[00:02] <jvargas> rockets: found a solution in: http://wiki.debian.org/NetworkConfiguration#Bringing_up_an_interface_without_an_IP_address
[00:10] <rockets> jvargas, it's working fine for me *right now* just using auto eth1
[01:27] <tkeith> After upgrading to Ubuntu 12.04, all my MySQL servers have an empty "test" database. Is that normal?
[01:40] <fosterdv> That don't sound normal to me....
[01:40] <fosterdv> Are you missing anything else?
[02:15] <cowboy> hello
[02:15] <EvilResistance> hiya
[02:16] <cowboy> anyone there pls point me to specific ubuntu server talking specific top pls?
[02:16] <cowboy> I like to talk about webpage development that can access Ubuntu server
[02:17] <cowboy> hello
[02:25] <cowboy> hello
[02:25] <cowboy> anyone there?
[02:25] <cowboy> help
[02:26] <cowboy> any expert ubuntu server help solve problems here...FORUM???
[02:27] <jsmith-argotec> cowboy: best off asking your question and waiting a bit and see if anyone responds
[02:28] <cowboy> okie..thanx jsmith
[02:29] <cowboy> beside install applications on chat ircd..is there a forum that help me step by steps from webpage to server...pls help
[02:40] <twb> cowboy: I don't understand your question.
[02:55] <sparc_> Hey hey.  Is reprepro still the way to manage your own apt-repositories?
[02:55] <sparc_> or maybe there are better tools now ...
[02:56] <mgw> sparc_: you just want a few .debs in a repo?
[02:57] <sparc_> yeah, there will probably be 20 to 30 of components of our in-house apps
[02:57] <sparc_> some perl modulse converted to packages, some whole web apps, some standalone python daemons
[02:57] <mgw> let me paste something for you
[02:57] <sparc_> i'm not trying to mirror any of ubuntu's major repos
[02:58] <sparc_> ok cool, thanks for the input!
[03:00] <mgw> https://gist.github.com/5faccb703054afdbbcae
[03:01] <mgw> then just use apache to serve that directory
[03:01] <sparc_> interesting, it's using dpkg to operate on an existing repo ...
[03:01] <mgw> yeah… but you don't need the whole dpkg-devel pacakge
[03:01] <mgw> one sec
[03:01] <sparc_> yeah, that makes sense.  you make the repo and then serve it over http
[03:02] <sparc_> creating it initially and then modifying the contents, are the processes that probably need management
[03:02] <sparc_> reprepro did that i think ...
[03:02] <qman__> that's what his script does
[03:02] <qman__> just dump the .debs in that directory and run it
[03:02] <sparc_> oooh
[03:02] <sparc_> ok
[03:03] <mgw> yeah, exactly, what qman_ said
[03:03] <qman__> it generates the repository based on what debs it finds
[03:03] <mgw> you can of course script that too
[03:03] <qman__> yeah, could cron it or automate it some other way
[03:03] <mgw> You'll have to set up gpg, or comment out that part and deal with an unauthenticated repo
[03:04] <sparc_> no problem, this looks pretty cool
[03:04] <mgw> I'm going to use ansible to script upload & run (could also use fabric or some such)
[03:04] <fosterdv> quit
[03:04] <sparc_> thsi blows away the repo every time?  i guess i should read the script
[03:04] <fosterdv> sorry
[03:04] <mgw> sparc_: just the Release and Packages files
[03:04] <sparc_> aah okie
[03:05] <qman__> it overwrites it with what's there
[03:05] <mgw> for 20 or 30 debs that shouldn't matter
[03:05] <twb> dpkg-scanpackages should not be used in that script; the apt-ftparchive script should simply be called multiple times
[03:06] <mgw> twb: thanks, I'll update it
[03:06] <twb> http://paste.debian.net/167967/ is mine, I run it in cron.hourly.  A beginner should stick to reprepro.  The apt-ftparchive technique is useful only if you are supporting a single release.  if you have >1 release it becomes unpleasant quickly.
[03:07] <twb> I also strongly recommend adding NotAutomatic: yes to your Release until you're confident it's production-ready.
[03:07] <twb> That attribute will cause it to be pinned to -1, the same as Debian's experimental repo.  IOW stuff from it won't accidentally be installed.
[03:10] <sparc_> thanks much for the input and background you all.  it's much appreciated.
[03:10] <twb> No worries.
