.
[08:06] <dtscodefish> morning to you as well
[08:07] <lordievader> o/
[11:22] <DonRichie2> When I install postfix on ubuntu some ncurses Menu wants to know if I want to install a "internet site" or only "local"..etc... Where do I find these sample config files to understand the mail system better?
[12:35] <peetaur> Hello, can someone help me with an 'apt-get update' performance issue? I am using apt-cacher-ng, and it takes over 5 minutes to "apt-get update" if I use localhost as the proxy. If I have no proxy, it takes under one minute. In sources.list, changing the us.archive.ubuntu.com stuff to de.* made it faster I think, and then the security.ubuntu.com ones don't have a us. so I don't know what to change them to; I tries setting them to de.
[12:35] <peetaur> archive instead of security, and now it takes only 48s to "apt-get update" but have I broken something? what is the difference between the *.archive.* and security.* sources?
[12:37] <peetaur> (and debian with apt-cacher-ng takes only 14s to update... I don't know why Ubuntu is so slow)
[12:42] <peetaur> based on results this thing generates, I guess de.archive.* is fine to replace security.*  http://repogen.simplylinux.ch/generate.php
[12:57] <peetaur> should it really take over 5 min? what is it hanging on? :/  https://bpaste.net/show/540e0985d260
[12:59] <peetaur> and without apt-cacher-ng, it takes 8s... https://bpaste.net/show/412d6d57d9ef
[13:02] <lordievader> peetaur: Perhaps using strace you can figure out what it is doing.
[13:03] <peetaur> how should I use it? does this tell me anything relevant? https://bpaste.net/show/27144b6fd814
[13:04] <lordievader> Select operations that timeout... Suppose it is related.
[13:05] <lordievader> Have to say that I am not sure.
[13:08] <lordievader> But I am not sure if the trace you pasted is of the getting/hitting the repo's.
[13:09] <lordievader> peetaur: Could you run 'strace -o apt.trace apt-get update' (or whatever apt-get command you use) and pastebin the apt.trace file once it is done?
[13:09] <peetaur> here is some in between timeouts ... but also I found that apt-cacher-ng has crashed and isn't running during this output I think https://bpaste.net/show/cbc3cfe84775
[13:09] <peetaur> once it's done...? it might never finish ;)
[13:10] <peetaur> but sure, it's running now... we'll see how long it takes
[13:11] <lordievader> That "No such file or directory" messages aren't nice.
[13:11] <peetaur> oh one thing I see in there that is very wrong:  accept(5, {sa_family=AF_INET, sin_port=htons(60651), sin_addr=inet_addr("10.2.0.87")}, [16]) = 18
[13:11] <peetaur> that .87 ip is like a year out of date
[13:12] <peetaur> oh but that Ip is not in a paste I sent you... I also ran strace on apt-cacher-ng itself
[13:12] <peetaur> https://bpaste.net/show/80e1919c2ef9
[13:12] <peetaur> so I'll go find the nth idiotic place the ip is duplicated and see if that fixes things
[13:12]  * peetaur hates duplicated useless redundant config
[13:12] <lordievader> peetaur: So your sources are mucked up?
[13:13] <peetaur> no, some other config has that IP
[13:13] <peetaur> sources.list is basically vanilla
[13:14] <peetaur> I have a puppet.list in sources.list.d/, and the proxy in apt.conf.d/
[13:15] <lordievader> Where is that non-existant(?) IP coming from?
[13:15] <peetaur> apparently it was in /etc/hosts, but that is one of the two places I should have looked long ago.
[13:16] <peetaur> so I don't know what happened... but it made not sense to look there, since to reach the proxy server (which it did) it would be using that ip, which it could not have since it was wrong
[13:17] <peetaur> and the best part is the machine worked fine for a year or however long it has been running with the new IP until today :D
[13:17] <peetaur> (and I doubt someone modified the hosts file since then)
[13:17] <peetaur> and now apt-get update takes 1m23 which is slower than without apt-cacher-ng, but it is perfectly reasonable
[13:19] <lordievader> peetaur: My machines are set to auto update at night, I rarely see the update process.
[13:19] <peetaur> apt-get update used the proxy which was set to "aptproxy" which in hosts pointed to 10.2.0.87 which doesn't exist... but it reached apt-cacher-ng anyway, and that's where the confusion happened I guess
[13:19] <peetaur> well when something is hung up you notice ;)
[13:19] <lordievader> Or it needed to timeout on the 10.2 first.
