[00:24] <kirkland> RoAkSoAx: you there?
[00:24] <kirkland> RoAkSoAx: looking at https://help.ubuntu.com/community/Cobbler/Installation
[00:25] <kirkland> RoAkSoAx: i think we should try to fix 1 and 2 in the cobbler packaging
[00:25] <kirkland> RoAkSoAx: set it to the host's ip address at least
[00:29] <sparc> Is there a way to get the installer to send it's messages to a remote syslog?
[00:46] <RoAkSoAx> kirkland: i'm here
[00:46] <kirkland> RoAkSoAx: cool;  i'd like you to review a cobbler change for me
[00:46] <RoAkSoAx> kirkland: yeah we can definitely do that
[00:46] <RoAkSoAx> kirkland: sure! :)
[00:47] <kirkland> RoAkSoAx: http://paste.ubuntu.com/615564/
[00:47] <kirkland> RoAkSoAx: something like that
[00:47] <kirkland> RoAkSoAx: what do you think?
[00:48] <RoAkSoAx> kirkland: looks good to me and I do think we should override those configs
[00:49] <kirkland> RoAkSoAx: do you have a setup where you could test this?
[00:49] <RoAkSoAx> kirkland: yes, applying patch now
[00:49] <kirkland> RoAkSoAx: cool, let me know if it works for you
[00:49] <kirkland> RoAkSoAx: i'm testing here too
[00:52] <kirkland> RoAkSoAx: bug
[00:52] <kirkland>  ... /etc/cobber/settings -> cobbler
[00:53] <kirkland> RoAkSoAx: other than that, worked here for me
[00:53] <kirkland> RoAkSoAx: http://paste.ubuntu.com/615566/
[00:53] <kirkland> RoAkSoAx: i've committed and pushed to bzr
[00:53] <RoAkSoAx> cool, installing now
[00:53] <kirkland> RoAkSoAx: did you fix that bug ^
[00:54] <kirkland> RoAkSoAx: i'll upload to oneiric if it works right for you
[00:54] <RoAkSoAx> kirkland: yes I'm using your new patch from paste
[00:54] <kirkland> RoAkSoAx: sweet
[00:57] <RoAkSoAx> kirkland: i think i just found another bug
[00:58] <kirkland> RoAkSoAx: okay?
[00:58] <kirkland> RoAkSoAx: whatcha got?
[00:58] <RoAkSoAx> kirkland: ok, so wlan0 is the one who has ip addr from the network
[00:59] <RoAkSoAx> while br0 is the bridge for eth0
[00:59] <RoAkSoAx> that i'm not using atm
[00:59] <kirkland> RoAkSoAx: ruh roh
[00:59] <RoAkSoAx> but when I run this: awk 'END {print $1}' /proc/net/route
[00:59] <kirkland> RoAkSoAx: hmm
[00:59] <RoAkSoAx> I get br0 as the interface, instead of wlan0
[01:00] <kirkland> RoAkSoAx: hmm
[01:00] <RoAkSoAx> kirkland: this is how ip route sh shows:
[01:00] <RoAkSoAx> default via 192.168.1.254 dev wlan0  proto static
[01:00] <RoAkSoAx> default dev br0  scope link  metric 1003
[01:00] <kirkland> RoAkSoAx: which interface makes most sense here?
[01:01] <RoAkSoAx> kirkland: in my case wlan0 as is the one who is really working and if I would like it to work out of the box it wouldn't
[01:01] <RoAkSoAx> kirkland: but maybe it is a corner case
[01:01] <kirkland> RoAkSoAx: hmm, this is a tough one
[01:02] <RoAkSoAx> kirkland: but giving a second though, let's just not worry about it just yet cause we just need to ensure it grabs a default route interface
[01:02] <RoAkSoAx> kirkland: and in my cause, I have 2 default routes, which should be avoided
[01:03] <RoAkSoAx> kirkland: and if this comes as a bug with someone having multiple interfaces, then we can deal with it, unless, we ask a debconf question
[01:04] <RoAkSoAx> kirkland: so we could just set the defaults the way you did it right now... but still asking the user whether we should keep that address or change it for another one, resolving the issue of having to deal with multiple interfaces on which two of them might have been misconfigured as default routes
[01:10] <RoAkSoAx> kirkland: cause otherwise it just works well
[01:10] <monaDeveloper> Hi, Every time I launch an ubuntu ec2 instance I just can login via ssh once with the key assigned after that I get Permission denied
[01:10] <monaDeveloper> (publickey)
[01:10] <monaDeveloper> I run ssh -i mykey ubuntu@ec2instance via my bash of course my OS is ubuntu natty.
[01:12] <monaDeveloper> any advice?
[01:12] <monaDeveloper> is there a specific room for ubuntu ec2?
[01:14] <rewt> are there instructions there saying that that's a one-time key and you have to generate your own key for subsequent logins?
[01:15] <RoAkSoAx> so yeah i/win 3
[01:15] <RoAkSoAx> arghh
[01:16] <monaDeveloper> no
[01:21] <rewt> are you running ec2-run-instances with the -k option?
[01:21] <RoAkSoAx> kirkland: updated docs aswell already
[01:27] <monaDeveloper> actually I'm accessing a running instance that was launched via aws console
[01:27] <monaDeveloper> so I login without firing ec2-run-instances
[02:09] <zul> hallyn: around?
[02:12] <hallyn> zul: yeah
[02:12] <zul> hallyn: do you extract the linux4tegra to the flash as well?
