[00:52] <webfox> Could someone help me reduce and keep it reduced after boot the console fonts please ?
[01:08] <hallyn> stgraber: trying out your daily lxc build.  lxc-ls doesn't work :(
[01:20] <halvors> JanC: No i know that. I'm sending outgoing SMTP traffic from my server via the ISP's relay. What's the problem is that i cannot connect, authenticate and send from it. I use the mail-stack-delivery package :) The host is s1.halvors.org telneting it on port 25 works just fine :)
[01:23] <stgraber> hallyn: how so?
[01:24] <stgraber> hallyn: it's still the old upstream shell version as my python implementation hasn't been acked yet :)
[01:24] <webfox> Could someone help me reduce and keep it reduced after boot the console fonts please ?
[01:42] <hallyn> stgraber: it doesnt' show running containers :(
[01:42] <hallyn> were there objections to the python version?
[01:43] <hallyn> all right lemme review right now :)
[01:46] <stgraber> hallyn: the only objection was about removing the shell version which I'll try to avoid doing (by adding some magic installing the python versions when possible and otherwise falling back to shell)
[01:46] <stgraber> hallyn: lxc-ls --active is what you want I believe
[01:46] <stgraber> hallyn: upstream's lxc-ls doesn't give you the same output as Debian's
[01:52] <hallyn> stgraber: no --active gave me junk
[01:55] <qman__> smw, thanks, that's what I was looking for
[02:37] <TheOnionRack> I've just installed Ubuntu 12.10 Server onto my VPS, and I've found that it's running Apache2 and Samba out of the box. Is that really the default behaviour, or is that something that my VPS provider has done to their VM image? It seems really odd to me that Ubuntu would ship the server with those services preinstalled and already running.
[02:38] <sarnold> TheOnionRack: that'd be something your VPS did for you
[02:38] <JanC> it's not the default
[02:38] <TheOnionRack> I thought that
[02:38] <TheOnionRack> I wonder what else they've changed
[02:38] <sarnold> https://wiki.ubuntu.com/Security/Features#ports
[02:39] <JanC> also, running Samba by default on a public VPS sounds stupid...
[02:39] <sarnold> ++
[02:39] <TheOnionRack> Yeah. That's what I thought. It has home shares set to read only by default :O
[02:41] <sarnold> maybe they're selling "pre-made cloud for you and your windows friends!" or something, but it seems strange...
[02:41] <sarnold> I'd expect OpenSSH, because it's the easiest way to make anything else work :) but that's it.
[02:42] <TheOnionRack> It has openssh running out of the box along with bind
[02:42] <JanC> certainly nothing else should listen externally
[02:42] <TheOnionRack> but it's not marketed as a cloud VPS, it's just bog-standard VPS
[02:44] <JanC> is samba actually configured to listen publicly?
[02:44] <TheOnionRack> yeah
[02:45] <TheOnionRack> you can mount \\vps.example.com\username and browse that user's homeshare (read only)
[02:45] <TheOnionRack> across the internet
[02:45] <TheOnionRack> It's kinda mindblowing
[02:47] <JanC> I wonder why ISPs insist on changing distro images at all...
[02:47] <JanC> except maybe if something is needed for the hardware or such
[02:48] <sarnold> JanC: that's almost a sign that they need better hardware :)
[02:48] <sarnold> TheOnionRack: very yikes,
[02:48] <sarnold> TheOnionRack: if it were tied to _your_ IP that'd be a little less scary
[02:49] <sarnold> but I'd be reluctant to ever use a mahcine that'd run samba on a public-facing IP -- I'm not sure I'd ever trust it to not be pre-rooted.
[03:12] <pndemc> Is there a way to wget an entire directory and all of its contents?
[03:15] <sarnold> pndemc: --mirror --no-parent is probably what you're after
[03:20] <pndemc> sarnold, that looks for an index.html, I just have a bunch of files in the folder that I need it to download
[03:22] <sarnold> pndemc: how else would the client discover the names of the files to request?
[03:22] <pndemc> they are not being hidden, simple accessing the directory should reveal the contents
[03:23] <pndemc> it does in the browser
[03:23] <sarnold> and your webserver does not automatically generate a directory listing?
[03:23] <pndemc> it does
[03:23] <pndemc> wget is not picking up on it I guess though
[03:23] <sarnold> hrm. I've nevre had a problem with wget --mirror --no-parent http://server/path/to/foo/ -- where there was no index.html, but the server generated one for me...
