/srv/irclogs.ubuntu.com/2012/05/08/#ubuntu-server.txt

aljosai'm getting "/usr/lib/lxc/templates/lxc-ubuntu-cloud: line 240: type: ubuntu-cloudimg-query: not found" when using "lxc-create -t ubuntu-cloud -n box2"00:59
aljosaany ideas?00:59
=== bazhang_ is now known as bazhang
resnoim considering splitting my server into mysql db and web server... which machine needs to be beefer?01:06
resnoor would it be better to just get a better server and be done with it?01:06
aljosaresno: it's usually database that needs a good hw but depends on your case. do you actually have perfromance issues?01:07
smoserDeathvalley122, your issue is that you have wildcard domain for that01:10
smoseri think.01:10
resnoaljosa: nah... just setting it up so, im curious if im just being silly or or not01:10
smoserbasically, cloud-init tried looking up 'ubuntu-mirror' in the domain was found in metadata server's local-hostname.01:10
resnoaljosa: im moving from a shared host to vps... so im unsure what my needs really are, but these are low inboxes01:10
smoserand got a hit.01:11
Deathvalley122smoser: was that a tab error?01:16
=== Stefan_ is now known as juicy
=== Stefan_ is now known as juicy
=== Jason__ is now known as juicy
mariooohi all. having consistent network issues with an upgraded 11.04 > 11.10 VM.02:35
mariooohoping for some help as to how to debug/diagnose the issue, lots of googling but not really sure where to go from here.02:35
mariooosome system info + things i've tried: http://pastie.org/387660502:36
twbDefine "issues"02:36
mariooobasically all networking seems to work fine until any significant traffic occurs, most notably apt-get update/upgrade. this seems02:36
marioooto crash the network adapter (if i'm SSHed my connection gets dropped) and can't do anything till networking is restarted02:36
Bronzetwb++ :)02:36
marioooI was intending to upgrade to 12.04LTS but wasn't aware of the required -d flag on do-release-upgrade, so now am stuck (do-release-upgrade fails as network crashes randomly while trying to get packages)02:36
qman__mariooo, release upgrades must be done incrementally; you'd have to go 11.04 -> 11.10 -> 12.04 regardless02:37
qman__the exception is that LTS can upgrade directly to LTS02:37
marioooqman__: damn. was fearing that might be the case.02:37
qman__8.04 -> 10.04 -> 12.0402:37
mariooosure. well in that case guess I really can't avoid this issue, all our VMs are on 11.04 =(02:38
twbmariooo: is the VM running a GUI?02:38
marioootwb: no, headless server VMs02:38
twbmariooo: is network manager (the backend daemon) installed?02:38
marioootwb: I'm unsure. how would I check?02:39
twbmariooo: dpkg -l | grep network, I guess.  I forget if there's a hyphen in the package name02:39
qman__there is02:39
qman__network-manager, network-manager-gnome, etc02:39
twbqman__: package is network-manager, process is NetworkManager, right?02:40
qman__yes02:40
qman__couldn't tell you why, makes about as much sense as the rest of it02:40
twbmariooo: pastebin output of "cat /etc/network/interfaces; ip a; ip r"02:41
twbmariooo: and tell us if NM is installed02:41
qman__and now 12.04 has mandatory resolvconf, even more headaches02:41
twbqman__: apparently if you rm the symlink it goes back to traditional mode02:41
ScottKNot on servers.02:41
twbScottK: ubuntu-standard isn't installed on servers?02:41
qman__it is on all the servers I installed02:42
ScottKIt is.02:42
mariooopastie updated. network manager doesn't appear to be there02:42
* twb digs out a precise box to check the exact deps02:42
qman__when I first experienced the issue I used chattr +i to stop it, but the next update broke it02:42
ScottKNevermind.  I was wrong.02:42
mariooohad an idle connection crash on me for the first time, so I guess I retract what I said before about high traffic. high traffic just seems to make it happen sooner.02:42
twbqman__: dpkg-divert --rename02:42
qman__granted that was a weak hack with no real research put into it02:42
twbAnyway wtf how could chattr +i be subverted02:43
qman__I don't know!02:43
virusuymariooo: that issue only happens in your VM ??02:43
virusuynot in other VM .. or somewhere else ?02:43