[02:13] <hallyn> no, you just install that onthe pc you'll connect with usb to the netbook
[02:13] <hallyn> and, i did need to patch the install program slightly for it to work for me
[02:16] <zul> gotcha
[02:17] <hallyn> zul: see http://paste.ubuntu.com/615583
[04:57] <kirkland> RoAkSoAx: agreed, sounds reasonable
[05:01] <kirkland> RoAkSoAx: okay, well my changes are committed to the lp:ubuntu/cobbler branch
[05:01] <kirkland> RoAkSoAx: feel free to release those whenever you're ready!
[05:05] <airtonix> if i followed this guide to creating a self signed cert for development purposes only, do i need to add more hostnames to the cert if they are just subdomains ?
[05:05] <airtonix> http://www.tc.umn.edu/~brams006/selfsign.html
[06:02] <airtonix> strange, if i $ mount --bind source target then add target to /etc/exports as /exports/target and mount from a client i can see the contents of target on the server, but if i use a equiv mount command from fstab the contents of target are hidden
[06:05] <airtonix> maybe not, probably just the funky /etc/exports options i was using as per the autofs wiki page
[08:41] <_ruben_> bugger .. my local repo mirror doesn't have enough space left to mirror natty (or even oneiric)
[10:23] <jits1998> hello
[11:18] <kickar> hey guys can someone help me to install php 5.2 and php 5.3 , so I can have both on the same machine?
[11:26] <soren> No.
[11:29] <koolhead11> soren, heh
[11:31] <kickar> is it possible at all ?
[11:31] <soren> no
[11:32] <soren> If it were, someone could help you.
[11:33] <koolhead11> kickar, why would someone do that?
[11:33] <qman__> it is technically possible, but would require two separate web servers compiled from scratch in their own segregated areas
[11:33] <qman__> which is definitely not a supported solution
[11:33] <kickar> I found some howtos and tried but no luck
[11:33] <kickar> I have a couple of sites that returns error in php 5.#
[11:33] <kickar> 5.3
[11:34] <soren> Fix them.
[11:34] <kickar> is there a solution for that ?
[11:34] <soren> 5.2 isn't supported.
[11:35] <kickar> one of the apps that I am using is openx, and I hardly can rewrite their code
[11:37] <zul> then its a bad app
[11:37] <qman__> 5.2 to 5.3 isn't a major leap
[11:37] <qman__> not like php4 to php5
[11:37] <qman__> some stuff broke, but it's not a complete rewrite type of fix
[11:38] <qman__> if the package STILL isn't updated for 5.3, I wouldn't want to run it anyway
[11:38] <kickar> i tried to install php 5.2 from source but couldn't get it to run anyway
[11:39] <qman__> as others have mentioned, 5.2 isn't supported
[11:39] <qman__> I'm talking about your web app
[11:39] <qman__> fix it yourself, complain to the authors, or find a different app
[11:39] <qman__> those are the best solutions
[11:43] <koolhead11> kickar, i doubt that would be a major issue. Openx is a well known ad-server IMHO
[11:45] <jkakar> I have an Upstart init script for my application, which depends on PostgreSQL... I can't seem to figure out how to make it start on boot, after PostgreSQL has started... any ideas/pointers?
[11:45] <koolhead11> kickar, you can also try #openx :P
[11:46] <kickar> thanks
[11:46] <soren> jkakar: I forget... Has postgresql been upstartified?
[11:46] <jkakar> I tried looking in the Upstart Cookbook and also Googling... but to no avail.
[11:46] <jkakar> soren: Nope, it hasn't.
[11:46] <jkakar> soren: At least not on 10.04, which is the environment I'm using.
[11:48] <jkakar> As much as Upstart is cool it's also a massive pain in the ass.  It'd be nice if we had one init system for everything.
[11:48] <jkakar> One that was easy to debug for bonus points. :)
[11:48] <koolhead11> jkakar, +1 :D
[11:56] <jkakar> Okay, I found a solution.
[11:57] <jkakar> I was using 'start on started network' before and it wasn't starting at all... now I'm using 'start on (local-filesystems and net-device-up IFACE!=lo)'.
[11:58] <soren> jkakar: Yeah, that's a fairly common pattern.
[11:58] <jkakar> There's a race with PostgreSQL, but we're using 'respawn' and our application does a start-up check to make sure it can connect to the database... if it can't, it bails... it gets restarted and eventually PostgreSQL is up so things work.
[11:58] <jkakar> A but ugly, but it works.
[11:58] <jkakar> soren: I don't understand why 'start on started network' didn't work, though.
[11:59] <soren> Not sure.
[14:48] <sommer> morning all :-)
[14:52] <ssureshot> can you specify */6 to run a job every six days ?
[14:52] <ssureshot> in crontab ?
[14:53] <patdk-wk> I don't believe it will
[14:53] <Pici> ssureshot: sure
[14:53] <Pici> why not/
[14:53] <patdk-wk> pretty sure the *count* starts over at 0
[14:53] <Pici> hmm
[14:53] <ssureshot> why not what I'm asking myself why it didn't run this morning .. I set it last thursday..
[14:54] <ssureshot> 0 9 * * */6 /usr/local/bin/app
[14:54] <patdk-wk> I'm saying it undrestands that as running */6 times
[14:55] <patdk-wk> and as only 6/6 will satasfy that, it will only run on day 6
[14:56] <ssureshot> ah,, roger that.. so any */# will run on the day that number corresponds to
[14:56] <zul> hey sommer
[14:57] <plm> Hi all
[14:58] <plm> people, my server a re with date correct. but when I create a directory ou send file to server date is worong.. why?
