[01:41] <codepython777> http://ppa.launchpad.net/yannubuntu/boot-repair/ubuntu/ - whats this ppa?
[04:05] <MACscr> anyone having issues with the ubuntu repos tonight?
[04:05] <MACscr> getting resolving issues
[04:10] <bradm> MACscr: what are you seeing?
[04:10] <MACscr> nvm, guess the system got some networking issue after upgrading from 12.04 to 12.10
[05:38] <MACscr> still havent solved why the networking is no longer working though =(
[05:38] <MACscr> doesnt appear to get an ip from dhcp anymore
[06:07] <Rallias> So... every time in the last month that I've done init 6 on my newly installed ubuntu servers, it's just shut down the machines... is this intentionally diversionary behaviour or is it something I should be concerned with?
[06:36] <rbasak> Rallias: "init 6" is an unusual way to do it. I think it should work, but it's an unusual path, given that we don't use Sys V init. Does "reboot" work?
[06:37] <Rallias> No...
[06:42] <Rallias> rbasak, The only reboot thing that works reliably is sysrq...
[06:42] <rbasak> Rallias: sounds like you have an issue there then.
[08:06] <laite> Hi, I have spent past hour trying to figure how to install Ubuntu server 12.04.3 from usb - no matter what I seem to do, I'm always stuck on "problem reading data from CD-ROM"
[08:06] <rbasak> !details
[08:06] <laite> I have tried creating install usb with dd, unetbooting and startup disc creator, and everytime the result is same
[08:07] <rbasak> So you can't boot the installer at all?
[08:07] <rbasak> At what stage do you get the error?
[08:07] <laite> installer boots, I can set locale and such, but it stops on error where it "can't find cdrom"
[08:07] <rbasak> I see.
[08:08] <rbasak> Have you checked the image you're writing?
[08:08] <rbasak> md5 or sha256 or whatever.
[08:09] <laite> ah, no, obviously I should've
[08:09] <laite> I'll see to it now
[08:11] <laite> yes, at least the downloaded file seems correct
[08:11] <laite> how do I check if it's correct on my usb?
[08:11] <rbasak> You can dd the other way
[08:11] <laite> ah, thanks
[08:12] <rbasak> dd if=/dev/... bs=128k|cmp - original.iso
[08:12] <rbasak> Or something like that
[08:12] <rbasak> (otherwise you'll go past the end and it'll come up wrong anyway)
[08:12] <rbasak> Or check the size and calculate and use dd count=
[08:19] <laite> rbasak:  I ran `dd if=/dev/sdd bs=128k|cmp - ubuntu-12.04.3-server-amd64.iso` and it gave just a line: "cmp: EOF on ubuntu-12.04.3-server-amd64.iso"
[08:19] <laite> does this mean it's all right?
[08:19] <rbasak> Yes
[08:20] <laite> so the disk isn't faulty
[08:20] <rbasak> Next, I'd try another machine and see if you can get past that stage. It's before anything actually gets done to the disk, right?
[08:20] <laite> right, I'll try with my laptop
[08:20] <rbasak> There might be some BIOS settings around USB boot emulation to tweak
[08:23] <laite> still the same "can't find cd-rom" error on another machine
[08:41] <kalle_> I have a problem with the package "ftp-proxy". I have set up a reverse proxy for my FTP on a firewall/NAT machine, but I get random "Connection reset by peer" when doing file listings in empty directories...
[09:05] <rbasak> yolanda: regarding https://launchpadlibrarian.net/147274320/buildlog_ubuntu-saucy-i386.rubyluabridge_0.7.0-2ubuntu1_FAILEDTOBUILD.txt.gz, the buildds resolve build dependencies slightly differently. I reproduced your sbuild result OK, and compared the build dep resolution against the archive build failure. I get these differences: http://paste.ubuntu.com/6043311/. It looks to me that on the buildds things like "require rdoc/task" and "require mkmf" a
[09:05] <rbasak> I never figured out if there's a way to get sbuild to do exactly what the buildds will do for build dependency resolution. It's annoying.
[09:05] <yolanda> rbasak, what i saw is that my local sbuild uses ruby1.9 and builder there ruby1.8
[09:05] <yolanda> that should be the key
[09:05] <yolanda> i need to force ruby1.9
[09:06] <rbasak> I get both ruby1.8 installed in both environments.