twbmariooo: I cannot see anything wrong there.  When it crashes, what still works?  e.g. does ping still work?  What VM technology (KVM, ESX, ...) are you using?02:43
mariooovirusuy: not sure what you mean... I've only got 1 VM that i've upgraded atm. I had a previous VM that had the exact same issues, but I trashed it in hopes of upgrading straight to 12.04, but have ended up on 11.10 again with exact same issues02:44
virusuyand you're using ..... KVM? XEN? ESXI ?02:44
marioootwb: I'm unsure what our underlying system is, I know what our hardware is but the guys at the colo are managing the VM pltform02:45
twbeek02:45
virusuyso.. let me see if i understand the issue...02:45
marioooI could probably ask & find out, but it might take a while. Any way I can do it from the system?02:45
twb"Unassigned class [ff80]: XenSource, Inc. Xen Platform Device (rev 01)" sounds like xen02:45
mariooooh yeah deffo Xen02:45
virusuyeek02:45
mariooosorry, it wasn't in the list so thought you were talking abotu something lower level02:45
twbxen doesn't do full virtualization, it does some magic hacks in the kernel02:46
twbSo you probably aren't running a stock 11.10 kernel02:46
marioooright, so it's an issue the colo guys will need to help resolve?02:46
twbOh wait, I take it back -- current kernels can be domU without any extra changes02:46
mariooohrm.02:46
twbmariooo: it is definitely worth talking to them about it, since they have access to the dom0 and can check things we can't02:46
twbBut it's a stock ubuntu kernel, going by the uname -a02:47
marioooyeah well I've mentioned it before but they've not been very helpful about it, so I have a feeling they have no idea02:47
virusuyif you suffered this issue before in other VM ... looks like a issue with that eth02:47
qman__you might try booting the old kernel after the upgrade02:48
twbmariooo: get a better colo provider ;-P02:48
virusuytwb: :)02:48
qman__or a newer one02:49
marioootwb: haha, they have been pretty awesome for everything aside from this. we're in australia so our choices seem pretty limited =(02:52
marioooqman__: sorry to noob-out on this, but any pointers for how I'd go about this? something to do with grub? =D02:53
qman__yep02:53
twbmariooo: I'm in .au too02:53
twbmariooo: most people I know don't bother hosting within .au because it's so fucking stupidly expensive02:53
marioootwb: financial clients. legal implications. yada yada.02:54
qman__connect to the 'local' console, use shift during boot, select old kernel02:54
twbmariooo: sorry to hear that :-(02:54
twbqman__: can he do that without dom0 access?02:54
marioootwb: thing ping is glorious though, our colo is like 2 streets away from work. nobody works locally02:54
qman__if you can't do that, you can modify grub config, but I don't know how to do that specifically in grub202:54
twbmariooo: you in sydney?02:54
marioootwb: indeedy02:54
twbWho is the provider?02:54
marioooqman__: hrm ok, I'll go poking around02:55
marioootwb: spectrum networks02:55
marioootwb / qman__: thanks for the pointers btw, really appreciate it!03:06
TohuwHow can I add a second image to my Ubuntu PXE server? I have the first one working, but am unclear on how to add the second and how to pick the image. I setup PXE per https://help.ubuntu.com/community/PXEInstallServer03:07
TenLettersI had a question if someone wouldn't mind answering...03:11
virusuy!ask03:11
ubottuPlease don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience03:11
virusuy!ask TenLetters03:11
TenLettersthanks03:11
TenLettersI was just trying to get somebodys attention03:12
TenLettersI found an old computer in my parents storage unit, Celeron 2.7 Ghz, 512 mb, 40 gb hd. I was wondering if you would think I would have significant problems running a server on it. I don't plan on anything major, I just wanted to learn to connect to and operate a server, and learn linux.03:12
TohuwTenLetters: I have successfully run Ubuntu 9.04 on a Pentium MMX. You will be fine, provided you don't overburden it. I'd consider picking up >= 512MB more RAM, but overall you should be fine as a starter. Some server applications/platforms will considerably slow performance, though.03:14
TohuwTenLetters: A base install of ubuntu server with OpenSSH server will run "fine" though.03:15