[03:24] <sarnold> normally apache generates six different stupid views, and wget will download all six of them, it feels like an utter waste of time :)
[03:26] <pndemc> sarnold, "domain.com/downloads/tf: No such file or directorydomain.com/downloads/tf/index.html: No such file or directory
[03:26] <pndemc> "
[03:26] <pndemc> sarnold, as you can see it's looking specificly for index.html
[03:30] <sarnold> pndemc: how about domain.com/downloads/tf/   ?
[03:30] <sarnold> (trailing slash)
[03:35] <pndemc> sarnold, same error
[03:39] <sarnold> pndemc: no kidding :(
[03:46] <ekaj> I installed the KDE desktop on ubuntu server, and I can't figure out how to run Dolphin as root - does anyone have any suggestions?
[03:50] <sarnold> ekaj: sudo dolphin  ?
[03:52] <ekaj> Ah sorry, it threw errors so I assumed it didn't work, but it did
[03:53] <sarnold> a great many kde programs spew a giant pile of debugging information to standard out or standard error, but continue to work just fine...
[03:53] <sarnold> I find it uesful to run the 'first' KDE program in a terminal window, then minimize that terminal, and run everything else from another terminal -- the messages all come from the first one (which starts the kdeinit process)
[05:01] <ScottK> If ekaj comes back, tell him Dolphin wasn't designed to run as root.  If he wants to run a KDE based GUI file manager as root, he should use Krusader.
[05:02] <sarnold> thanks ScottK :)
[07:50] <jamespage> morning all
[07:55] <philipballew> is running ksplise on my home server a bad idea?
[07:55] <philipballew> by bad idea I mean, is it gonna mess up my system?
[08:01] <sarnold> philipballew: what's ksplise?
[08:03] <philipballew> sarnold, http://www.ksplice.com/
[08:03] <sarnold> woah, this things still around?
[08:05] <philipballew> sarnold, apparently
[10:36] <iliv> hi
[10:37] <iliv> I need to run apt-get -s dist-upgrade on hardy for scripting purporses over ssh, but it wants me to do it via sudo
[10:37] <iliv> precise, for example, doesn't care about sudo and let's you run apt-get/aptitude
[10:38] <iliv> I'd like to be able to do the same with hardy
[10:38] <iliv> any ideas how?
[10:38] <Jeeves_> visudo NOPASSWD
[10:48] <iliv> Jeeves_, the point is, precise doesn't care about sudo at all. You could just run apt-get -s dist-upgrade
[10:48] <iliv> I don't like your suggestion because it means writing code for just one host
[10:49] <iliv> I've got mostly precise servers
[10:58] <Jeeves_> iliv: It's your call. I just gave you a suggestion
[11:05] <iliv> Jeeves_, right
[14:40] <roaksoax> jamespage: howdy
[14:40] <jamespage> hey roaksoax
[14:40] <roaksoax> jamespage: so looking at the quantum-gateway charm
[14:40] <roaksoax> jamespage: we only need HA when using OVS then?
[14:41] <roaksoax> jamespage: or in all cases?
[14:41] <jamespage> roaksoax, hmm
[14:41] <jamespage> roaksoax, that charm may only be required when using the ovs plugin
[14:41] <roaksoax> jamespage: right, since I see that charm installs different packages for OVS and NVP
[14:41] <jamespage> roaksoax, however as I've not actually worked with any of the other plugins as yet
[14:42] <jamespage> I'm not 100% certain
[14:42] <yolanda> jamespage, zul, is the merge proposal i have pending for Quantum very urgent? quite busy today....
[14:42] <jamespage> roaksoax, that may be fud
[14:42] <roaksoax> jamespage: right, so: utils.install(*qutils.GATEWAY_PKGS[PLUGIN])
[14:42] <roaksoax> the OVS plugin is the one that install the l3-agent and dhcp-agent
[14:42] <zul> yolanda: nah
[14:43] <jamespage> roaksoax, yes; the nvp one is not populated because I simply don't know yet
[14:43] <jamespage> need to RTFM
[14:43] <yolanda> ok
[14:44] <roaksoax> jamespage: alright, my questions was based on the fact that in the ha-relation-joined we don't need to send cluster data if the plugin is not OVS
[14:45] <jamespage> roaksoax, right now - yes
[14:46] <roaksoax> jamespage: ack!
[14:53] <jamespage> Daviey,  * radosgw: openstack keystone integration
[14:53] <jamespage> made it into ceph 0.55
[14:54] <jamespage> 0.56 will be the next LTS
[15:35] <hallyn> kirkland: you have a kill-a-watt or somesuch right?  What do you measure as the difference in draw in a normal laptop when idle and doing a package upgrade?
