[00:16] <twb> Urk, someone forgot to switch etckeeper to git on this box
[00:17] <twb> No --color either :-/
[00:18] <twb> Fortunately I have colordiff(1) on another box...
[00:18] <Veovis_Muaddib> I'm PuTTYing in to my server to create a few symbolic links to keep my television organized.  I made them all before realizing that bash shows them red.  My television directory is ~/Videos/Television/All/ and I'd like to have  ~/Videos/Television/Airing/ and ~/Videos/Television/Ended/, to start with.  But inside of the All directory, " $ ln -s Burn\ Notice/ ../Airing/ " Doesn't seem to work.  ls ../Airing shows the link but
[00:18] <Veovis_Muaddib> it is red, and SMB shares don't see the link.  Permissions are fine for all subdirectories of ~/Videos/, these would be the first symbolic links in ~/Videos/, and ~/Videos/ is not a symbolic link itself.  Suggestions?
[00:19] <twb> Veovis_Muaddib: you can't link to ~; link destinations don't expand ~
[00:19] <twb> Veovis_Muaddib: the red is because it's a broken link
[00:19] <twb> Veovis_Muaddib: you need to link to /home/muaddib/Videos instead of ~/Videos, for example
[00:19] <Veovis_Muaddib> twb: The exact command is $ ln -s SHOW/ ../Airing
[00:20] <Veovis_Muaddib> No Tilde involved
[00:20] <twb> You're probably still making broken links; try doing the ln from within SHOW
[00:21] <Veovis_Muaddib> veovis@Lambert:~/Videos/Television$ ln -s Burn\ Notice/ ../Airing/
[00:21] <Veovis_Muaddib> That's what I've been using
[00:22] <Veovis_Muaddib> Sorry, forgot the Airing after Television
[00:22] <Veovis_Muaddib> All I mean
[00:22] <Veovis_Muaddib> veovis@Lambert:~/Videos/Television/All$ ln -s Burn\ Notice/ ../Airing/
[00:22] <qman__> do an ls -l, it'll tell you exactly where the links are pointing
[00:22] <twb> Good idea
[00:23] <Veovis_Muaddib> Burn Notice -> Burn Notice/  As though it ignored the ../ when choosing where to point
[00:24] <Veovis_Muaddib> Okay, let me try up a directory
[00:25] <qman__> do it from the Airing directory
[00:26] <Veovis_Muaddib> Did that just as it was recommended.  It works.  Never had to do that before, any idea why now?
[00:26] <qman__> I don't know exactly why, and I don't recall it behaving this way for me in the bast
[00:26] <qman__> past*
[00:26] <Veovis_Muaddib> And thanks for the ls -l, didn't know that one
[00:26] <qman__> but it's taking your first argument literally, instead of relatively
[00:26] <Veovis_Muaddib> Yeah, that's pretty odd
[00:26] <Veovis_Muaddib> Thanks though, that worked
[00:28] <twb> Veovis_Muaddib: uh, you realize "ln -s foo bar" will create a link "bar -> foo" NOT a link "foo -> bar" ?
[00:29] <twb> I think either you're getting the args the wrong way round or forgetting that the link destination is relative to the link, not relative to pwd
[00:29] <Veovis_Muaddib> Yeah, probably, I'm pretty tired today.  :P  I bet that's it
[00:30] <Veovis_Muaddib> One of those two
[01:58] <idlemind324> I'm looking for recommendations for what I should use for file sharing on an all Linux network. SSH (SFTP/SCP) or NFS v4
[02:08] <twb> idlemind324: do you trust your network?
[02:09] <idlemind324> twb: i'd say this is a trusted network. it is attached to the Internet but it's behind an IPTables firewall box so I could control any traffic that might try and sneak out
[02:09] <twb> SFTP and kerberized NFSv4 would both be OK, but unkerberized NFSv4 shouldn't be run over an untrusted network.
[02:09] <idlemind324> ahh. kerberized nfsv4 ... i need to learn about the changes w/nfsv4
[02:10] <twb> All network filesystems have locking problems, so you're no worse off there.  SFTP is *really* easy to deploy, but if you intend to be shoving e.g. videos over it, NFS would be a better bet.
[02:10] <twb> The pathological case for NFS is untarring a tarball of lots of very small files, or e.g. a Maildir mailbox.
[02:13] <twb> unkerberized NFSv4 is slightly more painful than NFS3; kerberos is a major PITA
[02:14] <twb> For a trusted network where throughput is key I would probably use NFSv3 -- unless you had to also support non-Unix systems
[02:15] <qman__> for cross-platform support samba is generally the easiest and a pretty good compromise
[02:15] <twb> Yeah
[02:16] <qman__> in ideal conditions you can achieve 80MB/s over gigabit with it
[02:16] <twb> Although FWIW I just told my users "I'm not putting up with machine accounts or giving samba a root binddn, so you bitches can all just use SFTP and bloody well like it"
[02:18] <qman__> I use all three of the aforementioned on my LAN
[02:18] <twb> :-)
[02:18] <qman__> well, NFSv3
[02:18] <twb> I also said "I am *not* supporting >1 network filesystem"
[02:19] <twb> Well, technically I did cordon off part of the LAN and designate it "server only" and I run NFSv3 over that.