[09:06] <rbasak> And ruby1.9.1-dev in both environments.
[09:08] <rbasak> Some source package intentionally build for both ruby1.8 and ruby1.9.1, so dropping one and forcing another may cause a regression in that functionality. I'm not sure if that applies to this package, though.
[09:08] <rbasak> I think that adding a build-dep on ruby and/or ruby1.9.1 may fix it.
[09:10] <yolanda> rbasak, i was comparing my build log with that, mine is setting up update-alternatives: using /usr/bin/ruby1.9.1, and builder is setting up ruby1.8
[09:10] <yolanda> so yes, trying to add ruby 1.9 as dep
[09:12] <yolanda> i thought sbuilds for same platforms should work the same in all builders, then there is no way to be sure 100% that a package builds until it's uploaded?
[09:15] <Daviey> yolanda: You can download the chroot used by launchpad, to better match the remote environment
[09:16] <yolanda> Daviey, how can i do it, i'm interested on it
[09:24] <Daviey> yolanda: wget http://launchpadlibrarian.net/147277667/chroot-ubuntu-saucy-i386.tar.bz2
[09:26] <yolanda> thx
[09:58] <RoyK> huh - http://paste.ubuntu.com/6043505/ - any idea? whoever came up with the idea of three kernel tracks for an LTS release?
[09:59] <RoyK> damn - /boot is full -
[09:59]  * RoyK sticks his head down back in the sand
[10:08] <rbasak> RoyK: https://wiki.ubuntu.com/Kernel/LTSEnablementStack if you're not aware of it.
[10:08] <rbasak> It's to give users of newer hardware who need new kernels the LTS with the LTS userspace and just the new kernel
[11:38]  * xnox ponders if in "Investigate alternatives to mysql" session anyone mentioned PostgreSQL =)
[11:39] <jamespage> xnox, no because its not an alternative to mysql as it does not implement 'mysql'
[11:39] <jamespage> :-)
[11:39] <xnox> =)))))) i know, but would be funny if somebody writes compat for postgresql =)
[11:40]  * xnox says "what could possibly go wrong...."
[12:09] <geser> xnox: like http://andreas.scherbaum.la/blog/archives/657-PostgreSQL-9.0-Includes-the-new-MySQL-Emulation-Layer.html (but look at the date)
[12:12] <xnox> i love it!
[12:52] <oz0ne> hi, tryed to do a merge proposal from my branch. got an error "this branch is not mergeable into lp:serverguide." I did fill in Target branch (other) : ~ubuntu-core-doc/serverguide/saucy
[12:53] <oz0ne> is it affected of the naming when i pushed it? "bzr push lp:~ablacksheep/ubuntu-docs/saucy-serverguide"? or am i just doing something other wrong? ty.
[13:10] <oz0ne> i managed. : )
[14:42] <roaksoax> jamespage: yeah exactly the same issue
[14:43] <jamespage> yolanda, I think we need to disable the cinder-volume and cinder-backup DEP-8 tests - they appear to be racey
[14:44] <jamespage> if LVM is not configured then they exit out at some point in time after starting...
[14:47] <eSoul> Hey guys, I am running an Ubuntu 13.04 server box and I have a 2nd network card for my VirtualBox VMs.  Right now, I have to issue a "sudo ifconfig eth1 up" on boot to start the interface but I was wondering if there is something in /etc/network/interfaces I could put in to bring the interface up at boot.   It has no network configuration on it at all, just looking to "turn-on" the
[14:47] <eSoul> interface
[14:52] <yolanda> jamespage, ok, i will do it
[14:52] <roaksoax> jamespage: /win 15
[14:53] <roaksoax> err
[14:53] <roaksoax> sorry
[14:53] <jamespage> lol
[15:03] <roaksoax> jamespage: so destroying the environment would require to sync tools again?
[15:03] <jamespage> roaksoax, yep - raised a bug about that as well
[15:03] <jamespage> there is no 'public tools' in maas
[15:04] <roaksoax> i see
[15:08] <smoser> just fyi, jamespage utlemming anyone else interested
[15:08] <smoser> https://bugs.launchpad.net/ubuntu/+source/ubuntu-cloudimage-keyring/+bug/1218963
[16:06] <roaksoax> /q/win 15
[16:25] <banzounet> What port should I open for a mail server?