[13:21] <peetaur> why would it try the correct ip 10.2.0.8 or localhost at all though? how would it know what the right ip is?
[13:22] <lordievader> Depends on how things are set up I suppose. If there are multiple proxies set up, I think it will try them sequentially.
[13:22] <lordievader> Anyhow, this is just me rambling. Never used apt with a proxy or cacher.
[13:22] <peetaur> there was just the one I believe https://bpaste.net/show/7bda71d94909
[13:24] <peetaur> normally I find apt-cacher-ng so easy to set up, that you can use it with even 2 clients... it makes things very efficient
[13:24] <peetaur> it takes basically no config... just that Acquire line and the daemon running and it works already
[13:25] <peetaur> and today the problem was not apt-cacher-ng's fault ;)  (and btw apt-cacher works the same but crashes 50x per day, so use the -ng one)
[13:27] <peetaur> and last week I used apt-cacher-ng to install 28 machines in 1h using a slow ~500 kB/s link ;)
[13:27] <peetaur> netinstall
[13:29] <peetaur> lordievader: thx for the help by the way
[13:32] <bOynOiz> Hello, have somebody know about setup and configuration Ubuntu Server in VM before clone or export to real server machine? anywhere I can read about this?
[13:34] <peetaur> bOynOiz: any Linux should work simply by copying it to totally different hardware, with only a few exceptions...  like the r8168 network device, or some graphics and X
[13:34] <peetaur> and fakeraid or other rare stuff, wifi, etc.
[13:34] <peetaur> (desktop stuff mostly)
[13:35] <bOynOiz> Ah I see,
[13:35] <peetaur> and if you use the same kernel, etc. the r8168 issue won't happen
[13:36] <peetaur> er I mean if r8168/r8169 was already set up and confiured before
[13:38] <bOynOiz> I thinking to make virtual map drive in VMware Workstation and then rsync from map drive to real hdd before put them into real server.
[13:39] <peetaur> that should work fine, but don't forget the bootloader
[13:39] <peetaur> rsync copies the files but not the bootcode outside the filesystems
[13:39] <peetaur> here's my fully manual linux migration procedure http://pastebin.com/4NLFkdzQ
[13:40] <bOynOiz> Thank you so much :)
[13:40] <peetaur> is it a desktop or server install?
[13:40] <bOynOiz> it's server
[13:41] <peetaur> okay then I think the procedure is all you need (assuming no exceptional issues)
[13:43] <bOynOiz> yeah, my goal I don't want to close my web server for long times to setup and configuration.
[13:44] <zul> jamespage:  yeah so eventlet i dont see what you are seeing
[13:59] <zertyui> hi
[13:59] <zertyui> anyone experience with openssl ?
[14:00] <peetaur> zertyui: you should ask a real question and if someone knows the answer they'll say something
[14:01] <zertyui> ok
[14:01] <zertyui> i would like to know
[14:05] <peetaur> maybe everyone has experience with openssl, but there are so many things you can be thinking of, that nobody knows what you need to know if they know enough about it to say their the guy to ask for your problem
[14:05] <peetaur> s/their/they're/
[14:07] <lordievader> peetaur: No problem ;)
[14:22] <zul> jamespage:  lemme know when you are around
[14:27] <jamespage> zul, otp for 30 mins and then back
[14:28] <zul> jamespage:  ack
[14:30] <spidernik> hello. Anyone who uses the pam-auth-update utility? It's kinda cool but I have an issue running it in the debian-installer inside a late-script. It essentially hangs. Any similar experience? Ubuntu 14.04
[14:39] <Walex> spidernik: I would be surprised if it worked in all situations inside the installer.
[14:39] <Walex> spidernik: perhaps you may want to run it in some post-basic-install step
[14:44] <spidernik> Walex: the odd thing is: if a switch tty, chroot and run it it will work fine... it hangs just when called via the late bash-script as a late command
[14:47] <Walex> spidernik: surprising indeed.
[14:47] <Walex> spidernik: maybe it is not being run inside the '/target' chroot.