[03:14] <pmp6nl> Hey everyone.  I am looking to backup my server to my home computer.  What are good folders to rsync, what is not needed? Thanks
[03:17] <twb> pmp6nl: /
[03:18] <pmp6nl> twb, that was my initial thought.  But thought I would check
[03:18] <pmp6nl> Thanks
[03:18] <twb> pmp6nl: if you care about user data, you want /home and maybe some things like /var/mail, depending on what services you're using.
[03:18] <twb> pmp6nl: for the OS itself, /etc is critical, and databases and things generally live in /var.  But if you have the space, back it all up.
[03:18] <qman__> it depends on what's important to you
[03:18] <twb> pmp6nl: recommend you use -xdev and list filesystems explicitly, to avoid backing up things like /proc/kcore which is 16TB
[03:19] <qman__>  /etc and a list of installed packages are needed to rebuild a lost server, /var is usually a good idea
[03:19] <qman__> but if you want everything, / is it
[03:19] <twb> -xdev is more reliable than --exclude, although -xdev leads to false negatives instead of false positives.  YMMV etc.
[03:19] <twb> Also strongly recommend you deploy rsnapshot or similar, so you get incremental backups for free.
[03:20] <pmp6nl> twb and qman__ ok thanks.  Right now I am rsyncing / .... so I should skip /proc?
[03:20] <twb> Definitely
[03:20] <qman__> yes
[03:20] <pmp6nl> woops.
[03:20] <twb> And probably some other things like /sys and /dev and /run
[03:20] <qman__> and /dev, and all the other virtual and temporary filesystems
[03:20] <twb> Which is why I said just use -xdev
[03:20] <pmp6nl> so -xdev will skip /dev and /proc
[03:21] <twb> xdev limits it to the specified filesystem
[03:21] <twb> So e.g. if / and /home are separate and you rsync -xdev /, it will skip /proc but it will also skip /home
[03:21] <qman__> yeah, you limit it to one filesystem, and configure a backup for each filesystem you want backed up
[03:21] <twb> RTFM is advisable, too :-)
[03:22] <pmp6nl> ah ok. So I should do separate for each filesystem ... /home /var etc
[03:22] <qman__> if you have those as separate filesystems, yeah
[03:22] <linocisco> I want to have myname@ubuntu.com
[03:22] <qman__> IMO it's easier to do it that way then to attempt to exclude all the other filesystems and locations you don't need
[03:23] <twb> Agreed.
[03:23] <qman__> because there's only a few filesystems to back up (sometimes just one) while there's dozens of things to exclude
[03:24] <pmp6nl> ok, html/php files etc. are in /home    databases are in /var  ... what is in all of the other random folders?  Anything else important to backup?
[03:25] <qman__> no, user data is in /home, which may or may not contain html and php files depending on your configuration
[03:25] <qman__>  /etc is crucial, it contains all the config for the server, otherwise you'll be reconfiguring from scratch
[03:26] <twb> qman__: if /home contains a web app, you screwed up
[03:26] <twb> pmp6nl: note that you cannot rsync RDBMS databases, because they are not guaranteed coherent/quiescent
[03:26] <qman__> it shouldn't, but I was referring to userdir config
[03:26] <twb> pmp6nl: to back up a db correctly you must dump it to .sql and rsync that
[03:26] <pmp6nl> qman__, twb I have public_html in home  I will add /etc.  Any others?  RDBMS?
[03:27] <twb> I also strongly recommend etckeeper
[03:27] <qman__> yeah, etckeeper is cool
[03:27] <qman__> I haven't actually tried to restore anything from it, but auto saving everything at the right times is great
[03:28] <twb> qman__: it's not about restore, it's about blame
[03:30] <qman__> having a nice list of everything that changed after updating is nice too
[03:41] <linocisco> I want to have myname@ubuntu.com
[03:43] <pmp6nl> twb, and qman__ ok. thanks. :)
[03:43] <pmp6nl> Hello, is it possible to rename a site in /etc/apache2/sites-enabled ?  If so how?  Thanks
[03:44] <twb> Go edit the file?
[03:45] <pmp6nl> I mean the name of it.  Its called example.com
[03:45] <twb> So rename the file?
[03:47] <pmp6nl> twb, its that easy.  Is the command just rename?
[03:49] <twb> No mv
[03:49] <twb> You probably want to a2dissite first, edit the master version in sites-available, then a2ensite it
[04:53] <linocisco> hi all
[05:23] <weallknowit> morgen :)
[06:41] <linocisco> I want to have myname@ubuntu.com
[06:42] <linocisco> i have two VSAT networks on each office. according to Teleport's ISP infrastrure, we can ping each other with ping time 2000ms. I would like to know how to establish VPN between it
[07:09] <linocisco> i have two VSAT networks on each office. according to Teleport's ISP infrastrure, we can ping each other with ping time 2000ms. I would like to know how to establish VPN between it
[07:10] <owh> linocisco: With great patience - having done it myself.