[16:25] <banzounet> 25, ... ?
[16:25] <banzounet> (Hi guys)
[16:29] <e_t_> Yes, port 25. Perhaps other ports if you want to do more than receive mail.
[16:40] <plokami666> Hello
[16:41] <plokami666> Can somebody tell me if it's possible to make users able to write to a directory that is owned by a different user but have the owner user able to read/write to and from the files uploaded by the others?
[16:42] <plokami666> Ubuntu server 12.04, going to use sftp
[16:44] <plokami666> I need to do the above because the main user has a valid shell and I don't want the people I'll be giving sftp access to being able to ssh to the server.
[16:54] <ikonia> plokami666: please don't cross post questions, you're already asking/getting help in #ubuntu
[16:54] <plokami666> I'm not really getting any help and when I posted here nobody had replied to me.
[16:56] <ikonia> ok, so please don't cross-post then
[16:56] <ikonia> find the right channel for your question and post it
[16:57] <plokami666> I wasn't aware of this channel before. I looked for it after getting silence at first in the #ubuntu one. :)
[17:34] <vedic> Hello Friends. I am trying to setup a small team of 5 to 10 people to work on a project. I am looking to buy VPS from a hosting provider and run bug tracker and repository on it. How can I ensure that the team members can only access to the bug tracker site and pull/push to server only from particular systems.
[17:35] <vedic> I use 12.04 LTS 64bit.
[17:35] <sarnold> vedic: iptables; the ufw frontend is friendly and uncomplicated.
[17:35] <vedic> I want to restrict that if they are not in office they should not be able to access the site
[17:35] <rdw200169> vedic: why not use github organizations for private repos?
[17:36] <vedic> rdw200169: We use bzr
[17:37] <vedic> sarnold: ok, I take a public IP for office internet and only enable that ip on port 443 right?
[17:37] <rdw200169> vedic: yup then ufw/iptables is the way to go; you can also make your vpc connect to your office (if you have  linux box there) thru openvpn to lock down access
[17:38] <vedic> rdw200169: Yea, we use only Ubuntu 12.04
[17:38] <rdw200169> sarnold: if your services run on port 443; i don't know what the bzr port is, but you'll need that one too
[17:38] <vedic> rdw200169: oh yea
[17:39] <rdw200169> vedic: really depends on your source ip on outbound traffic from the office; i assume the office will have a nat, and the external IP (or IPs) of that nat is what you'll you want to use on the vpc firewall
[17:41] <rdw200169> vedic: if your office users are hidden behind a nat, the only way to classify traffic firewall rules per user at the vpc would be with a vpn (since nat will hide the IP of your users)
[17:42] <rdw200169> vedic: or you could use a captive portal or something at the vpc to keep people out that don't have creds (the ways of doing this are outside the scope of this convo tho)
[17:44] <vedic> rdw200169: Currently I don't have static ip. But that is what really required I will get another net connection and the PCs in office will be behind that running on DHCP. Then on the remote server I can enable ufw (or via iptables) only the static ip on specific ports. Hope that will work out access denied from anywhere else
[17:46] <tonyyarusso> vedic: Why do you want to block access to working from home btw?
[17:47] <vedic> tonyyarusso: Because as of now I don't want to allow work from home. The access to server involves access to user data which I don't want anybody to use outside the office space
[17:48] <tonyyarusso> That didn't really answer why...
[17:49] <tonyyarusso> Do you prohibit any electronic devices from entering or leaving the office space and do bag checks to enforce that?
[17:55] <vedic> tonyyarusso: More restrictions doesn't make the argument to allow to access from anywhere.
[17:55] <vedic> tonyyarusso: There are levels of security which you can add. And crossing each level has different motives and efforts
[17:55] <sarnold> it might be easier to just hire people you can trust and give them the access they want to do their job best.
[17:56] <vedic> sarnold: Will do. But not at this stage
[17:56] <sarnold> restricting access to a limited number of IPs makes a ton of sense, no reason why the application source has to be the only line of defense..
[17:56] <sarnold> .. but letting employees work from home tends to improve moral and performance in my experience
[17:57] <vedic> sarnold: sure. I agree.