[15:36] <hallyn> stgraber: I'm goign to push the two patches from yesterday (rsync -H) to staging, unless you have a comment?
[15:37] <stgraber> hallyn: I must admit I didn't really follow that thread but it looks like everyone agrees now? if so, fine :)
[15:37] <hallyn> k
[15:42] <zul> jamespage: https://code.launchpad.net/~zulcss/nova/nova-testsuite-fix/+merge/137652
[15:52] <hallyn> stgraber: have you used lxc.network.type = phys?  I'm a little surprised lxc hasn't been renaming the nic on shutdown all this time...
[15:55] <stgraber> hallyn: nope, never used it. I actually started working on adding network devices support to lxc-device yesterday, though for lxc-device, there's no way the name can be restored on container exit...
[15:58] <hallyn> ok, just testing to see if my naive lazy attempt at renaming works
[16:09] <stgraber> hallyn: I think that at some point I'll want the equivalent of lxc-device in the API so that lxc can do the right thing on shutdown (at least for network devices), but it's not really a priority
[16:09] <stgraber> hallyn: basically having some call to pass a new network device from the host to the container, renaming it in the process and adding it to the list of interfaces to rename back on shutdown
[16:12] <tgm4883> What is the best way to add iscsi storage to libvirt?
[16:12] <tgm4883> There doesn't seem to be any good docs on it and it fails in virt-manager
[16:21] <stgraber> hallyn: hey, so I've been looking at your patch for the loglevel stuff. Is it me misreading or are you always doing lxc_log_priority_to_string() on the value, even though config_loglevel supports both strings and digits?
[16:22] <hallyn> stgraber: i store an int in the conf
[16:22] <hallyn> config_loglevel, if i didn't mess up, shoudl store an int
[16:24] <stgraber> hallyn: oh, I see c->loglevel is an int and config_loglevel requires a char*
[16:25] <stgraber> ok, makes sense, I'll ack it :)
[16:25] <hallyn> stgraber: suppose i could store the const char* in config...
[16:25] <hallyn> thanks
[16:25]  * hallyn back to server mtg :)
[16:53] <ttx> smoser: http://www.cloudbase.it/cloud-init-for-windows-instances/
[16:53] <RoyK> oh, does ubuntu support asterisk these days - that's some task ;)
[16:53] <smoser> yeah, i just saw that.
[16:53] <smoser> apiolitti asked me if i'd accept windows patches.
[16:55] <RoyK> windows?
[16:55] <RoyK> oh, there's a windows version of * now, it seems
[16:56] <RoyK> must be wonderful
[16:56] <RoyK> mixing the spaghetti code of asterisk with windows - two negatives must make a positive, right? http://karlsbakk.net/fun/asterisk-installation.wav
[17:16] <stgraber> hallyn: submitted the patch to allow shipping "legacy" scripts in lxc
[17:22] <hallyn> stgraber: cool
[17:23] <hallyn> stgraber: other than high prio/critical bugs, for now my plan is to push fixes to github and let the fix roll into raring with the next merge
[17:28] <stgraber> hallyn: that's what I've been doing for a while. Anyway, we should be getting alpha1 next week, so really not worth the effort of making patches for ubuntu
[17:46] <hallyn> stgraber: jinkeys, the error paths in lxc_spawn() are pretty messed up.
[18:37] <bjf> arosales, you asked about kernel tests using utah? the answer is no.
[18:40] <arosales> bjf: roger that. Thanks for the reply.
[18:40] <bjf> arosales, np
[18:40] <bjf> arosales, any particular reason for the question?
[18:41] <arosales> it was a follow on comment smb actually had at todays ubuntu-server IRC meeting.
[18:42] <arosales> sorr bjf^
[19:15] <streulma> hello, is there a ppa for PHP 5.3 Daily builds on Ubuntu?
[19:22] <stgraber> hallyn: I'm going to trigger a daily of lxc very soon, want your nic renaming patch in there?
[19:22] <stgraber> hallyn: I did a quick review and just posted a single question on the list about it
[19:22] <RoyK> isn't php 5.3 in upstream?
[19:29] <zul> hallyn: getting the new qemu to test now
[19:32] <zul> hallyn: :( http://pastebin.ubuntu.com/1410947/
[19:41] <tdelam> Hi, I've installed Ubuntu 12.04 LTS server on my server (go figure!) it's been running for a while now and it's merely a web proxy server, it drops connection sporadically ( every few days or weeks ), I can't figure out why. Can anyone recommend a few reasons or things to look for as to why this is happening?