[14:58] <plm> where I configure to proftpd use date of Operating System?
[15:02] <plm> anyone?
[15:09] <jMCg> plm: have you checked proftpd's documentation?
[15:10] <jMCg> By default, you see, most sane servers just use UTC and so should your OS, btw :)
[15:11] <BrixSat> hello im trying to boot a live cd on a pc with disk problems, how can i disable ubuntu to check disk errors?
[15:11] <BrixSat> so i can recover something
[15:28] <luite> is there a customized selinux ubuntu policy available from somewhere, something that's a bit more usable (that allows sshd to run /bin/bash for example...)
[15:30] <jMCg> BrixSat: why would a live cd do that?
[15:30] <jMCg> luite: AppArmor :)
[15:34] <hallyn> luite: i'm hoping to get to that soon in the ubuntu policy.  fel free to fix it if you can
[15:34] <hallyn> there is an ssh module, so it is a bug
[15:36] <luite> hallyn: I'm still investigating whether to use apparmor or selinux, don't have experience with customizing profiles for either.
[15:36] <hallyn> oh, then stick with aa - supported :)
[15:36] <sommer> I have some time to do some iso testing... how much of that is automated?  just don't want to waste my time double checking something that isn't needed :-)
[15:39] <luite> I need something that lets some daemon change its own security context (looks like both can do that) to something safe enough for execution of arbitrary code uploaded by users
[15:41] <luite> apparmor policies look a bit easier to make
[15:42] <hallyn> luite: seccomp?  :)
[15:42] <luite> hallyn: nah that's too limited, the code still needs some file access, just very limited (no network though)
[15:44] <hallyn> put it in empty netns
[15:44] <hallyn> then aa for file access
[15:45] <jdstrand> luite: for apparmor, you might be interested in http://wiki.apparmor.net/index.php/Documentation, specifically http://wiki.apparmor.net/index.php/Pam_apparmor_example
[15:45] <jdstrand> luite: iirc, there is some stuff with changehat for ssh. you could ask in #apparmor on oftc for more info on either
[15:47] <luite> ah thanks, was already reading the first link, but the example looks instructive :)
[15:47] <duli> How do I know if a service is permanently on, meaning it will be started after reboot?
[15:49] <luite> are there major differences between selinux and apparmor in what can be defined in policies?
[16:01] <jamespage> hey sommer: take a look here - http://jenkins.qa.ubuntu-uk.org/view/oneiric-server/
[16:02] <jamespage> this is what is currently automated.
[16:02] <sommer> jamespage: ah sweet
[16:02] <jamespage> that said its nice to have manual tests as well as they do (sometimes) pickup UI issues.
[16:39] <ScottK> jamespage: How does one get additional tests added to that?
[16:42] <jamespage> ScottK: well to start with the preseed and the tests need to be developed and integrated into the ubuntu-server-iso-testing project (see http://launchpad.net/ubuntu-server-iso-testing)
[16:42] <SpamapS> sommer: o/ !
[16:42] <sommer> SpamapS: hello hello :-)
[16:42] <ScottK> jamespage: I was thinking of seeing if we could get some Kubuntu tests added in there.
[16:42] <jamespage> I'm hoping to fine some time this cycle to make it a bit easier as at the moment it requires the project to be re-build and re-installed on all of the test nodes
[16:42] <SpamapS> sommer: BTW, the RAID1 tests aren't automated, so that would be a good one to do. :)
[16:43] <sommer> SpamapS: working on that now actually... for i386 anyway
[16:43] <SpamapS> sweet!
[16:43] <ScottK> jamespage: If I aim a Kubuntu/QA oriented person at you, could you help them get set up?
[16:43] <jamespage> that would be a good use case - jibel has done alot of work automating the desktop and alternate ISO install testing using the same framework
[16:43] <SpamapS> sommer: I put in a fix early in oneiric so let me know if you have any problems
[16:44] <sommer> SpamapS: will do, thanks man
[16:44] <jamespage> ScottK: absolutely
[16:44]  * jamespage thinks we might need a bigger server soon
[16:46] <ScottK> Great.
[16:57] <kirkland> RoAkSoAx: hiya
[16:57] <kirkland> RoAkSoAx: would you mind releasing/uploading cobbler sometime soon-ish?
[16:59] <RoAkSoAx> kirkland: howdy! yeah I wanna do that right now... I'm gonna dd the debconf stuff later... just finishing 1 minor powernap thingy
[16:59] <kirkland> RoAkSoAx: sweet
[16:59] <kirkland> RoAkSoAx: regarding powernap... i had something i wanted to ask you about
[16:59] <RoAkSoAx> kirkland: shoot
[17:00] <kirkland> RoAkSoAx: i'm wondering if it ever makes sense to run powernap in a virtual machine
[17:00] <kirkland> RoAkSoAx: i'm guessing "no", except for our own testing purposes
[17:00] <RoAkSoAx> kirkland: I actually though about it and wasw trying to find its uses
[17:01] <RoAkSoAx> kirkland: the only thing that I came up with is that it will serve as a monitoring system to pretty much poweroff a VM whenever it was not needed
[17:01] <kirkland> RoAkSoAx: i bet it'll have a very minimal impact on the host
[17:01] <kirkland> RoAkSoAx: hmm, yeah
[17:01] <RoAkSoAx> kirkland: cause in terms of power savings, doesn't really make any sense (specially using pm-powersave)
[17:01] <kirkland> RoAkSoAx: anyway, I was thinking we should have a check, and if not explicitly enabled, I think we should disable powernap if we can tell we're running in a virtual machine
[17:02] <kirkland> RoAkSoAx: at which point, I'd like to make orchestra-client recommend powernap
[17:02] <kirkland> RoAkSoAx: such that orchestra-managed clients are running powernap in powersave mode by default
[17:03] <RoAkSoAx> kirkland: yeah... I guess we could have a pre-start in the upstart script that actually checks that
[17:03] <kirkland> RoAkSoAx: yeah, that sounds good
[17:04] <kirkland> RoAkSoAx: how do we tell if we're running in a vm, though?  is there a common function we can use
[17:05] <RoAkSoAx> kirkland: that's the problem :).., might there be anything under /proc that tells us that?