[17:57] <sarnold> (who wants to sit in the office at ten pm? no one.. but from the couch at home? sure..)
[17:57] <sarnold> but it's your staff and your data :) hehe
[17:57] <vedic> sarnold: yup :)
[17:57] <hallyn_> sarnold: no sometimes i do like being in an office at 10pm :)
[17:57] <vedic> :)
[17:57] <sarnold> hallyn_: haha :D
[18:04] <banzounet> Hey, is it possible to use apache to cover URL like domain.com:8000 and have thing like aaa.domain.com ?
[18:04] <banzounet> and still have the port 8000 block from outside
[18:04] <ikonia> mod_proxy
[18:04] <ikonia> mod_rewrite
[18:06] <banzounet> ikonia: thanks i'll google that
[18:26] <tonyyarusso> It doesn't make any sense to have "security" that doesn't actually accomplish anything except annoy people.
[18:26] <tonyyarusso> That's like locking your front door while leaving the garage door open.
[19:09] <thurstylark> I'm having trouble establishing a pptp tunnel. I'm completely new to vpn, and I want to learn, so any guides or help you can give me would be great
[19:11] <Guest36475>  hey world
[19:40] <garbagegod> Question regarding a high-volume web server setup
[19:43] <garbagegod> Can anyone recommend a host / strategy for a web server which will need to serve several hundred thousand clients a minute
[19:43] <garbagegod> I'm thinking an amazon ec2 with nginx serving flat html
[19:43] <garbagegod> Do you guys think that'll cut it?
[19:43] <garbagegod> Would you recommend using memcached?
[19:47] <tonyyarusso> garbagegod: Key question:  several hundred thousand clients a minute EVERY minute, or sometimes, with several orders of magnitude less much of the time?
[19:48] <garbagegod> most likely the latter
[19:48] <tonyyarusso> Then Amazon's stuff may be worthwhile.
[19:48] <garbagegod> I'm expecting no less than several thousand hits a minute every minute, and I'd like to be able to accomodate up to around 1 million hits/minute
[19:48] <tonyyarusso> Some coworkers are trying to set things up on Amazon, and are frustrated by the added complexity.
[19:49] <garbagegod> Do tell
[19:49] <tonyyarusso> The alternative is just spin up normal VPS systems.
[19:49] <garbagegod> I think it would be adventageous to implement a cloud based solution in this instance... a single VPS ain't gonna work
[19:49] <tonyyarusso> Never said a single VPS.
[19:49] <tonyyarusso> You can have a zillion of them if you want.
[19:50] <garbagegod> yeah, so now it's not really that much added complexity in comparison, is it ;)
[19:50] <tonyyarusso> It's whether it needs to be dynamically elastic that's complicated.
[19:51] <tonyyarusso> If it really is just flat HTML and nothing else, your solution can be equivalent to a hostname with a thousand A records and a thousand Raspberry Pis.
[19:53] <garbagegod> EC2 it is
[19:53] <tonyyarusso> Meanwhile, you should balance the relative value of predictable costs for your business.  If for some reason there was a day when you got a TRILLION hits/minute, would you prefer to pay a million times more or have the site go down?
[19:54] <tonyyarusso> I think Amazon might let you limit how big things can grow now though.
[19:54] <garbagegod> Right... Obviously 1 million would be crazy high, but I would like the security because that's the high end I was given to support
[19:54] <garbagegod> What kind of specs do you think I should get
[19:54] <garbagegod> I assume I just need to get a small instance, do some benchmarking, and scale up as needed
[19:55] <tonyyarusso> Not a damn clue :)
[19:55] <garbagegod> Yeah. Just multiplying the avg RTTs and whatnot
[19:55] <tonyyarusso> Yeah, if you have the ability to benchmark with some real-world data, that's obviously your best shot.