[07:10] <owh> linocisco: You need to make sure that the VSAT isn't blocking VPN traffic which some do. I ended up using an ssh tunnel.
[07:11] <linocisco> owh, really? where is your country?
[07:11] <owh> linocisco: AUS
[07:11] <linocisco> owh, how can I make sure or check if VSAT or ISP is blocking VPN traffic?
[07:11] <owh> In my experience the quickest way is to ask them.
[07:12] <owh> Normally the IP range is internal to the VSAT operator. If both VSAT terminals are from the same provider. If not, you generally have a problem because they generally don't expose the IP address to the outside world.
[07:12] <linocisco> owh, here it is not legal to ask official. but according to their struture, ping is ok and some org said they can establish VPN between two VSAT sites
[07:13] <owh> I'd start with attempting to ssh to the other end.
[07:13] <owh> What traffic are you hoping to share?
[07:13] <linocisco> owh, boths sites are from the same ISP
[07:14] <linocisco> owh, web, mail, telephony
[07:14] <owh> Yeah, that's really not going to work well.
[07:14] <linocisco> owh, the priority is web and mail.
[07:15] <owh> What is happening for most VSAT traffic is that it goes from your terminal to the satellite, then down to the ground station, then back up to the satellite, then to the other terminal. Gilat offers VSAT to VSAT functionality, but most countries block that because it prevents monitoring.
[07:17] <owh> (Which is why you're seeing 2 second ping times)
[07:34] <linocisco> owh, yes
[08:11] <ome> Does ubuntu comes with a preinstalled alternative to 'sysv-rc-conf or sysvconfig' ?
[08:13] <ikonia> ome: it uses upstart, so it works different
[09:50] <railsraider> what should i check for if a 8 servers on rackspace went down all on the same time? can it be d-dos attack?
[09:52] <railsraider> i see a lot of AIF:PRIV connect attempt: on th elogs
[10:08] <ikonia> railsraider: contact rackspace and ask them what happened
[10:15] <sonne> greetings!
[10:15] <sonne> what is the ubuntu equivalent of debian's "locales-all"?
[11:34] <sonne> and.. anyone knows why cron is started when building a chroot?
[11:35] <sonne> i think it has something to do with upstart not respecting policy-rc.d
[11:44] <sonne> i think i got it
[11:57] <brainysmurf> What sort of tools are available if I want to modify headers of emails sent to particular addresses. My ubuntu server is using postfix
[11:57] <sonne> to anyone that might be interested in what i was tlaking about: https://bugs.launchpad.net/ubuntu/+source/xen-tools/+bug/997063
[11:58] <mardraum> brainysmurf: modify them how?
[11:59] <mardraum> your mua can modify the usual things. otherwise people will generally assume you are a spamming dick.
[12:00] <brainysmurf> My boss wants "to: group@" changed to "to: undisclosed"
[12:01] <mardraum> just send to bcc
[12:01] <brainysmurf> mardram: I know...
[12:01] <mardraum> oh?
[12:02] <brainysmurf> We don't want this group's addy publicized, and users have been known to avoid the bcc like the plague
[12:03] <brainysmurf> Just wondering if there is some sort of filtering tool that can be used (and that wouldn't make you look like a spamming dick)
[12:03] <mardraum> you want to give users your address, but then hide it from them?
[12:03] <brainysmurf> (I work at a very busy school)
[12:04] <brainysmurf> This address is actually an alias, not a human account
[12:04] <mardraum> an address is an address. if the mta accepts mail, it's all the same thing.
[12:05] <mardraum> I guess I don't really have a clue what you are trying to do that bcc doesn't already do.
[12:06] <brainysmurf> hehe I share your frustration :) I'm trying to make it so that if I send to group@ (which is an alias to several people) when the indiviuals get it they don't see the address
[12:07] <brainysmurf> I know that bcc is what we're supposed to use, but my users keep insisting on using to
[12:07] <mardraum> they won't see it, the MTA expands it
[12:07] <mardraum> they'll see you address, and their address
[12:07] <mardraum> your*
[12:08] <brainysmurf> I tested this. I see the group address though.