[02:19] <twb> Because otherwise how do you get the authorized_keys files distributed to allow SFTP? ;-)
[02:20] <qman__> I just don't have enough systems to justify the work setting up centralized authentication
[02:21] <qman__> or users, for that matter
[02:29] <patdk-lap> hmm, I dunno how people run unkerberized nfs4, it's not just completely useless
[02:32] <twb> patdk-lap: well, I got to the point where it *looked* like it should work
[02:32] <twb> The SMB2 protocol sounds pretty interesting, but it's no use to me until linux has a client and samba has a server
[02:35] <patdk-lap> well, nfs4 is only useful when using kerberos
[02:35] <patdk-lap> and I haven't gotten it to work yet, so I'm still not sure of the limitations
[02:35] <patdk-lap> so till I have more time for testing, I still have to use nfs3
[02:36] <twb> How so?
[02:36] <patdk-lap> making uid/usernames match up across nfs4 is a pain in the ass, and requires kerberos
[02:36] <twb> I mean on paper NFSv4, even unkerberized, sounds better than NFSv3
[02:36] <twb> Hum
[02:36] <patdk-lap> why? unkerberised nfs4 removes all usernames and uses a nobodynfs user
[02:36] <patdk-lap> that completely destroys any cross user permissions and usability
[02:36] <twb> Because it's connection-oriented.  But I didn't know that ^^
[02:37] <patdk-lap> nfs4 uses usernames cross everything, requireing kerberos auth to verify
[02:37] <twb> Fair enough, then
[02:37] <patdk-lap> then it requires uid's to be in sync also, cause they didn't think the username though totally :)
[02:38] <patdk-lap> it's just so painless for me to sync uid, and just mount nfs3
[02:39] <twb> Of course NFSv3 also provides zero protections permission-wise, unless you can guarantee users can't get root on any existing box, nor bring in a box on which they have root...
[02:39] <patdk-lap> well, bringing in a box, is easy
[02:39] <twb> Not in a prison :P
[02:39] <patdk-lap> can't bring a box into a locked room :)
[02:40] <patdk-lap> as for root, that is what monitoring is for
[02:40] <patdk-lap> but ya, nfs4 attempts to fix the *root* issue, while not really fixing it at all, and breaking all other things
[02:42] <patdk-lap> so far I have spent a week playing with nfs4, without any real success
[02:42] <patdk-lap> but I also have solaris in that mix, and found out linux nfs4 and solaris nfs4 don't like eah other at all
[02:42] <twb> oh ow
[02:42] <twb> good luck with THAT
[02:43] <patdk-lap> isn't mainly that solaris kerberos isn't compatable with linux kerberos mit
[02:43] <patdk-lap> and solaris kerberos is default to a much higher security level
[02:44] <patdk-lap> and that was about where I said, this has taken too long, time to stop playing and work on other things
[04:39] <linocisco> how to configure openfire to use through VSAT link with high latency ?
[05:01] <twb> Haha, he wanted to run an XMPP server over a satellite link
[05:01] <EvilResistance> lool
[07:03] <SpamapS> anybody want to try pounding on this website? curious to see how much punishment it can take...
[07:03] <SpamapS> http://ec2-50-16-128-14.compute-1.amazonaws.com/
[07:23] <Corey> SpamapS: Urm... that sounds a bit sketchy. :-)
[07:23] <SpamapS> Haha
[07:23] <SpamapS> I swear I won't pwn you w/ XSS
[07:23] <SpamapS> :)
[07:24] <SpamapS> What would be the point? I can already PWN you w/ SRU's ;)
[07:29] <SpamapS> ./config-defaults.php:$databasetabletype  =   'myISAM';	   // Storage engine mysql should use when creating survey results tables and token tables (if mysql is used). Default is myISAM.
[07:29] <SpamapS> *sigh*
[07:29] <SpamapS> a "modern" app that still uses that waste of space myisam. :-/
[07:38] <onre> myisam is painful
[07:38] <SpamapS> its seriously useless
[07:38] <onre> agreed
[07:42] <SpamapS> at least in 5.5, InnoDB is the default
[09:03] <jamespage> morning all
[09:14] <xranby> jamespage: morning
[09:15] <jamespage> morning xranby
[09:15] <jamespage> apologies but I have not found time to try that new kernel out yet
[09:27] <lynxman> morning o/
[09:27] <jamespage> morning lynxman
[09:27] <lynxman> jamespage: morning sir :)
[09:34] <koolhead11> hi all
[09:44] <auston> hi, i have 2 NICs installed on server act as a firewall. eth0 is connected to router and eth1 is on LAN. Both using ip forwarding between the 2 NICs. How to block all incoming traffic from internet?
[09:45] <auston> Using iptables, i have tried only can block the server itself but client connect to it still can access internet.
[09:49] <lynxman> auston: if you want to block clients you need to add a FORWARDING rule
[09:50] <lynxman> auston: since the traffic through your bastion host is forwarded to the clients, the input, output will only affect the host itself
[09:51] <auston> what is the best command or method to use for filtering the forwarding rule?
[09:52] <lynxman> auston: I'd recommend you have a look at the firewall guide https://help.ubuntu.com/11.10/serverguide/C/firewall.html
[09:53] <auston> i have no idea why when I enable ufw, after restarted, ufw become inactive.
[10:04] <koolhead11> smoser: around
[10:06] <auston> thanks. lynxman. I'll try out the FORWARDING rule.\
[10:06] <lynxman> koolhead11: way too early!