TenLettersTohuw: Thank you so much! Will have to see what I can find in the way of ram, could probably bump it to 1 GB..03:16
TohuwTenLetters: not a must, but it will certainly help if you start tooling around with Apache, SQL, etc.03:17
ResistanceTenLetters:  a Celeron is kind of old... you might not be able to get 1GB of RAM in something that old03:17
Resistanceyou'd have to check the specs for that board03:18
Resistancefor that system (and that may be hard to find)03:18
TenLettersResistance: Good point, will have to look and see. It's a DDR400 board03:18
Resistanceoh god, DDR?03:18
TenLettersyeah >.>03:18
Resistancei highly doubt you'll be able to even *FIND* RAM for that03:18
TenLettersI know a few places03:18
Resistanceat least not within a normally-affordable range03:18
TenLettersI used to work at a computer store, I know they had a lot laying around.03:19
Resistancei'd still go search the specs up for that board, it sohuld have information about the maximum amount of RAM those systems can have03:19
Resistanceif i remember when i worked on boards with Celeron in those days03:19
TenLettersah03:19
Resistancethey had 2 or 4 slots for memory, and could handle 128MB per slot03:19
Resistancewhich normally maxes out at 51203:19
TenLettersWell03:20
Resistancewhat's the make/model of the system03:20
TenLettersI have 2 slots with 256 in them03:20
Resistancei *might* be able to find it03:20
TenLettersI have to open it up03:20
TenLettersI replaced the mobo03:20
TenLettersgive me a sec03:20
ResistanceDesktop / Server?  or Laptop?03:20
* Resistance assumes Desktop / Server for something that old03:20
TenLettersDesktop03:20
TenLettersI could find a better use for a laptop03:20
TohuwResistance: in my experience, motherboards supporting 2+ GHz Celerons rarely failed to support at least 1 GB maximum.03:21
Resistanceah, if you replaced the mobo, give me those specs03:21
ResistanceTohuw:  depends on the board, some could, others couldnt03:21
qman__TenLetters, it depends entirely on what you want to run -- but as another example, my shell server is a K6 200MHz with 256MB EDO and a 10GB hard drive03:21
Resistancei.e. mobo manufacturer and model (in case, i'll check what the max memory is in that, if possible)03:21
TenLetterswell damn qman__03:21
qman__if it ain't broke03:21
TenLettersResistance: Prescott 800 model P4i65G03:22
TenLettersManufacturer: Asrock03:22
TenLettersmy old POS when I was a little kid...03:22
TenLettersrunning it makes it feel like opening a time capsule.03:22
Resistancethis one?  http://www.asrock.com/mb/overview.asp?model=p4i65g03:23
Tohuwqman__: K6's are indestructible processors. It's scientifically proven. I still have a working K6/128 MB RAM box sitting around, running some antiquated Slackware. I think it will outlast all my other systems... combined.03:23
ResistanceTenLetters:  see my last message03:24
qman__that's why I use it03:24
qman__it's never going to stop03:24
qman__at least not before everything else I have03:24
TenLettersResistance: That's it03:24
ResistanceTenLetters:  mem specs of relevance: 2 x DDR DIMM slots, dual channel tech.  Supports DDR400/333/266 non-ECC, un-buffered memory.  Max. capacity of system memory: 2GB03:25
TenLettersTohuw, qman__: I wish things were like the old days, all of these old products seem to last much longer than the newer ones. SNES are still going while XBox's and PS3 are dropping out of the sky03:25
TenLettersResistance: Thanks, Supports 2 GB, which will cost an arm and a leg for DDR :/03:26
qman__plus the fact that it can literally only handle about 20 SSH login attempts a minute turns out to be a feature03:26
ResistanceTenLetters:  indeed.03:26
TenLetterswow...03:26
twbqman__: you can simply tell sshd to do that03:26
ResistanceTenLetters:  you *might* find some on ebay, or that shop you mentioned, but for all intents and purposes, i dont think you'll affordably find any03:26
TenLettershere's some on newegg03:27
twbqman__: MaxStartups 8:16:3203:27
TenLetters$25 :/03:27
marioootwb: oh sorry, when you were asking if network-manager was installed, was it a good or bad thing for it to not be?03:27
qman__I use iptables to cull it further and stop the disk activity I was getting during brute force attempts before03:27