[19:43] <sarnold> tdelam: anything in the proxy logs? anything in dmesg?
[19:44] <tdelam> sarnold: no, nothing out of the ordinary
[19:45] <tdelam> dmesg was the first place i looked
[19:45] <tdelam> typical stuff that happens on boot
[19:47] <sarnold> tdelam: are there a handful of long-lived connections that die? or dozens or hundreds or thousands of short-lived connections that die?
[19:53] <tdelam> sarnold: one long-lived one
[19:53] <sarnold> tdelam: is there a NAT firewall between the two peers?
[19:54] <tdelam> yes, cisco PIX, i've ruled that out being that hundreds would go down if it were that device.
[19:54] <tdelam> once the connection was dropped, I was able to SSH while local and test outgoing connections, those were the not responsive
[19:55] <sarnold> tdelam: how long is that connection idle when the connection dies? Many NAT firewalls impose an inactivity timeout, and it's often very short
[19:55] <sarnold> is it ever idle? sometimes idle? idles for two hours sometimes?
[19:55] <tdelam> sarnold: ms, it's a web proxy for an ISP site which has a lot of traffic.
[19:55] <tdelam> no, never
[19:55] <sarnold> damn. :D
[19:55] <sarnold> tdelam: any chance that pix has logs around the time the connection drops?
[19:55] <tdelam> ;) it' very weird.
[19:56] <hallyn> zul: what the heck?  those were supposed to conflict+replace
[19:56] <tdelam> sarnold: there could be, I haven't checked and to be honest that pix is managed by a 3rd party vendor oddly.
[19:56] <tdelam> i ruled out the pix considering the fact that all others would drop. I could get someone to login to check the logs though
[19:57] <sarnold> tdelam: hrm, thjat's an unfortunate complication, but one that makes a certain amount of sense. (I wouldn't want to manage a cisco machine, I don't have the first clue where to start with them. hehe.)
[19:57] <tdelam> its not fun
[19:57] <tdelam> if it were hardware dmesg would report it right?
[19:58] <hallyn> zul: guess i need to add some.  sorry about that.  trying another push
[19:58] <sarnold> tdelam: you could throw tcpdump on one or both peer and try logging the details of the connection .. maybe some simple grep filtering to knock out the most common packets, or some way to save only the most recent ten minutes or ten thousand packets or something. oof.
[19:58] <sarnold> tdelam: probably. check also the /var/log/mcelog -- if you've got the mcelog package installed and and MCE happened, it'll log verbose messages there...
[20:00] <tdelam> ok
[20:08] <zul> hallyn: otherwise looks ok
[20:09] <hallyn> zul: new version pushed (but i'm not convinced it's ok)
[20:09] <hallyn> zul: ok, thanks.  you just had qemu-kvm installed and did apt-get dist-upgrade?
[20:09] <zul> right
 swear i did that...
[20:09] <hallyn> ok i'll test that once it builds, then if it's ok ask you to try one more time to make sur ei'm not doing something funky
[20:10] <hallyn> heck, might be easier to just port the source to 1.3 :)
[20:10] <hallyn> that'll fix the -z versioning crap too
[20:10] <hallyn> which btw i'm not sure how to handle when p ushign to the archive.  is there a precedent for that?
[20:17] <hallyn> stgraber: didn't say it in the email, but pushed the physical nic renaming change into staging with your comment fixed, thanks
[20:18] <stgraber> hallyn: was just about to ask you if you did :) good
[20:21] <lvmer> I've got a kind of annoying question... I'm about 3 weeks into ubuntu-server, & I'm struggling with a 'custom program' install. I'm trying to install & configure minidlna-transcode  alongside: minidlna, imagemagick, & ffmpeg.
[20:22] <lvmer> Information on 'minidlna-transcode' can be found here: https://bitbucket.org/stativ/minidlna-transcode    (for reference)
[20:39] <lvmer> is there an easy way to page down to your next search term in a 'man' file? or do I just /searchterm   then press f repeatedly to find the highlights myself?