[19:55] <garbagegod> Okay, thanks a lot
[19:55] <tonyyarusso> Good luck - sounds like a fun project
[19:55] <garbagegod> Indeed :)
[20:57] <DammitJim> this is the weirdest thing
[20:57] <DammitJim> I installed server 12.04 lts
[20:58] <DammitJim> but when selecting the packages, I picked openssh and manual
[20:58] <DammitJim> after that screen I got a black screen where I could install stuff or upgrade stuff. I quit out of that
[20:58] <DammitJim> and finished the installation
[20:58] <Lequtix> when i install it i only choose openssh
[20:58] <Lequtix> then i install the rest once i reboot and setup IP address and hostname
[20:58] <DammitJim> now I'm trying to install a window manager because my boss can't do terminal
[20:58] <DammitJim> and I ended up adding a ubuntu-desktop ppa
[20:59] <DammitJim> but during the install, it's giving me size mismatch
[20:59] <Lequtix> email servers don't do well until u give the box a static ip and setup the hostname and FQDN
[20:59] <DammitJim> I got an email address
[20:59] <DammitJim> haven't gotten tot he email thing :D
[20:59] <Lequtix> DamnitJim ..  install ubuntu server .. only choose ssh server during setup
[20:59] <Lequtix> once you reboot
[21:00] <DammitJim> ok
[21:00] <Lequtix> set static IP.. and set FQDN
[21:00] <Lequtix> you know how?
[21:00] <DammitJim> why do I need an FQDN?
[21:00] <DammitJim> I have a static IP address
[21:00] <Lequtix> email servers work off the domain names
[21:00] <DammitJim> oh, I'll get to that
[21:00] <DammitJim> the first part I need to do is the window manager piece
[21:00] <Lequtix> well.. i recommend using webmin
[21:00] <Lequtix> instead of gui
[21:00] <DammitJim> LOL
[21:01] <Lequtix> gui adds complexity in terms of device automounting etc.
[21:02] <Lequtix> if my boss wanted a gui for his mail server i'd tell him to pony up for exchange server and server 2012
[21:02] <Lequtix> hahaha
[21:02] <Lequtix> that changes his mind
[21:02] <Lequtix> whats the point of a gui when you have to still manage exim from terminal anyhow
[21:02] <Lequtix> a terminal in a window is the same as an SSH window
[21:03] <Lequtix> with a webmin interface he can manage the server through a browser from anywhere
[21:03] <Lequtix> just my opinion
[21:04] <TJ-> Always dangerous; allowing a non-technical manager to do such a thing. Make sure and deploy git to /etc/ and a cron job to commit changes every 5 minutes or so
[21:04] <mosh> ummm
[21:04] <Lequtix> installing a gui on a server is just retarded in my opinion
[21:05] <mosh> You can also use Redmail, for a /type/ of GUI
[21:05] <Lequtix> exim4u.org also has a webmin thingie for exim only
[21:05] <mosh> but I don't know if it supports ubuntu server, I know it works on debian.
[21:06] <mosh> Which really it should support Ubuntu Server.
[21:06] <mosh> If it works on Debian.
[21:06] <DammitJim> ok, guys, I just started working here
[21:06] <Lequtix> HAHAHAHA
[21:06] <DammitJim> and I've got the ultimate decision
[21:07] <Lequtix> i feel your pain man..
[21:07] <DammitJim> LOL... I actually would have installed debian
[21:07] <Lequtix> i have had bosses like that
[21:07] <DammitJim> but he pushed for Ubuntu
[21:07] <DammitJim> LOL
[21:07] <Lequtix> ugh
[21:07] <Lequtix> so what is the mandate for the project then?
[21:07] <Lequtix> ubuntu linux, Gnome desktop, and exim4 mail server?
[21:08] <TJ-> If you go with Ubuntu you have a faster upgrade cycle. For example, if you want to host HTTPS, then 13.10 has apache 2.4 which support's perfect forward secrecy via SSLCipherSuite, which is provided by OpenSSL, and can be implemented also in Postfix, Dovecot, Stunnel, and others that require TLS
[21:09] <Lequtix> fast upgrade cycle sucks in business cus there's always better things to be doing with your tine
[21:09] <Lequtix> if it works don't fix it
[21:09] <Lequtix> my opinion
[21:09] <Lequtix> im sure i'll get yelled at
[21:09] <Lequtix> :O
[21:09] <TJ-> Not when it comes to security
[21:09] <DammitJim> actually, he listened to some other dude that has been here for over a decade and he said lxde
[21:10] <Lequtix> i keep everything behind a firewal.. only necessisary ports are exposed
[21:10] <DammitJim> no need to be on the fast lane for this project
[21:10] <DammitJim> dammit... I have this size mismatch crap going on
[21:10] <Lequtix> lxde is nice
[21:10] <DammitJim> doing apt-get upgrade
[21:10] <Lequtix> try apt-get -uy dist-upgrade
[21:10] <DammitJim> why dist-upgrade?