[17:05] <kirkland> RoAkSoAx: hacky, but there's /proc/cpuinfo
[17:06] <kirkland> RoAkSoAx: i swear there was a new tool that did this ...
[17:06] <kirkland> RoAkSoAx: maybe facter?
[17:06] <RoAkSoAx> negronjl: ^^ can facter tell us if we are running in a VM?
[17:07] <negronjl> RoAkSoAx: no but, you can use fact-add ( facter-customfacts-plugin ) to add a fact to that effect.
[17:07] <RoAkSoAx> negronjl: ok thanks
[17:07] <kirkland> RoAkSoAx: yeah, facter
[17:07] <smoser> jamespage, around ?
[17:07] <jamespage> smoser: yep
[17:07] <RoAkSoAx> kirkland: but, if we use facter, would it make sense to add the check logic to the upstart job?
[17:08] <kirkland> RoAkSoAx: sudo apt-get install facter
[17:08] <kirkland> RoAkSoAx: facter
[17:08] <smoser> we need test of oneiric
[17:08] <smoser> images
[17:08] <negronjl> RoAkSoAx: ie:  fact-add <some_new_fact> <some_new_value>
[17:08] <kirkland> RoAkSoAx: facter virtual
[17:08] <negronjl> RoAkSoAx: ie: fact-add vm true
[17:08] <zul> kirkland: im going to be doing an upload for cobbler on friday
[17:08] <negronjl> kirkland:  virtual is also used on EC2/UEC so it can be confusing
[17:08] <jamespage> smoser: right - I'll get it setup to execute then :-)
[17:08] <kirkland> zul: okay
[17:08] <smoser> JAMES PAGE
[17:08] <kirkland> negronjl: right
[17:09] <kirkland> negronjl: no, it already has support for virtual
[17:09] <kirkland> negronjl: we just need to consume that
[17:09] <kirkland> RoAkSoAx: my only concern with doing it in upstart is that it's kinda slow
[17:09] <kirkland> RoAkSoAx: (takes 0.4s here)
[17:09] <negronjl> kirkland:  the problem with virtual is that it won't tell you whether it is in a VM or in the cloud
[17:09] <kirkland> RoAkSoAx: that said, i think it's the right way to solve this
[17:10] <kirkland> negronjl: we don't care about that
[17:10] <negronjl> kirkland: ok
[17:10] <kirkland> negronjl: we just want to know if we're running in a vm, or on physical hw
[17:10] <kirkland> negronjl: so that we can conditionally start powernapd, or not if we're in a vm
[17:10] <RoAkSoAx> kirkland: yeah that's what I was thinking but other than that the check would make sense
[17:10] <kirkland> RoAkSoAx: looks like it's in main
[17:10] <kirkland> RoAkSoAx: so we can depend on it
[17:11] <kirkland> RoAkSoAx: okay, i like it;  you want to hack that into powernap's upstart, or should I/
[17:11] <RoAkSoAx> kirkland: but, if we add that we can cause that powernap will never start in a VM if I want it to, if I'm using poweroff as action_method
[17:11] <kirkland> RoAkSoAx: right, we should allow for an override
[17:12] <RoAkSoAx> kirkland: or in the daemon itself
[17:12] <kirkland> RoAkSoAx: or you could just edit the upstart script locally
[17:12] <RoAkSoAx> kirkland: something like "if vm and action_method is powersave, kill daemon"
[17:13] <kirkland> RoAkSoAx: hmm, i was thinking "if vm and not vm_override then don't start daemon (exit true)"
[17:13] <RoAkSoAx> kirkland: ok, we'll have to document that though!
[17:13] <kirkland> RoAkSoAx: sure
[17:14] <smoser> jamespage, before going full out, let me sniff boot of a couple different instance types.
[17:14] <smoser> at least recently m1.large would not boot
[17:14] <RoAkSoAx> kirkland: alrighty then, makes sense to me
[17:18] <RoAkSoAx> negronjl: thanks btw ;)
[17:18] <negronjl> RoAkSoAx: np
[17:20] <jamespage> OK - I'll do a quick test of the setup
[17:20] <jamespage> smoser: ^^
[17:25] <jamespage> smoser: http://tinyurl.com/3omfdej - seems to work OK; lemme know when you want to run the big one
[17:28] <smoser> jamespage, what does that mean ? m1.small worked?
[17:28] <smoser> if you can ... feel free to run a quick single region single ami launch of all types.
[17:28] <smoser> if m1.small works i'd guess that c1.medium does.