[10:06] <lynxman> auston: best of lucks :) it's in default iptables like INPUT and OUTPUT there's FORWARD too
[10:06] <koolhead11> lynxman: ooh okey. :P
[10:07] <auston> I got it :)
[11:19] <Ursinha> good morning :)
[11:21] <vAx1> hello
[11:22] <vAx1> need some help with syslog-ng 3.3.1 (compiled for my 10.04 LTS)
[11:22] <vAx1> someone?
[11:48] <soren> vAx1: Noone can help you if you don't ask a question.
[11:53] <patdk-lap> vax1, works fine here
[12:03] <vAx1> hey, that was just tu knock ;)
[12:03] <vAx1> to
[12:03] <soren> Don't.
[12:03] <vAx1> so... my problem
[12:03] <soren> It's a waste of time.
[12:04] <vAx1> syslog-ng works fine after upgrading my conf
[12:04] <vAx1> but, when I stop it, the subprocess (lanched by the program() destination)
[12:05] <vAx1> dont stop
[12:06] <vAx1> I change the option tu start syslog withous the "supervising"
[12:06] <vAx1> so, just in background... and it's the same
[12:06] <vAx1> still have my 2 process (perl) running
[12:08] <vAx1> here is my problèm :D
[12:19] <vAx1> Ah, the question....
[12:19] <vAx1> Does someone has syslog-ng 3.3.x installed a server? Same problem ?
[12:36] <lynxman> dannf: ping
[12:36] <lynxman> er sorry
[12:36] <lynxman> Daviey: ping
[12:37] <lynxman> Ursinha: Bom dia soleiado
[12:39] <Daviey> lynxman:
[12:41] <pmatulis> vAx1: what can syslog-ng do that rsyslog cannot?
[12:47] <vAx1> pmatulis: have a compatible configuration
[12:47] <vAx1> :D
[12:47] <vAx1> here, we have about 200-300 servers ;)
[12:47] <vAx1> and some are named loghosts, relaylogs ...
[12:48] <vAx1> you dont want to see the configuration files
[12:48] <Exopaladin> Better than us, we scp/mail the logs around all over the place including some hacky reverse ssh tunnels to get around firewalls
[12:49] <vAx1> nice!
[12:49] <Exopaladin> It's truly horrible
[13:12] <Alan> maybe somebody can help me here - not sure if i'm reading these errors and/or the output of smartctl properly
[13:13] <Alan> firstly, grep "read error" /var/log/syslog gives me http://paste2.org/p/1809521 just before a disk drops out of my raid5
[13:14] <Alan> secondly, smartctl -a is telling me 3 pending sectors and 80 reallocated sectors
[13:14] <Alan> does this look like a drive which has run out of spare sectors and needs to be RMA'd? the other 3 drives show 0 for both statistics
[13:14] <Alan> and all 4 drives are the same make/model
[13:15] <Alan> i'm running a raid recheck again to see if it barfs on the same sectors again...
[13:16] <Alan> http://paste2.org/p/1809527 <-- the results from SMART self-test
[13:17] <_ruben> Alan: i'd suggest running the vendor's healthcheck tool on it, it'll tell if you should rma it or that there's else wrong with the system
[13:18] <Alan> _ruben: I swear samsung's healthcheck tool is rigged...
[13:18] <Alan> I have another 2 samsung drives of varying ages showing similar symptoms in other machines, and every damned time the samsung tool is like "eh, no problem here"
[13:18] <Alan> but it's the same section of the drive, again and again, that has the problem :|
[13:19] <Alan> and doesn't depend on SATA cable or SATA controller
[13:19] <slakcphil> good morning, has anyone had usb 3.0 stop working after this last slew of updates?(10.04 LTS)
[13:20] <slakcphil> not recognized with fdisk -l and dmesg gives errors about not adressing in port 1
[13:21] <slakcphil> 2TB external usb with rocketfish 3.0 pcix card
[13:22] <pmatulis> vAx1: 'compatible configuration'?
[13:24] <Alan> _ruben: given that the healthcheck tool hasn't said my disk is dead, what options do i have for further troubleshooting the problem?
[13:25] <vAx1> I can estimate an average of configuration lines number to 1000 per server
[13:25] <vAx1> and about 200 serveur in production
[13:26] <_ruben> Alan: the SMART errors might be enough to warrant a RMA, might wanna contact samsung about it. dunno what their rma policy is
[13:27] <_ruben> or your vendor, they might be more/less flexible
[13:27] <Alan> unfortunately i bought from 4 vendors at the same time (to source from different batches), so i wouldn't know which vendor :(
[13:28] <_ruben> they probably won't know either, so pick the vendor you have the most confidence in, or contact several/all ;)
[13:28] <Alan> haha :P
[13:28] <Alan> I did wonder about doing that :p
[13:29] <Alan> that being said, last time i tried to get a RMA from a vendor they said "eh, we'd just have to then RMA it to the manufacturer anyway, so you'd be waiting twice as long"
[13:29] <_ruben> atleast you wouldn't have to deal with the packaging hell, i know WD for instance has *very* strict packaging rules before even accepting a rma
[13:30] <_ruben> assuming the vendor is local
[13:30] <Alan> nope, internet vendor
[13:31]  * Alan waits for the "fail event" emails....