[14:49] <jamespage> rbasak: fyi that ceph point release will hold in proposed a little longer - performance regression which upstream are dealing with
[14:49] <jamespage> zul, around
[14:49] <peetaur> if you want to run in target, you use "in-target", eg.    d-i preseed/late_command string in-target apt-get install blah
[14:51] <zul> jamespage:  so 3 things: (1) im going to be updating dependencies today (2) im going to be making sure everything is green for kilo (3) i dont see the failures you have for eventlet
[14:51] <spidernik> I should probably try to run the utility with the in-target as suggested, instead of running it in the late bash script. Maybe some variable does not get set properly and makes the pam-auth-config fail
[14:51] <spidernik> thanks everyone
[14:51] <jamespage> zul, lemme recheck on eventlet
[14:51] <jamespage> I'm missing something
[14:51] <zul> jamespage:  okie dokie
[14:51] <jamespage> 1 & 2 ++ thanks
[14:51] <jamespage> zul, pull on corey a bit as well
[14:52] <zul> jamespage: i dont think corey would like that, hes got other stuff hes behind in
[14:52] <jamespage> zul, hmm ok
[14:52] <jamespage> better catchup with him soon ten
[15:01] <jamespage> zul, so what's the six PY3 failure about?
[15:02] <zul> im not 100%
[15:02] <zul> sure
[15:06] <jamespage> zul, you know I don't like "|| true"
[15:08] <zul> im pretty sure im not the one who put it there
[15:08] <zul> but yeah lemme have another look
[15:17] <Walex> spidernik: 'pam-auth-config' is not an installer command, it is a here-on-this-system command.
[15:20] <zul> jamespage:  trying with eventlet 0.17.1
[15:22] <jamespage> zul, oh ok
[15:34] <zul> jamespage:  *groan*
[15:35] <jamespage> zul, oh
[15:35] <zul> jamespage:  we are running python3 tests for eventlet when we arent building eventlet for python3 (yet) so thats why the asertion is being thrown
[15:36] <jamespage> hrm
[15:36] <jamespage> zul, ahyes!
[15:51] <zul> jamespage:  ok builds fine now
[16:08] <jamespage> zul, sameplace?
[16:08] <zul> jamespage:  eventlet builds fine when you build only for python2 since we arent using python3
[20:14] <smoser> https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1427829
[20:14] <smoser> hallyn_, ^ is that just me?
[20:14] <smoser> anyone else able to reproduce that ?
[20:17] <dtscodefish> https://letsencrypt.org/howitworks/
[20:17] <dtscodefish> oops wrong channel
[20:24] <setuid> How does one get invited to #ubuntu-cloud?
[20:24] <setuid> Looking for some dev-side help getting the cluster up and operational
[20:24] <cyphermox> smoser: I could try in a bit, but can I ask why you remove "--" ?
[20:25] <cyphermox> (wondering if there's something else than "because otherwise console= won't get applied")
[20:28] <smoser> doens't really matter, but bug 1402042 is why.
[20:28] <smoser> the kernel very helpfully changed behavior, and doesn't pay attention to things after '--' any more.
[20:28] <smoser> (note sarcasm in word 'helpfully')
[20:29] <cyphermox> right, I was fixing that this morning
[20:29] <cyphermox> tomorrow's image might work
[20:29] <cyphermox> that's why I was wondering if there was another reason too
[20:30] <cyphermox> instead of --, you'll have ---, and won't need to move things around or remove the separator; that should fix OEM mode and accessiblity and such too
[20:30] <smoser> cyphermox, oh?
[20:31] <smoser> ah. you'll chnage hte isos
[20:31] <cyphermox> well, yes, the isos should be fixed tomorrow, I think
[20:31] <cyphermox> anything that uses d-i directly (mini.iso) should already be fine, unless I missed some
[20:31] <smoser> so willthat change d-i to copy after '---' ?
[20:32] <cyphermox> d-i will continue to pass what's after --- to the installed system
[20:32] <smoser> as previously it (and curtin as a follow-on) would copy anything after '--' over to the installed system.
[20:32] <smoser> was that previous behavior ?
[20:32] <cyphermox> yes
[20:32] <smoser> really. hm..
[20:33] <cyphermox> it's meant so that you can pass, well, console= to an installed system and not have to blindly try to reapply it after installation
[20:33] <cyphermox> for example
[20:33] <cyphermox> I don't know about curtin though
[20:33] <cyphermox> is that good or bad?
[20:34] <cyphermox> hmm, perhaps curtin needs its fix too unless it uses some of the d-i bits
[21:02] <mfisch> zul: do you have an approx ETA on Juno.2 landing in updates?
[21:06] <zul> mfisch:  not yet
[21:22] <hallyn_> smoser: well, it's just you in that noone else would do that
[21:23] <hallyn_> smoser: does ctrl-m or ctrl-j work?
[21:32] <jamespage> coreycb, did that eventlet update get out of the door for utopic?