[17:28] <smoser> so amd64 types would be the ones i question
[17:29] <smoser> bug 784345
[17:29] <smoser> i'm testing m1.large now
[17:36] <RoAkSoAx> kirkland: just updated the powernap branch with the second stage action method if you wanna test it (and upload it for me :) )
[17:37] <smoser> m1.large is good, jamespage
[17:37] <smoser> go ahead and kick a full test.
[17:40] <zul> JAMES PAGE
[17:40] <kirkland> RoAkSoAx: you need to apply for core-dev, or per-package at least for powernap :-)
[17:40] <RoAkSoAx> kirkland: I already did, meeting is next week :)
[17:41] <kirkland> RoAkSoAx: sweet
[17:42] <kirkland> RoAkSoAx: did you ever hear back from zonker?
[17:42] <kirkland> RoAkSoAx: about testdrive?
[17:42] <kirkland> (I didn't)
[17:43] <RoAkSoAx> kirkland: nope, didn't hear anything I was wondering if he uses irc
[17:44] <RoAkSoAx> and track him down here
[17:45] <kirkland> RoAkSoAx: talk to akgraner
[17:45] <kirkland> RoAkSoAx: she knows him, works with him regularly
[17:45] <RoAkSoAx> kirkland: will do
[17:55] <RoAkSoAx> kirkland: btw.. did you ever ask for the UPS/PDU?
[17:55] <kirkland> RoAkSoAx: asked, have not received it yet
[17:56] <RoAkSoAx> kirkland: neither have I
[17:57] <RoAkSoAx> kirkland: ok so I'm uploading your changes to cobbler now cause I don't wnat that branch to get lost or further conflicts and when I add the debconf stuff I'll just do a new upload
[17:57] <kirkland> RoAkSoAx: k
[17:58] <kirkland> RoAkSoAx: just make sure you stay in sync with zul
[17:58] <kirkland> RoAkSoAx: he said he's working on it too
[17:58] <RoAkSoAx> kirkland: ok will do
[17:58] <RoAkSoAx> zul: ping! I'm uploading recent kirkland's changes to oneiric
[17:58] <zul> for what cobbler?
[18:24] <smoser> jamespage, did you start test ?
[18:27] <jamespage> smoser: http://jenkins.qa.ubuntu-uk.org/view/oneiric-ec2/job/oneiric_server_ec2/1/
[18:28] <smoser> gratze
[18:35] <hallyn_afk> soren: bug 540446, did you in fact have a patch for that?
[18:45] <ilya> Hello  :)
[18:46] <ilya> I can't get my euca server to talk to the node..network issues i'm assuming
[18:46] <ilya> I'm running 11.04
[18:48] <ilya> All help is very much appreciated :) Thank you
[18:51] <adam_g> is openstack known to be working on oneiric? i noticed yesterday nova-manage was broken and still is today
[18:55] <evilsushi> what are you guys using for a iptables frontend? anyone using shorewall?
[18:55] <evilsushi> just wanted to get some thoughts on what everyone in here may be using.
[18:58] <ilya> Can anyone help me out with getting the cloud/node talking to one another in a private network with a switch?
[18:59] <sommer> evilsushi: ufw is pretty great for servers... it's host based though
[18:59] <guntbert> evilsushi: I tried ufw and shorewall - liked shorewall better
[19:00] <evilsushi> host based?
[19:01] <evilsushi> explain.
[19:01] <evilsushi> guntbert: what did you like about shorewall?
[19:01] <evilsushi> the idea of zones for the network interfaces seems kinda "neat".
[19:01] <evilsushi> are you using shorewall with these type of features?
[19:02] <evilsushi> sommer: does ufw have the idea of a config file?
[19:03] <guntbert> evilsushi: I had the task to provide a very tightly closed router - and it was easier accomplished (for me) with shorewall (I better understood its concept than that of ufw)
[19:05] <ilya> silly uni internet
[19:06] <ilya> here is my ifconfig output for server/node: http://pastie.org/2004371, Am I supposed to add the dns ip to the interfaces file?
[19:07] <sommer> evilsushi: host based meaning not using the server to transfer traffic... ufw doesn't have a config file exactly, but integrates with some packages like samba to open the necessary ports
[19:07] <sommer> evilsushi: if you need to setup a router something like shorewall is probably better suited than ufw
[19:08] <sommer> evilsushi: ufw can be used though, but may require more work :-)
[19:09] <sw0rdfish> when I ssh'd into my vps it said i have mail, lol how do i check that?
[19:12] <guntbert> sw0rdfish: type mail (one way, install and use alpine is another)
[19:12] <sw0rdfish> yep i used my imagination and just typed mail
[19:13] <sw0rdfish> thanks guntbert :)
[19:14] <guntbert> sw0rdfish: :) pine aka alpine is my preferred mail client on CLI
[19:14] <sw0rdfish> i'll install it now
[19:14] <ilya> can anyone see me?
[19:15] <genii-around> ilya: We can see what you type, yes
[19:16] <ilya> genii-around: great :) now if only they could have acknowledged that :) Thank you for doing that
[19:23] <sw0rdfish> wow guntbert thanks alpine is pretty cool
[19:24] <sw0rdfish> listen is there a program or just a way to download stuff into my vps
[19:24] <sw0rdfish> via http links
[19:24] <sw0rdfish> but well i guess wget will have to do it...
[19:24] <guntbert> sw0rdfish: you're welcome :-)  - w3m is apretty web browser, wget for direct download
[19:25] <sw0rdfish> w3m
[19:25] <sw0rdfish> a cli web browser?