[13:34] <Alan> crap... S2H7JD1B216417
[13:34] <Alan> oops
[13:35] <Alan> http://www.samsunghddvalue.com/webrma/closed.html
[13:35] <Alan> the engrish, it burns
[13:45] <zul> sgood morning
[13:47] <_ruben> Alan: this might also mean that seagate's tools might work on samsung drives ;)
[13:48] <Alan> "Samsung warranty was handled within the UK and Seagate warranty you have to send your drive back to Holland"
[13:48] <Alan> yet more fail
[13:48] <Alan> _ruben: it also means RMA is completely in limbo at the moment...
[13:49] <Alan> _ruben: samsung's tools are atrociously difficult to use, given that no machine in existance now has a floppy drive...
[13:49] <zul> good morning
[13:50] <_ruben> Alan: that's what floppy-emulating-usb-sticks are for! ;)
[13:50] <Alan> yeah, never got that to work with the samsung tools either
[13:50] <_ruben> or even cdrom, if it wouldn't require write access ;)
[13:50] <ninjix> zul: morning
[13:50] <Alan> I curretly have a CD with the tools on that works, somehow
[13:50] <Alan> and only then after i turned off soemthing in the bios...
[13:51] <_ruben> nice
[13:52] <_ruben> i've been using western digitals for quite some time now, their tools is actualy semi-decent (on windows at least, not sure if they support any other OSs)
[14:00] <Alan> that's actually a big deal for me... at least the samsung tools will work without installing windows
[14:00] <Alan> since my server doesn't run windows, i'm not going to install windows just to test some drives
[14:09] <hallyn> Daviey: feel like chopping off etherboot's head today?
[14:09] <Daviey> hallyn: hell yes
[14:10] <zul> hmmm?
[14:10] <hallyn> great, you still have the package source handy?
[14:11] <_ruben> Alan: my servers don't run windows either, tho i do have some windows boxes available for that kinda stuff :)
[14:11] <Daviey> hallyn: no, but i'm looking :)
[14:12] <Daviey> hmm, it's not a branch
[14:12] <Alan> _ruben: yeah, it'd be a pain to move a drive between machines but i guess i could do it if i had to
[14:13] <Alan> thing is... isn't this what SMART is supposed to be fore?
[14:13] <Alan> *for
[14:13] <hallyn> Daviey: http://people.canonical.com/~serge/kill-etherboot/etherboot_5.4.5.dsc
[14:13] <_ruben> Alan: in theory: yes, but i never really managed to get smart to give me actually useful/reliable/etc info, tho that might just be me :)
[14:15] <_ruben> Alan: I'd just order a replacement drive anyway, wait for the limbo to pass, then ask samsung/seagate how to get eligable for rma .. when it comes back from rma, use it as cold spare :)
[14:15] <_ruben> (or replace the new disk with the rma, and use the "new" one as spare)
[14:16] <Alan> _ruben: order a replacement drive at about 2.5x the price it should be :|
[14:16] <_ruben> Alan: how so?
[14:16] <Alan> _ruben: because thailand?
[14:17] <Alan> don't know about the rest of the world, but hard drive prices are still inflated by about 150% here
[14:17] <Alan> i.e. 2TB drives are £150 instead of the £60 they were
[14:18] <_ruben> not following prices that closely, but didn't notice incredible increases either
[14:18] <koolhead11> smoser: around
[14:18] <_ruben> lets see what a 2tb drive goes for around here
[14:18] <patdk-wk> heh? the 2tb drive  Iwas going be getting, was $120, topped at $340, and now $280
[14:19] <_ruben> heh .. my "favorite" vendor actually has a note regarding thailand on the website .. prices and delivery times can't be guaranteed currently
[14:19] <patdk-wk> I need to turn on a 40 drive system before feb, so just waiting as long as I can
[14:20] <_ruben> samsung f4 2tb for 156 euros .. but not in stock...
[14:20] <Daviey> hallyn: I'm being a plum, but can you remind me why we are not doing this via ipxe?
[14:21] <patdk-wk> seems like half the laptop drvies are still normal priced
[14:22] <_ruben> patdk-wk: what kinda chassis for those drives btw?
[14:22] <koolhead11> hello Daviey :)
[14:22] <patdk-wk> probably a supermicro
[14:23] <Daviey> hey koolhead11
[14:23] <_ruben> patdk-wk: ah ok, been looking at that 88/72 drive chassis the other day .. looks kinda sweet :)
[14:23] <_ruben> (the 417 series)
[14:24] <patdk-wk> yep
[14:25] <_ruben> the extra costs for sff over lff is less than the costs of an extra chassis when going lff .. nice scaling option
[14:25] <hallyn> Daviey: say what?  we are doing it via ipxe?
[14:25] <soren> Daviey: The version of etherboot is way ahead of that of ipxe, so to do this from ipxe alone (and provide an upgrade path), ipxe would either need an epoch bump or the kvm-pxe binary would need to be built with another version than the rest of the binaries from the same source package.
[14:26] <patdk-wk> ruben, that depends on how much space you need
[14:26] <hallyn> oh i c
[14:26] <patdk-wk> the price of 600gig sff vs 2tb lff
[14:26] <hallyn> misunderstood q
[14:26] <soren> hallyn: Or maybe I did :)
[14:26] <soren> We'll see once Daviey responds. :)
[14:27] <_ruben> patdk-wk: we were comparing 1tb sff with 1tb lff
[14:28] <patdk-wk> laptop sff?
[14:28] <Daviey> soren: ahhh yes!