[21:32] <jamespage> mfisch, will look tomorrow - should have been today but had some other plates spinning - sorry
[21:34] <coreycb> jamespage, not yet, bug 1423675
[21:34] <coreycb> https://bugs.launchpad.net/ubuntu/+source/python-eventlet/+bug/1423675
[21:34] <coreycb> I'll check with Brian
[21:34] <jamespage> coreycb, ack - thanks
[21:35] <jamespage> coreycb, everything is in juno proposed for the CA - have we run the testing yet?
[21:35] <jamespage> if not I can trigger tomorrow
[21:35] <coreycb> jamespage, I've not yet run against the ca
[21:36] <jamespage> coreycb, ok leave it for me tomorrow
[21:36] <coreycb> jamespage, alright, thanks
[21:36] <jamespage> its been in proposd for a while - we can push to updates if it tests ok
[21:37] <hallyn_> smoser: seems most likely to be a seabios/vgabios/whatever bug...  do you know whether it used to work in trusty or precise?
[21:55] <mfisch> jamespage: thanks, been reloading the /Packages file every morning! ;)
[21:55] <jamespage> mfisch, http://reqorts.qa.ubuntu.com/reports/ubuntu-server/cloud-archive/juno_versions.html
[21:57] <mfisch> jamespage: that shows 2.2 is out?
[21:58] <jamespage> mfisch, almost - 2.2 is in proposed finishing testing - when the column on the rhs goes green we're all done
[21:58] <jamespage> mfisch, you can always test from proposed - its exactly the same binaries as will end up in updates
[22:05] <mfisch> yeah the testing looked ok
[22:15] <beisner> jamespage, coreycb, i can kick of trusty & utopic juno-proposed deploys.  just say when.
[22:16] <junix> hello, would like to know if someone can help me if i can get help with a problem that i have with a ubuntu server
[22:17] <bekks> junix: Just ask your actual question. If someone is able to help you. ,you will get a response.
[22:17] <junix> ok
[22:17] <multiverse> Hello, how do I load the NFS module into the 14.10 kernel?
[22:18] <bekks> multiverse: More likely, you start the NFS client/Server, depending on what you need.
[22:18] <multiverse> I wish
[22:19] <multiverse> I am unable to start an ordinary nfs server on 14.10, and it tells me I need to install the NFS kernel module.
[22:19] <bekks> multiverse: So try this: https://help.ubuntu.com/14.04/serverguide/network-file-system.html
[22:20] <multiverse> That has not resulted in the error going away.  It’s the first thing I read.
[22:20] <junix> i have a virtual server that was powered off wrong and i have a problem rebooting, and a message as follows: "Kernel panic - not syncing: Attempted to kill init!"
[22:21] <bekks> multiverse: And did you install the nfs kernel server module?
[22:21] <multiverse> The question remains:  “How do I install the nfs kernel module when the apt-get install nfs-etc… are already installed, but the service one start.
[22:21] <multiverse> That’s what I am asking
[22:21] <multiverse> Howdo I do that?
[22:22] <nuno> Hello. I've just installed ubuntu 12.04 server, however i cannot install packages using apt-get
[22:22] <rww> nuno: do "sudo apt-get update" if you haven't already
[22:22] <bekks> multiverse: Like this: https://help.ubuntu.com/14.04/serverguide/network-file-system.html#nfs-installation
[22:22] <multiverse> That doesn’t work
[22:22] <nuno> rww: i did, and resulted in error at end : : Failed to fetch bzip2:/var/lib/apt/lists/partial/us.archive.ubuntu.com_ubuntu_dists_precise_main_source_Sources  Hash Sum mismatch
[22:22] <bekks> multiverse: Why not?
[22:23] <multiverse> thanks for the help.
[22:23] <rww> nuno: hrm, there's your problem. if you repeat the command, does it still happen?
[22:23] <rww> (sometimes that's a one-off blip)
[22:24] <nuno> rww: checkin atm
[22:24] <nuno> rww: yes it does
[22:26] <rww> hrm. do other WWW connections work (e.g. through w3m, wget)
[22:28] <nuno> rww: yes i have connectivity to the internet
[22:29] <nuno> rww: I guess i found it. at least I can now install git. other packages I'm not sure. I followed this http://askubuntu.com/questions/41605/trouble-downloading-packages-list-due-to-a-hash-sum-mismatch-error
[22:37] <nuno> rww: thanks for the help though =)
[23:33] <bdmurray> Is there anybody who could give me a hand with haproxy logging?