[19:25] <sw0rdfish> nice
[19:25] <sw0rdfish> i think i'll install it
[19:25] <sw0rdfish> i'll install webmin too....it looks cool
[19:25] <guntbert> sw0rdfish: of course :), and while you are installing: take screen too  - no, don't use webmin
[19:25] <guntbert> !webmin
[19:26] <sw0rdfish> oh :o
[19:26] <sw0rdfish> aww damn!
[19:26] <sw0rdfish> i loved their screenshots
[19:26] <sw0rdfish> in their website
[19:26] <sw0rdfish> :(
[19:26] <sw0rdfish> it seemed like a cool program to handle your server via gui stuff
[19:26] <sw0rdfish> via web-based gui server thingy
[19:26] <sw0rdfish> but oh well
[19:27] <sw0rdfish> is there something like it out there good with ubuntu?
[19:27] <guntbert> well, the *screenshots* won't hurt anything :)
[19:27] <sw0rdfish> is there something like it out there good with ubuntu?
[19:27] <guntbert> sw0rdfish: not to my knowledge
[19:27] <sw0rdfish> guntbert: i'ma google it, :)
[19:28] <sw0rdfish> while i'm installing it take screen? what ya mean by that guntbert
[19:28] <guntbert> !info screen
[19:29] <ilya> when setting up eucalyptus, do i need to add dns information to the interfaces file?
[19:30] <guntbert> sw0rdfish: yes, links2 even can disply images
[19:30] <sw0rdfish> :o
[19:30] <sw0rdfish> NiCE!
[19:30] <sw0rdfish> think i'll get links2
[19:30] <sw0rdfish> then
[19:30] <kirkland> RoAkSoAx: pulled powernap, will review on the plane
[19:31] <guntbert> sw0rdfish: please don't crosspost across ubuntu channels, and take it easy on the <enter> key - especially in #ubunutu
[19:32] <RoAkSoAx> kirkland: cool, thanks!
[19:32] <sw0rdfish> lol ok
[19:32] <sw0rdfish> i thought they're different channels, but no big deal...
[19:33] <frewsxcv> i'm installing ubuntu server and it's asking me what i want to install. what is "Basic Ubuntu Server"
[19:33] <sw0rdfish> whats with irc channels and enter key, can they really get hurt with it
[19:33] <guntbert> sw0rdfish: they are, but some helpers read both (and some more...)
[19:33] <guntbert> !enter
[19:34] <guntbert> busy channels are scrolled even fatser
[19:34] <guntbert> *faster
[19:35] <sw0rdfish> ok gotcha.
[19:59] <sidnei> is the kernel-ppa the right place to get lts-backport kernels for lucid?
[19:59] <frewsxcv> anyone have any idea why when i boot grub i get: error: fd0 read error
[19:59] <frewsxcv> i think it has to do with me having a software raid
[20:19] <smoser> jamespage, still around ?
[20:20] <smoser> http://jenkins.qa.ubuntu-uk.org/view/oneiric-ec2/job/oneiric_server_ec2/1/ what does red mean ?
[20:21] <osmosis> all of a sudden out of nowhere my server started having major IO issues. related to any updates?
[20:39] <osmosis> My io times dropped without any detectable cause.  system reboot seems to have fixed.  bug?
[20:39] <jamespage> smoser: well it means that something did not happen according to plan
[20:39] <jamespage> for example a) server did not start-up correctly first time and had to be rebooted.
[20:39] <jamespage> b) unable to run tests for some reason
[20:39] <smoser> ...
[20:41] <smoser> cloud-config failed across the board
[20:41] <smoser> but i dont know why
[20:42] <smoser> ie, please help, jamespage
[20:42] <jamespage> ERROR:root:Something bad happened Unable to execute in-instance tests.
[20:45] <smoser> where do you see that?
[20:47] <smoser> ah... it looks like our udev friend is much more common now
[20:47] <jamespage> hmmm - it would appear that executing the in-instance tests results in a non zero return code
[20:48] <jamespage> that message is at the bottom of the log in all of the cloud init tests I've looked at so far.
[20:48] <smoser> i think that the instances are not coming up with networking
[20:48] <smoser> http://jenkins.qa.ubuntu-uk.org/view/oneiric-ec2/job/oneiric_server_ec2/1/ARCH=amd64,REGION=ap-southeast-1,STORAGE=instance-store,TEST=simple-user-data,label=ubuntu-server-ec2-testing/artifact/None/amd64/m1.large/instance-store/i-793b8f2c/268a2649-0f8c-4a6b-aa66-621243928a07-running.console.txt
[20:48] <smoser> yeah.
[20:48] <smoser> well, that is good that it is more reproducible
[20:50] <jamespage> yep - there is alot of that
[20:50] <jamespage> even if a reboot fixes that it will cause a test failure = red blob
[20:51] <smoser> well, thats good
[20:51] <smoser> "reboot fixes my problem" isn't really ok here
[20:51] <smoser> :)
[20:52] <jamespage> really? and I though servers just not starting was normal :-)
[20:52] <jamespage> I think the other issue might be a bit of a bug in the framework
[20:53] <jamespage> all of the cloud-init tests add a ppa - which does not have oneiric content
[20:53] <jamespage> apt-get update might return a non-zero
[20:53] <soren> It certainly will.
[20:53]  * soren had this problem on his first oneiric test box
[20:56] <jamespage> would that cause the calling script to return non-zero as well?  it does not use '-e' so will not fail
[20:58] <jamespage> bah - thats definately the problem
[20:59] <jamespage> smoser - so that looks like a change in return code for something called during the in-instance tests - we can make that script a bit more rugged
[20:59] <jamespage> I want to change it so it runs out-of-band from Jenkins anyway
[21:00] <smoser> ah...