[14:29] <_ruben> patdk-wk: we no longer using 5.25" disks, so lff is 3.5" and sff is 2.5" in our books :) .. and sata, as it's for backups and other just-storage tasks
[14:31] <patdk-wk> I wouldn'
[14:31] <patdk-wk> t be using laptop 2.5" in a storage system
[14:31] <smoser> koolhead11, here.
[14:31] <patdk-wk> a real 2.5" drive, sure
[14:31] <koolhead11> smoser: i followed the guide https://help.ubuntu.com/community/UEC/Images#Ubuntu_Cloud_Guest_images_on_Local_Hypervisor_Natty_onward
[14:32] <koolhead11> and customized the image by adding the proxy information
[14:32] <koolhead11> now am not able to get instance in running state. :(
[14:32] <koolhead11> its stuck at "Pending"
[14:33] <koolhead11> even euca-get-console-output not helping me much :(
[14:33] <smoser> well, get-console-ouptut will only show after it starts running.
[14:34] <smoser> do you hvae access to the hosts?
[14:35] <koolhead11> smoser: http://paste.ubuntu.com/760481/  this is how i made it
[14:35] <koolhead11> am not able to ping the host
[14:35] <smoser> host
[14:35] <smoser> not guest
[14:35] <koolhead11> smoser: yes i have
[14:35] <ninjix> smoser: do you at least see the kvm process running?
[14:36] <koolhead11> ninjix: nopes
[14:37] <smoser> koolhead11, you will need to look in /var/log/libvirt
[14:37] <smoser> but your pastebin is missing some stuff.
[14:37] <strat-o-caster> I tried apt-get upgrade and it says bind9 is held back.  If I just do apt-get install bind9, will that blow away all my current live configurations? I am doing this from remote so I'm a bit nervous...
[14:37] <smoser> or maybe not. i didn't realiez you were doin gvnc to it.
[14:38] <smoser> so... thats fine, i' prefer you did everything in the user-data, as its then reproducible, but ok.
[14:39] <ninjix> koolhead11: I'm jumping in late here. You trying this in EC2 or private cloud?
[14:39] <koolhead11> ninjix: private cloud
[14:40] <koolhead11> smoser: there is no log created for the running instance :(
[14:41] <ninjix> no log makes me think that the host node is failing to find a resource
[14:41] <smoser> i dont think you should have to remove 70-persistent-net-.rules as kvm network devices should be ignored. if they're not i want to know.
[14:41] <smoser> https://bugs.launchpad.net/ubuntu/+source/udev/+bug/341006
[14:41] <smoser> koolhead11, look in the logs for that instance id
[14:41] <smoser> or if you are not getting that far, you're going to have to look at openstack.
[14:41] <smoser> logs
[14:42] <ninjix> koolhead11: this a euca or openstack based private cloud?
[14:42] <koolhead11> ninjix: openstack
[14:43] <koolhead11> smoser:  /var/lib/nova/instances/ even it does not show id for that instance
[14:43] <ninjix> but you see a "pending" right?
[14:43] <smoser> grep the logs for it.
[14:44] <ninjix> you see anything in /var/log/libvirt/qemu ?
[14:45] <koolhead11> ninjix: notthing even there :(0
[14:47] <hallyn> cmagina: ppetraki: any plans to look into the multipath-tools FTBFS?
[14:47] <koolhead11> smoser: ninjix  the only information i modified to the cloud image was assding
[14:48] <koolhead11> apt.cong file in /etc/apt
[14:48] <koolhead11> and providing it with proxy server infoemation
[14:48] <koolhead11> beacuse my whole infra is behind proxy
[14:48] <ninjix> this the first guest you've tried to bring up on the stack?
[14:50] <koolhead11> ninjix: i already have other instances running
[14:50] <koolhead11> my custom build images
[14:50] <koolhead11> and i can do ssh and stuff
[14:51] <ppetraki> hallyn, what's FTBFS?
[14:51] <smoser> fails to build from source
[14:52] <smoser> koolhead11, i'll try to reproduce here, but i have access only to a diablo level openstack to test it on.
[14:52] <koolhead11> smoser: yes that will be fine :)
[14:52] <ppetraki> hallyn, oh, that's news to me, sure.
[14:53] <koolhead11> smoser: my nova and running instance both are using same proxy address
[14:53] <hallyn> ppetraki: https://launchpad.net/ubuntu/+source/multipath-tools/0.4.9-2ubuntu1/+build/2965923
[14:55] <koolhead11> smoser: you want me to put it as questions at launchpad with cloud-init section?
[14:55] <koolhead11> with alld etails
[14:55] <koolhead11> *details
[14:58] <smoser> koolhead11, this is not cloud-init related.
[14:58] <smoser> the instance will get to "running" before cloud-init has the opportunity to screw up.
[14:58] <koolhead11> smoser: ok.
[14:58] <smoser> :)
[14:58] <koolhead11> ooh :P
[15:11] <strat-o-caster> I tried apt-get upgrade and it says bind9 is held back.  If I just do apt-get install bind9, will that blow away all my current live configurations?
[15:16] <smoser> koolhead11, did you find any evidence of that instance-id on a node ?
[15:16] <smoser> do you have multiple nodes ?
[15:24] <Daviey> RoAkSoAx: can you drive bug 898268?