[12:08] <mardraum> right now I suspect you want #postfix
[12:09] <mardraum> if that's the MTA you are choosing to handle the mail
[12:09] <brainysmurf> I am using postfix
[12:09] <brainysmurf> they said that postfix doesn't touch the content of the email, I need something else
[12:11] <mardraum> you're probably going to have the most luck telling your boss that your MUAs need to be configured to send from a "donotreply" address
[12:11] <mardraum> otherwise you'll need to use some sort of custom software that can do the expansion itself
[12:12] <brainysmurf> Okay thanks mardraum that's what I thought
[12:13] <mardraum> find a new job :p
[12:26] <sanderj_> Is there any known prodecure to follow.. to upgrade from 10.04 to 12.04?
[12:27] <sanderj_> ? upgrade
[12:29] <koolhead17> !upgrade
[15:06] <rockets> Psi-Jack, So, I have my old virtual machine imported and booted via kvm now, and I have it set up with a bridged interface, but for the life of me I cannot get networking working properly. Also, virt-manager doesn't actually show the "br0" I set up as an available interface, it still only shows "Default". I had to enter it manually.
[15:07] <rockets> I can ping the bridge, but nothing using it is reachable.
[15:17] <Psi-Jack> rockets: YOu can ping the host's endpoint IP?
[15:17] <rockets> Psi-Jack, I can ping the ip I assigned to the bridge interface
[15:18] <Psi-Jack> Which is on the host OS?
[15:18] <rockets> yes
[15:18] <Psi-Jack> Okay, then it's likely a networking or firewall issue in your guest.
[15:18] <rockets> but i cannot reach the IP assigned internally to the guest
[15:18] <Psi-Jack> Or possibly even host if you're running firewall there.
[15:18] <rockets> no firewall
[15:19] <Psi-Jack> OKay, so what exactly did you do, network-wise?
[15:19] <rockets> I did this:
[15:19] <rockets> http://pastie.org/3884753
[15:20] <Psi-Jack> OKay. That tells me literally nothing.
[15:20] <rockets> And the guest is set to use br0
[15:20] <rockets> which I set by using virt-manager
[15:20] <Psi-Jack> ok
[15:21] <rockets> and then inside the guest I assigned an IP to eth0 as I would normally and to eth1 as i would normally
[15:21] <rockets> and neither are reachable
[15:21] <Psi-Jack> "normally?"
[15:21] <rockets> As I would in any normal ubuntu linux host, via /etc/network/interface
[15:21] <rockets> auto eth0, iface etc etc
[15:21] <Psi-Jack> Uh huh.
[15:21] <Psi-Jack> Think of me as a blind man.
[15:21] <Psi-Jack> DESCRIBE EVERYTHING!
[15:21] <rockets> lol
[15:22] <rockets> inside the guest there is a standard interfaces file describing the network config
[15:22] <rockets> just like this: http://pastie.org/3884762
[15:22] <Psi-Jack> I only see eth0 in this.
[15:23] <rockets> That's correct, I'm just giving you an example
[15:23] <rockets> hold on
[15:23]  * Psi-Jack closes that "Example." Don't give me examples. I'm a blind man.
[15:23] <Psi-Jack> I know what a Debian interfaces file is. :p
[15:24] <rockets> I know you do!
[15:24] <rockets> Hold on, I'll get the guest's interfaces file
[15:29] <Psi-Jack> Welp, taking too long, and it's lunch time. Though I will be back afterwards.
[15:36] <rockets> Psi-Jack, sorry heh. I really do appreciate your time. I'm going to try a few things and come back in a bit.
[15:40] <rockets> Psi-Jack, ok, so here's something more concrete for you. I'm attempting to set up a brand new virtual machine. I set it up using virt-manager to use the br0 bridge for the guest's ethernet cards. E.g. http://i.imgur.com/Hn6vC.png
[15:41] <rockets> And yet the guest vm (which is running the ubuntu 12.04 server installer) cannot get an IP via dhcp. This is during the install process, no actual OS installed yet. My network definitely has a DHCP server, we're all using it.
[15:41] <rockets> So I'm sure I have the bridge set up wrong somehow, I just don't know how.
[15:43] <billybigr> hey all, anyone alive? just need to talk to someone about the serverguide and openvpn
[16:41]  * Psi-Jack sighs/.
[16:41] <Psi-Jack> Seriously wish people would stop giving their "status" away through nick changes.
[16:42] <patdk-wk> changes nick to patdk-wk-away-from-psi-jack
[16:54] <billybigr> can someone here help me with an openvpn install problem?