[21:00] <smoser> yeah, the ppa will screw things up.
[21:00] <smoser> suck
[21:00] <smoser> what ppa is it?
[21:00] <smoser> i will add a oneiric so it doesn't fail
[21:00] <smoser> we should assert that the ppa is sane
[21:01] <jamespage> http://jenkins.qa.ubuntu-uk.org/view/oneiric-ec2/job/oneiric_server_ec2/1/ARCH=amd64,REGION=eu-west-1,STORAGE=ebs,TEST=cloud-config,label=ubuntu-server-ec2-testing/artifact/tests/cloud-config/ud-cloudconfig-01.txt
[21:02] <jamespage> hmm - prob should have tiny url'ed that one
[21:02] <sidnei> smoser, speaking of ppa, do you know which ppa can i find the lts-backport kernels in? i suspect kernel-ppa it is, but the description says 'daily builds'
[21:02] <jamespage> thats the cloud-config
[21:02] <h4lfl1ng> how do i check if two computers are talking to one another over a private network using a switch?
[21:02] <smoser> sidnei, don't knwo personally. i think i did the same google as you once.
[21:04] <|rt|> h4lfl1ng: need more information to answer for certain especially if there are more than one route between the hosts etc
[21:04] <|rt|> h4lfl1ng: but traceroute should show you the path that the traffic is taking if any routers are involved
[21:05] <|rt|> h4lfl1ng: if it's only one hop then you can assume that both computers are on the same broadcast domain....but not necessarily the same switch if that's what you're looking for
[21:07] <h4lfl1ng_> |rt|: the second isn't getting an internet connection, but it should through the first right?
[21:08] <h4lfl1ng_> the server is connected to the internet and switch, and the other pc is connected to switch only
[21:08] <h4lfl1ng_> |rt|: what other information would you need?
[21:08] <|rt|> only if the first is setup to route the traffic
[21:09] <jamespage> smoser: interestingly the instances of the udev issue seem to be isolated to instance-store instance rather than ebs
[21:09] <jamespage> or at least it seems much worse in those tests
[21:09] <smoser> so faster local disk
[21:10] <jamespage> ebs is faster than instance-store?
[21:11] <h4lfl1ng_> |rt|:should ping work to check if there is a connection between the two machines?
[21:13] <|rt|> h4lfl1ng_: yes...assuming they are in the same subnet you should be able to ping them
[21:15] <h4lfl1ng_> |rt|:they are in the same subnet 255.255.255.0, but the second machine has a bridge, which shouldn't matter right?
[21:15] <h4lfl1ng_> |rt|: and the 2nd also has it's gateway set as the 1st's eth1 ip
[21:16] <h4lfl1ng_> but they don't ping
[21:19] <zertyu> hello there
[21:19] <zertyu> how to check system performance remotely ?
[21:20] <|rt|> h4lfl1ng_: I don't think the bridge would matter as it's essentially a switch
[21:20] <_Neytiri_> how to i install updates via cmd line?
[21:21] <|rt|> h4lfl1ng_: I would check the firewalls to make sure you're not blocking pings
[21:22] <zertyu> apt-get update _Neytiri_
[21:23] <zertyu> how to check system performance remotely ?
[21:23] <zertyu> is that possible ?
[21:25] <|rt|> zertyu: are you looking for monitoring and alerting?
[21:25] <zertyu> let me explain the problem what i got, i got small web server i need to upgrade hardware in case  there is lots of usage, i don't know what i have to do
[21:25] <zertyu> to see when i have to upgrade the hardware, as that thing depends on visitors usage
[21:26] <|rt|> zertyu: sar can collect data locally but I tend to prefere munin, cacti, or Zenoss
[21:26] <|rt|> I've also heard good things about Zabbix
[21:26] <zertyu> sorry i don't get you
[21:27] <zertyu> what i need exactly ?
[21:27] <zertyu> do you understand what i mean ?
[21:28] <|rt|> http://munin.ping.uio.no/
[21:28] <|rt|> is that what you're looking for
[21:28] <zertyu> i need to upgrade hardware depending the visiotors
[21:29] <zertyu> where we don't know about the number visitors will come to my site
[21:30] <|rt|> If I understand what you're looking for is that you want to monitor usage and trending data so that you know when you need to upgrade your server(s)
[21:30] <zertyu> yes i simply want to check my system performence
[21:30] <zertyu> regulary
[21:31] <zertyu> and to be notify in case needing to upgrade the hardware
[21:31] <zertyu> is that possible ?
[21:31] <|rt|> sure using any of the packages I mentioned earlier
[21:32] <|rt|> sysstat package includes sar which will collect data on the computer
[21:32] <h4lfl1ng_> |rt|: this is my ifconfig info for both machines: http://pastie.org/2005277
[21:32] <zertyu> wich one is the best  ?
[21:33] <|rt|> but munin, zenoss, cacti, zabbix, icigna, nagios will all do this sort of thing too
[21:33] <|rt|> zertyu: depends on your needs some of those I mentioned are monitoring and alerting solutions others just collect and display trending data
[21:33] <zertyu> i tell you the needs
[21:34] <zertyu> just only want to notify about my system performence
[21:34] <zertyu> in case of lots of demands
[21:37] <donnie> I just installed Ubuntu server 11.04 on Dell Inspiron E1505. Do i have to install bcmwl-kernel-source like i do in Ubuntu to setup my wifi?