[15:24] <RoAkSoAx> Daviey: sure, will put it in my TODO
[15:24] <Daviey> RoAkSoAx: fence-agents pulled it in btw
[15:25] <smoser> koolhead11, did you try the same image unmodified ?
[15:25] <zul> Daviey: do you want me to handle the keystone MIR
[15:25] <RoAkSoAx> Daviey: uhmmm... I guess debian maintainer added that since the latest fence-agents is pretty much a resync
[15:25] <smoser> ie, you're poinging to "/current". have you tried using something with /release ? just for a constant. (current link moves twice weekly)
[15:26] <RoAkSoAx> Daviey: ah never mind just remmebered that It hit main this cycle :)
[15:26] <Daviey> zul: sounds great
[15:27] <Daviey> zul: looks like an rdepends issue
[15:28] <Daviey> RoAkSoAx: nah, i think fence-agents was promoted; which has exposed it
[15:28] <RoAkSoAx> Daviey: yeah
[15:28] <koolhead11> smoser: let me do that again but AFAF i tried that as well
[15:29] <RoAkSoAx> Daviey: i though I was promoted last cycle, but I just remember that I filed the MIR to get it promoted this cycle
[15:30] <Daviey> ah cool
[15:41]  * RoAkSoAx just uploaded new redhat-cluster... gfs2-utils next
[15:51] <koolhead11> smoser: yes it was running well before i modified it, i could see the logs as well via euca-get-console command
[16:14] <alex88> hi guys, sometimes my server is really slow because of a process "fuser -s ./sess_io............" which uses all cpu, plus i see sometimes cannot allocate memory and too many open files error...what can it be?
[16:14] <zul> RoAkSoAx: hey did you see the email about "debian development" environment on the cobler ml
[16:16] <RoAkSoAx> zul: yeah! I was planning to reply to them to basically say that I build packages and then test instead of building from their branch
[16:16] <zul> RoAkSoAx: ok
[16:16] <RoAkSoAx> zul: any ideas? I was gonna point them to us in case they need further help
[16:17] <RoAkSoAx> zul: i think we should just request them to be co-maintainers
[16:17] <zul> RoAkSoAx: nah i was thinking of anything
[16:17] <zul> RoAkSoAx: sure
[16:18] <RoAkSoAx> zul: cool
[16:21] <zul> RoAkSoAx: yeah i just have gotten to it, will do so in a minute (stuck in SRU purgatory)
[16:21] <RoAkSoAx> heh ok
[16:22] <hallyn> stgraber: jjohansen: fyi, finally started https://wiki.ubuntu.com/LxcSecurity
[16:23] <stgraber> cool
[16:43] <kpettit> can linux using samba or something else be a active directory server?  Have a crappy Win2003 server I'd love to get rid of, but the computers on the network login via active directory.
[16:44] <zul> kpettit: yes it can
[16:44] <zul> check google there is alot of docs
[16:44] <SpamapS> kpettit: http://wiki.samba.org/index.php/Samba4
[16:44] <SpamapS> Still alpha unfortunately
[16:45] <kpettit> that's what I was wondering.  I know samba can be a relay but last I checked it culdn't be a PDC
[16:45] <kpettit> thanks.
[16:45] <SpamapS> http://www.samba.org/samba/news/developers/ms-patch.html
[16:45] <SpamapS> Nice
[16:45] <SpamapS> MS patching samba
[16:46] <zul> no no we arent a monoploy ;)
[16:47] <kpettit> that's pretty cool.
[16:48] <dkn> hmmm why does my server keep asking me for a password when i've disabled passwords for the account and i'm using RSA keys?
[16:48] <kpettit> does your RSA key use a password?
[16:48] <dkn> ah.... i don't think so
[16:49] <kpettit> You might want to check, if your RSA key has a password then you still need to enter that password when loggin in.
[16:49] <dkn> lol
[16:49] <kpettit> ssh-keygen asks for a password when you generate the keys.  You have to leave it blank if you want to do passwordless
[16:49] <dkn> ya.. i thought i did... i guess i'll just generate a new one and try again..
[16:51] <SpamapS> dkn: why not use a password?
[16:51] <dkn> SpamapS, backuppc needs to run automated
[16:53] <stgraber> hallyn: looks good, did you see any news for the shutdown/reboot patch?
[16:54] <hallyn> stgraber: yes, daniel has been pushing it over the weekend - 3 versions :)
[16:54] <hallyn> (oleg has been reviewing)
[16:54] <dkn> i've got a vm with a disk running backuppc as a server, and the main filesystem as the client, then filesystem runs it's own backuppc server for two removable disks
[16:55] <dkn> i can have backuppc login be both a client and server can't i?
[16:55] <SpamapS> dkn: passwordless keys are fine for automation.. but make sure you understand the risks. :)
[16:56] <stgraber> hallyn: ah, that's very good news
[16:56] <dkn> oh i know :)
[16:57] <hallyn> strgraber: I should probably do a manual test so I can chime in and keep the thread going...
[16:57] <stgraber> hallyn: hopefully we can get the kernel team to cherry pick once accepted (assuming it won't make it for 3.2)
[16:57] <hallyn> what do we bribe smb with?
[16:58] <smb> Beers work usually...