[16:54] <billybigr> trying to follow the serverguide and i hit a wall
[16:55] <billybigr> trying to contact someone on the -doc team isn't working either
[18:07] <bananapie> I have a server running ntpd. The clock is currently 4 seconds out of alignment. I want to see a status from ntpd to make sure that ntpd knows about the descrepancy and how long it will take before the clock is resynchronized
[18:10] <RoyK> bananapie: first, check the logs
[18:11] <bananapie> I checked both debug, boot.log, syslog and dmesg. I don't see anything of used in there
[18:11] <bananapie> other than max descriptors, but that's not useful
[18:15] <bananapie> I uncommented the statsdir line
[18:15] <bananapie> I can see that my server is now aware of the 4 second difference
[18:18] <bananapie> But I don't see where it says how long it takes to resynch
[18:30] <Pici> bananapie: ntpdc -p and ntpq -p  should be helpful.
[18:31] <Pici> Its been forever since I needed to fix that on my server, but I remember those outputs being very handy.
[18:32] <bananapie> ok
[18:32] <bananapie> it's the first time it's happened to me in about 3 years
[18:33] <cultavix> what are you trying to do
[18:34] <RamJett> If I have 4 ethernet interfaces and I want one to be set to dhcp but not change the, name servers, default gateway, ntp.conf, etc. How would I do that? Basically I just want it to set the IP address only
[18:34] <patdk-wk> RamJett, edit dhclient.conf
[18:37] <RamJett> There is no way to do it per interface. Like in the interfaces file? It has to be a global dhcpclient ?
[18:38] <rockets> What's a good way of backing up live kvm VMs?
[18:41] <pdtpatrick> RamJett, you just want one interface to be dhcp right?
[18:42] <pdtpatrick> RamJett, in /etc/network/interfaces - just set that interface to something like "iface ethX inet dhcp"
[18:42] <pdtpatrick> RamJett, or you can do "dhclient ethX"
[18:42] <patdk-wk> pdtpatrick, one interface with dhcp, for IP only, no dns/ntp/....
[18:42] <patdk-wk> no gateway even
[18:44] <pdtpatrick> patdk-wk, good catch. Thanks
[18:47] <RamJett> Right. I want to pick which interface does the degalt gateway etc .. Like in Gentoo you just do eth0_dhcp="nodns,nogateway". I think I understabd how to do in in the dhclient.conf file. But is I use dhcp on eth1 but want it to use the gateway .. I just globally disabled that
[18:47] <pdtpatrick> i don't understand that last sentence
[18:48] <pdtpatrick> nvm s/is/if
[18:49] <RamJett> if I do it in the dhcpclient.conf. I'm doing it on all interfaces, right ?
[18:49] <RamJett> Sorry bout the typing
[18:50] <Pici> RamJett: Per the manpage use: interface "name" { declarations ...  }
[18:50] <patdk-wk> that depends on *how* you modify dhclient.conf
[18:50] <pdtpatrick> is it possible to use post-up if u use inet dhcp?
[18:50] <Pici> man 5 dhclient.conf
[18:50] <RamJett> ok. Didn't see you could do it per interface. Thanks
[19:47] <axisys> how to clean up /boot ? it is 86% now
[19:48] <axisys> what's the recom way to remove old kernels?
[19:48] <axisys> i want to remove all but last 3 kernels
[19:48] <axisys> i think i have 10+ of them now
[19:49] <ogra_> dpkg -l|grep linux-image ....
[19:50] <ogra_> and remove the old versioned packages you dont need
[19:58] <axisys> whats the one without numbers?
[19:58] <axisys> http://dpaste.com/745683/
[20:26] <jmedina> hi there
[20:26] <jmedina> anyone is able to use vmbuilder in precise with lvm storage?
[21:10] <resno> is 10x1 fast enough to run a few sites?
[21:20] <jmedina> I cant create machines with file images
[21:20] <jmedina> but not with lvm :S
[21:43] <jmedina> anyone here using kvm in 12.04 server?
[21:46] <StevenR> jmedina: yes
[21:46] <jmedina> StevenR: do you use lvm for vm storage?
[21:46] <jmedina> In trying to create a vm using vmbuilder with --raw=/dev/vg/lvx and fails
[21:47] <jmedina> with image files works fine, it only fails with lvm
[21:47] <jmedina> I have seen a few bugs with similar errors
[21:50] <jmedina> Im using this command:
[21:50] <jmedina> vmbuilder kvm ubuntu --hostname server5 --suite precise --flavour server --arch i386  -o --libvirt qemu:///system --raw=/dev/poolb2/storage5 --addpkg=acpid --addpkg=openssh-server --user=administrador --pass=123456 --tmpfs=-
[21:55] <StevenR> jmedina: sorry, I don't use LVM with it.
[21:55] <jmedina> StevenR: thanks, I just tried the same in a second machine and same result