[21:10] <Lequtix> maybe  clean it first
[21:10] <Lequtix> apt-get --purge clean-all
[21:11] <Lequtix> apt-get --purge clean all
[21:11] <Lequtix> i forget exactly what it is
[21:11] <Lequtix> i find the dist-upgrade works better for me
[21:11] <TJ-> "apt-get clean" will remove the packages in /var/cache/apt/archive/.
[21:12] <Lequtix> yea.. he's having issues with the packages and the database i think
[21:12] <Lequtix> clean it all out.. re-update
[21:12] <Lequtix> then re-upgrade
[21:12] <Lequtix> see if that solves the upgrade issues
[21:13] <Lequtix> doesn't it clean out the repository database too?
[21:15] <DammitJim> the clean didn't work
[21:15] <DammitJim> ugh, I'm out of time
[21:15] <DammitJim> I'll have to drill this on Monday
[21:15] <DammitJim> no, Tuesday
[21:15] <Lequtix> i would do a fresh install then
[21:15] <DammitJim> yup
[21:16] <Lequtix> save urself some time
[21:16] <DammitJim> I'm changing the settings in vmware to boot from iso
[21:16] <TJ-> You should pastebin the error messages
[21:16] <Lequtix> so did ur boss mandate how you are to partition it too?
[21:16] <Lequtix> so this is a VM?????
[21:16] <Lequtix> why the hell put a gui on a fricking VM
[21:16] <Lequtix> thats even more stupid
[21:16] <Lequtix> no offence
[21:16] <Lequtix> lol
[21:17] <TJ-> it's offensive to put a GUI in a VM! :D
[21:17] <Lequtix> its completely pointless
[21:17] <Lequtix> first you need a windows workstation to run the vsphere client
[21:17] <Lequtix> secondly the performance sucks
[21:17] <Lequtix> thirdly the vmware tools are a pain in the ass
[21:18] <Lequtix> Quit that job Jim .. find another one
[21:18] <Lequtix> lol
[21:19] <Lequtix> just kidding..
[21:20] <Lequtix> i'm sure there's a vmware appliance for a standard ubuntu 12.04 LTS install
[21:20] <Lequtix> already packaged
[21:23] <DammitJim> weird... what is that manual package selection option where one picks openssh ?
[21:23] <Lequtix> just don't pick anything
[21:23] <DammitJim> should I select that or install ubuntu-desktop later?
[21:23] <DammitJim> when do I do the ubuntu-desktop piece?
[21:23] <Lequtix> you want the desktop for sure?
[21:23] <Lequtix> in a vm?
[21:23] <DammitJim> yes
[21:23] <DammitJim> it sucks
[21:23] <Lequtix> ok..  install it then
[21:23] <DammitJim> but I gotta do it
[21:23] <DammitJim> but how is the question
[21:23] <Lequtix> let the installer do it for u
[21:24] <Lequtix> it's not in tasksel?
[21:24] <DammitJim> tasksel
[21:24] <DammitJim> where is that?
[21:24] <Lequtix> maybe that's debian
[21:24] <DammitJim> where you pick the packages like openssh mail samba, etc?
[21:24] <Lequtix> ya
[21:24] <DammitJim> on debian you at least see a window manager option LOL
[21:24] <Lequtix> ok.. don't pick anything
[21:24] <DammitJim> here it just says manual selection or something
[21:24] <Lequtix> just install base ubuntu
[21:25] <Lequtix> then when u get into the os..  configure all the networking and FQDN
[21:25] <Lequtix> then update/upgrade
[21:25] <Lequtix> then apt-get install tasksel
[21:25] <Lequtix> then sudo tasksel
[21:25] <Lequtix> it will give u the software selection menu that you are used to
[21:25] <Lequtix> https://help.ubuntu.com/community/Tasksel
[21:26] <Lequtix> you should probably install the vmware tools first tho
[21:26] <Lequtix> before installing everything
[21:32] <DammitJim> thanks
[21:32] <DammitJim> running