[16:59] <dkn> hmmmm it's still asking me for a password when i try testing by ssh in for the backuppc user
[16:59] <smb> But the earlier the better for special consideration
[17:02] <hallyn> smb: great, i hear it's cheap in budapest  :)
[17:03] <smb> hallyn, Even more compared to US prices
[17:03] <adam_g> RoAkSoAx: ping
[17:04] <RoAkSoAx> adam_g: pong
[17:06] <RoAkSoAx> adam_g: what's up
[17:07] <adam_g> RoAkSoAx: would you mind targeting precise for bug #893352 and marking fixed released, if thats the case? looks like it is, but wasn't sure with the merge into resource-agents. just checked again and it looks okay in cluster-agents 1:1.0.4-0ubuntu2
[17:08] <RoAkSoAx> adam_g: ok, i'll take a look
[17:09] <adam_g> RoAkSoAx: sweet, thanks
[17:11] <hallyn> Daviey: whoa!  i thought the kvm-ipxe package got pushed?  but i don't see that version of ipxe in the publishing history
[17:11] <hallyn> can't do the etherboot push until kvm-ipxe is in archive
[17:22] <Daviey> hallyn: thanks for the heads up
[17:23] <hallyn> Daviey: but so i'm confused, i thought you uploaded htat on friday?
[17:23] <Daviey> hallyn: on a call right now, will confirm after
[17:23] <hallyn> kthx
[17:25] <x404x> I changed workgroup in the /etc/smb.conf file but still it uses wrong group after restart , what to do ?
[17:26] <xranby> x404x: can you post the config some place?
[17:27] <xranby> preferably using paste.ubuntu.com
[17:30] <zul> adam_g: fyi quantum and horizon has been packaged as well
[17:30] <hallyn> ahs3: i rolled the libnl fix as well into http://people.canonical.com/~serge/netcf-0.1.9-2.debdiff
[17:30] <hallyn> builds for me on sid
[17:33] <smoser> well, koolhead11 not here, but http://paste.ubuntu.com/760635/ "worked for me"
[17:36] <Guest35318> how to make nano select all text ?
[17:43] <blkperl> Guest35318: i don't think you can..
[17:44] <blkperl> Guest35318: http://superuser.com/questions/196425/how-do-i-select-all-text-from-a-file-with-nano
[17:54] <Guest35318> http://paste.ubuntu.com/760674/
[17:55] <Ursinha> Daviey: how busy are you right now?
[17:55] <Daviey> Ursinha: in a call
[18:03] <F-3000> Hello! What term is used, when a single data (ie a file) is written simultaneously on two harddrives?
[18:03] <kpettit> mirroring?
[18:04] <F-3000> kpettit: Ay, thanks.
[18:05] <cloudgeek> how to setup a python or django hosting on ubuntu
[18:05] <kpettit> Raid 1 might be what your looking for as well.  Raid 1 is hard drive mirroring
[18:06] <kpettit> cloudgeek, I like to use pip.  pip is like a cli python package manager.   "pip install django"
[18:06] <x404x> anyone look at my smb.conf file ? why does it not change workgroup ? hhttp://paste.ubuntu.com/760674/
[18:06] <cloudgeek> kpeyyit:i know that :)  but i looking it config with django , for student project ,that from one server system we can hosting for our projects
[18:07] <kpettit> cloudgeek, sorry didn't understand that.  Can you rephrase?
[18:08] <kpettit> i use django on cloud servers alot.  Most give me ubuntu to start with, then from there I install pip then django and other python stuff I need.
[18:08] <tash> need some samba help, anyone in here worked with it enough to speak on it knowledgeably?
[18:08] <ahs3> hallyn: thx.  i'll take a look here in a bit.
[18:09] <kpettit> tash, go ahead and ask.  If somebody can help they will.
[18:09] <tash> kpettit: k
[18:09] <cloudgeek> kpettit:hey we need a hosting for our student project we can't efford the price , so looking setup out own for our projects o we can use for out projects
[18:10] <tash> Have several windows users using a program called autocad to make drawings, they then save them to the samba share so others in the office can view them and edit them.  However, the files seem to be getting added with 744 permissions for some reason.
[18:10] <tash> I'd like them to be defaulted to 775
[18:10] <kpettit> cloudgeek, so you want to create a service on your own hardware so you can host student django projects?
[18:11] <tash> I thought create mask = 0775 in smb.conf would be my friend, but not so sure that is the case b/c create mask 0755 was already set, and like I said, the files are defaulting to 744
[18:11] <dkn> /var/lib/backuppc is the home directly and the server directory for the backuppc user on the filesystem.... aha got it now..
[18:11] <cloudgeek> kpettit:yep ,for our use only ! any guide or tutorial
[18:12] <patdk-wk> tash, create mask is a restriction
[18:12] <patdk-wk> you need force create mode, to ADD missing permissions
[18:13] <kpettit> cloudgeek, I can't think of a app that does that off hand.  I think it depends a bit on how you want the students to be able to add new projects.  I'd probally give the students ubuntu logins and have apache look in there users directory for the .wsgi file.
[18:13] <tash> patdk-wk: so, would force create mode = 0775 do what I am asking then?  And, also for directories, force directory mode = 0775 ?
[18:14] <patdk-wk> probably
[18:14] <kpettit> tash.  If your not that familiar with samba I'd use SWAT which is samba's web admin tool.  It can show you ever option and has a nice help next to each option so you can see what it does.
[18:15] <cloudgeek> kpettit:thanks , i try to implement , if fails i ask gaian for help
[18:16] <tash> kpettit: thanks for the tips
[18:16] <kpettit> cloudgeek,  ok.  Also remember if students don't need the website to be permanent they can use django's built in webserver.