[21:37] <zertyu> so which is the best ?
[21:39] <|rt|> zertyu: I don't think there is a "best" but if you want alerting and monitoring you'll need to use Zenoss, Zabbix, icinga, or nagios
[21:39] <|rt|> zertyu: cacti may have added some alerting though
[21:39] <|rt|> Zenoss is pretty heavy though so I don't know that I woudl use it to monitor one host
[21:39] <|rt|> icigna or nagios would be my choice
[21:40] <|rt|> but you'll need plugins to get the trending data if you want to collect that as they are less integrated than Zenoss and Zabbix in that area
[21:42] <|rt|> h4lfl1ng_: what are the frame errors on machine 2 eth0
[21:42] <zertyu> just find xymon what do you think about that ?
[21:44] <|rt|> zertyu: looks similar to icigna and cacti but looks unmaintained if the latest release was a couple years ago
[21:44] <|rt|> zertyu: nagios is in ubuntu's repo
[21:45] <hggdh> zul, is the test rig available?
[21:46] <zertyu> well is it possible to monitoring icinga by iphone 4 ?
[21:47] <|rt|> zertyu: I believe so....icinga is just a fork from nagios...think of it as a more open nagios with a nicer web UI
[21:48] <h4lfl1ng_> |rt|: where do you see the errors? by the way, i just got them talking, but now to figure out how to route the internet connection to the second
[21:49] <zertyu> is it flexible for managing tasks ?
[21:49] <zertyu> i don't have any experience with monitoring services
[21:49] <zertyu> is it flexible for managing tasks ?
[21:50] <|rt|> h4lfl1ng_: look under the 2nd machine eth0 ifconfig output
[21:50] <|rt|> zertyu: what do you mean by tasks?
[21:51] <zertyu> administrative task
[21:51] <zertyu> administration task
[21:51] <|rt|> zertyu: most monitoring and alerting systems are very flexible to the point that they can be somewhat complex to setup
[21:52] <h4lfl1ng_> |rt|: oh, i see it now. It's no longer there on the machine. How would i check what the error is anyways?
[21:53] <zertyu> well icinga is it easy for setup ?
[21:55] <|rt|> I don't have any experience with it so no idea
[21:56] <|rt|> I would assume that it's similar nagios though which is a bit picky about it's configuration files which are not always intuitive in the way the different config files interact with each other
[21:56] <h4lfl1ng_> |rt|: how would i route the traffic with a switch, i found the guide for a system with crossover cable..but this is slightly different. I'll have to do it to 7 other machines
[21:56] <|rt|> but perhaps icinga has moved most config stuff to the web ui
[21:57] <zertyu> thanks lrtl
[21:58] <|rt|> h4lfl1ng_: have a look at https://help.ubuntu.com/community/Router and https://help.ubuntu.com/community/Internet/ConnectionSharing
[22:00] <|rt|> h4lfl1ng_: at the very least you'll need to enable ipv4 forwarding and setup ip masquerading (NAT)
[22:01] <|rt|> h4lfl1ng_: but why use ubuntu as your gateway?  I tend to use something more geared towards that task like pfSense
[22:02] <h4lfl1ng_> |rt|: and that would be enough to run apt-get?
[22:02] <|rt|> pfSense is freebsd based so no
[22:03] <h4lfl1ng_> |rt|: oh, it's a distro..ll, i'm creating a cloud setup, this is only temporary. Maybe later we could have a dedicated firewall/router
[22:05] <|rt|> well you can run pfSense as a virtual machine if you set up the virtual networking part correctly
[22:35] <ammar> ubuntu vi editor keybord problem it not wirting my words down how do i fix that
[22:38] <Pici> ammar: This might sound a bit silly, but do you know how to use vi?
[22:48] <zul> hggdh: always has been
[22:48] <hggdh> zul, :-) just checking, better safe, and all that
[22:55] <ammar> ya
[22:56] <ammar> type something it dosnt work but my keybord works
[23:16] <zul> adam_g: around?
[23:20] <pvthodson> So whats new in the server world anything worth trying?
[23:23] <hallyn_afk> ensemble :)
[23:23] <pvthodson> So whats new in the server world anything worth trying?
[23:32] <lynxman> ensemble
[23:36]  * hallyn_afk waits for it....
[23:37] <lau> i am having lots of jbd2/sda6-8(732): WRITE block 273600 on sda6 (8 sectors) on 2.6.38-8-generic #42-Ubuntu SMP Mon Apr 11 03:31:50 UTC 2011 i686 i686 i386 GNU/Linux
[23:38] <lau> it looks like #607560 but I haven't find any work around any idea ?
[23:48] <earthmeLon> Just installed postfix on my new server.  My old server had postfixadmin and used mysql for alias and such.  I am trying to figure out how to get postfix to use the mysql databases that i've imported and such.  Anybody familiar with this?
[23:52] <luite> earthmeLon: you should probably look at /etc/postfix/main.cf from your old server and look for the mysql: type maps
[23:52] <earthmeLon> Thanks luite
[23:52] <luite> (it's been a while since I configured postfix, and I don't know the specifics of postfixadmin, so things might have changed a bit in the meantime)
[23:53] <luite> earthmeLon: anyway these mysql maps refer to other config files that describe how to connect to the database and which column's to use
[23:53] <JanC> there will be separate config files that define the mapping to the database too
[23:53] <JanC> ☺