[20:40] <sarnold> lvmer: 'n'
[20:40] <sarnold> lvmer: N will go backwards
[20:41] <sarnold> lvmer: ? will start searching backwards; in that case, n will keep going backwards and N will then go forwards
[20:42] <lvmer> sarnold: thank you again. :)  it worked perfectly. No more manual hunting xD
[20:42] <lvmer> do you have any advice on installing a custom program or compiling it? like minidlna-transcode? by chance? xD
[20:44] <sarnold> lvmer: normally, you follow the untar, ./configure --prefix=/usr/local ; make ; sudo make install   dance -- you lose the distro packaging, but most things support the autotools ./configure script
[21:06] <lvmer> yah he said it supports ./autogen.sh and ./configure & make & make install
[21:07] <lvmer> I've just never done it before so I'm a little hesitant on all of it
[21:08] <lvmer> I was afk, because I had a firewall protocol warning & had to fix it. xD
[22:00] <jamespage> adam_g, I think I've changed my mind about the network-manager config in the nova-compute charm
[22:01] <adam_g> jamespage: o rly? :)
[22:03] <adam_g> jamespage: you prefer to keep it set as a config option in both places?
[22:03] <jamespage> adam_g, I going to sleep on it and take another look tomorrow....
[22:03] <lvmer> sarnold: no such file:  ./configure
[22:03] <adam_g> jamespage: okay. also, did you say there was some discussion somewhere recently about kombu / rabbit HA wrt nova?
[22:24] <sarnold> lvmer: heh, not all programs supply a ./configure -- you should figure out how to modify the build scripts for your program to install it into a directory of your choosing, so you can install it into /usr/local or ~/local/miniwhatever...
[22:24] <lvmer> http://paste.kde.org/619832/
[22:25] <lvmer> sarnold: yah I'm running into an './autogen.sh' problem
[22:26] <lvmer> I had to install a bunch of packages: libtool autopoint autoconf automake
[22:26] <sarnold> that's too bad, quite often the software distributor will do that for you
[22:27] <lvmer> this is a fairly small package or w\e
[22:27] <lvmer> for dlna transcoding with 'minidlna' which happens to be one of the only working dlna's for my tv at the moment. :/
[22:27] <lvmer> What do you think I am missing?
[22:28] <lvmer> I have to install AM_ICONV && AM_GNU_GETTEXT ?
[22:29] <sarnold> lvmer: try installing libc6-dev
[22:29] <lvmer> already installed
[22:29] <lvmer> 0 upgraded 0 newly installed 0 remove 0 not upgraded
[23:02] <stgraber> hallyn: and a bunch more python changes for your review :) making lxc-device a lot more useful and allowing for non-root lxc-ls
[23:15] <hallyn> stgraber: cool.  one reviewed, other may have to wait.  good ngiht
[23:15] <stgraber> hallyn: ok, good night
[23:15] <hallyn> (cause the lxc-ls one looks complicated :)
[23:16] <stgraber> hallyn: it's actually much less complicated than the lxc-device one :)
[23:16] <lvmer> sarnold: still no luck xD  I emailed the creator. Can't find anything on 'm4_pattern_allow' in the configs. :/
[23:17] <stgraber> hallyn: just moving code around and changing the loop to only grab the object if absolutely needed
[23:20] <sarnold> lvmer: the solution will be easier than that; it'll proabably just be installing one or another package. I just don't know which one, nor do I know off-hand how to figure out what to install.
[23:22] <lvmer> sarnold: ah.. well I did try 'autotools-dev' && 'autogen'  but still no luck
[23:25] <sarnold> lvmer: is there no README with good details? :)
[23:25] <lvmer> no lol. I asked already :(
[23:25] <sarnold> be sure to contribute back a README with details on the pakages to install :) the next guy will thank you
[23:26] <lvmer> I am already making one
[23:26] <lvmer> sarnold: :)
[23:26] <sarnold> yay
[23:26] <lvmer> sarnold: actually lol I did find part of one brb... lmao
[23:45] <lvmer> sarnold: if I have a file:  new.txt    & it contains: sudo apt-get install libexif 12 libjpeg 62 libid3tag 0 libFLAC 8 libvorbis 0a libvorbisenc2 libsqlite3-0 libavformat53 libavformat-extra-53       how do I run said file?
[23:47] <sarnold> lvmer: with those spurious numbers in there your best bet is probably to just copy-and-paste the package names out by hand
[23:49] <lvmer> do I save the file as .sh?
[23:50] <sarnold> lvmer: mv new.txt new.sh
[23:51] <lvmer> yah that moves it
[23:51] <lvmer> how do I run the command inside though?
[23:53] <sarnold> lvmer: sh new.sh  or sh new.txt
[23:55] <lvmer> bash
[23:55] <lvmer> xD got it