[18:17] <kpettit> tash, no problem.
[18:17] <cloudgeek> kpettit they need as staic for 4 month till the project
[18:18] <kpettit> cloudgeek, then I'd go the approach of hardcoding apache to look for wsgi file in students home directory.  That way the students have there own private dev area.
[18:19] <cloudgeek> kpettit:okay any step by step tutorial for that
[18:19] <kpettit> cloudgeek, do you know django or how to get python apps running in apache?
[18:21] <cloudgeek> kpettit: i try it with mod wsgi
[18:21] <kpettit> If your new to it.  You can look at this:  https://docs.djangoproject.com/en/dev/howto/deployment/modwsgi/   In ubuntu you need to make sure wsgi is installed
[18:21] <cloudgeek> most of our work on ubuntu desktop now we switch server
[18:22] <kpettit> Ah got ya.  To install wsgi in ubuntu "apt-get install libapache2-mod-wsgi"  then from there you can follow instructions on that link I sent
[18:22] <kpettit> there is also a django channel on this server that has lots of good people if you get more specific questions on it.
[18:22] <cloudgeek> thanks all for help need to sleep exam on tomorrow
[18:23] <kpettit> no problem.  Have fun
[18:23] <cloudgeek> kpettit::)
[18:54] <zul> SpamapS: im probably going to upload the SRU on thursday
[18:57] <lynxman> Daviey: ping
[19:11] <ipl31> Hello, I am noticing when I stop and start LXCs on 11.10 server that is seems to pause the networking on Host OS
[19:17] <SpamapS> zul: please make *one* bug for tracking the SRU, instead of 40 ;)
[19:18] <zul> SpamapS: you sure? :)
[19:18] <zul> SpamapS: include the changelog, debdiff, etc, etc?
[19:19] <SpamapS> zul: you can leave the bugs in the changelog
[19:19] <SpamapS> zul: just open a bug task for the overall process
[19:19] <zul> SpamapS: thanks
[19:59] <adam_g> SpamapS: is precise a valid release for lp:charm/$rel/$pkg branches?
[20:01] <SpamapS> adam_g: since you asked, I'll create the precise series... tho your push will probably be alone in it. ;)
[20:03] <SpamapS> adam_g: should work now. :-D
[20:04] <adam_g> SpamapS: ah, it wont be alone for long, im sure. thanks
[20:04] <SpamapS> adam_g: I hope to have a simple way to copy oneiric -> precise when the time comes.
[20:30] <smoser> jamespage, i just pushed to https://code.launchpad.net/~ubuntu-server-ec2-testing-dev/+junk/ec2-automated-tests/
[20:59] <jamespage> smoser: looks good - thanks for letting me know!
[21:00] <dob_> Hi, i have some special question. How can i add a [OK] at the end of every line i echo?
[21:01] <smoser> myecho() { echo "$@" "[OK]"; }
[21:01] <smoser> myecho "hi mom"
[21:02] <smoser> { echo hi mom; echo bye mom; } | sed 's,$, [OK]'
[21:02] <smoser> { echo hi mom; echo bye mom; } | sed 's,$, [OK],'
[21:02] <smoser> echo -n "hi mom"
[21:02] <smoser> echo "[OK]"
[21:02] <smoser> echo -n "bye mom"
[21:02] <smoser> echo "[OK]"
[21:04] <dob_> smoser: is it also possible to right append the [OK]. So that it's like at starting up linux?
[21:07] <smoser> i believe if you want to do that you should look in /lib/lsb/init-functions
[21:08] <smoser> at log_daemon_msg, log_progress_msg, log_end_msg
[21:08] <dob_> smoser: thank u. I will have a look
[22:15] <hggdh> hallyn: kvm is failing to start all my VMs with 'permission denied' on the disk image. No apparmour errors reported
[22:15] <hggdh> hallyn: any suggestion on how to debug?
[22:17] <hallyn> hggdh: afk atm, but check /var/lib/libvirt/ and try kvm by hand
[22:18] <hallyn> sorry meant log not lib :)
[22:20] <hggdh> k
[22:23] <mtaylor> hey guys!
[22:23] <mtaylor> 14:22 <vishy> mtaylor: in openvz channel they are commenting that they think ubuntu doesn't support it anymore
[22:23] <mtaylor> anybody know anything about that? and/or what i'd need to do to support openvz on an ubuntu server?
[22:28] <zul> mtaylor: patches accepted ;)
[22:34] <SpamapS> mtaylor: openVZ is a massive host kernel patch that AFAIK has been rejected upstream
[22:34] <mtaylor> SpamapS: awesome. I learn new things every day
[22:35] <SpamapS> mtaylor: its sort of the reason we're backing LXC.. lighter weight patch that is upstream for good.
[22:35] <SpamapS> history repeats itself w/ KVM vs. Xen all over again. ;)
[22:49] <hallyn> yeah incidentally, zul - i'm implicitly leaving libvirt-xen bugs for you.  poke me if that's wrong.
[22:53] <xiris> Hey guys, please ignore my english but ... someone can help-me to execute one .sh when i create a user?
[22:56] <RoAkSoAx> zul: are we still importing daily checkouts from cobbler?
[23:16] <zul> RoAkSoAx: no
[23:17] <zul> hallyn: oh