TenLetterswhich isn't terrible, but I don't think it's worth buying for an experiment Resistance.03:27
twbmariooo: NM causes problems, it should not be installed03:28
twbmariooo: at least not on a server03:28
marioooright. ok03:28
qman__it's convenient on a laptop with wifi03:28
Tohuw<qman__> plus the fact that it can literally only handle about 20 SSH login attempts a minute turns out to be a feature <-- okay, that made me actually laugh out loud03:28
qman__that's about the only time it should be used, ever03:28
ResistanceTenLetters:  its not *bad*, but its not experiment-like costs :P03:30
Tohuwqman__: any system with mobile broadband is also easier with nm, and any system that has to readily support multiple wired settings (for whatever reason) is also easier.03:30
qman__for multiple wired networks I usually just configure them all simultaneously as subinterfaces03:31
qman__I can see how that would be useful though03:31
DestreyfI was wondering if anyone had a recommendation for setting up openstack up and running with only having 4 machines (1 machine as MAAS/juju). https://help.ubuntu.com/community/UbuntuCloudInfrastructure <- says 10 machines, though i don't have that many on hand at the moment.03:38
qman__Destreyf, I don't know why exactly that is, but it does use 'minimum' twice in once sentence, so I assume that's important03:40
qman__then again, it appears to contradict itself, implying you need different machines for the juju and MAAS, then saying you can install juju on the MAAS a couple lines later03:42
Destreyfwell i know you can use lxc to create virtual containers and deploy them, but i have near zero experiance getting there, i was trying to enter ubuntu-cloud channel, but it requires an invite apparently lol03:45
qman__I really don't know much about it, never used it03:46
Destreyfdang.03:46
qman__I run all my stuff bare metal, mostly on old junk, just don't have those kinds of requirements03:47
Destreyflol my work has a bunch of powerful servers, that sit around 90% of the time idle, and we're trying to migrate into a cloud infastructure/virtualization so we can use those environments so we can utilizes only what's needed on the smaller machines.03:47
qman__good in theory, though I'd be interested to see at what point it actually pays off03:49
qman__in my case I'd have to buy a bunch of new hardware to do it, so a few extra kilowatt hours are cheaper for me03:49
TohuwI'm baffled at why the example deployment they provide requires a "minimum of 10 machines" -- OpenStack itself doesn't really care. Juju might somewhere in there, though I don't know why...03:50
qman__a few grand can power the old K6 for a looong time03:50
Tohuwqman__: I usually argue virtualization/private cloud infrastructures not based on power savings, but balanced resource consumption. Being able to spike key services across more underlying hardware is nice.03:52
twbTohuw: the main benefit of VMs is not having to deal with stupid hardware issues anymoer03:53
twbe.g. if a HDD dies, that is someone else's problem, not yours (the VM maintainer's)03:53
twbYour OS suddenly only has to care about a little old ne2000 and an ISA bus ;-)03:54
Tohuwtwb: to an extent, yes. But I'm usually proposing virtualization to the parties for whom the dead HDD is a problem.03:54
qman__don't get me wrong, it's great tech, just wondering at what point it's worth doing, just how many servers you have to have to see the value03:55
twbqman__: IMO it's not worth the extra infrastructure hassle for less than, oh, five, ten or twenty boxes03:55
twbIf you are at the end of a hw/os cycle ANYWAY, it might make sense for the next build to consolidate several boxes onto a single beefy one using VM tech to partition that hw03:56
qman__I've only got six, and a few of them wouldn't virtualize well anyway03:56
Tohuwqman__: That answer has been downshifting as the solutions get cheaper. I have clients of < 25 employees/~5 servers who see significant returns on their virtualized infrastructure, and that's with VMWare, so licensing is in the mix.03:56
qman__one's the router, one's the file server, and one's a game server (needs the performance)03:56
twbTohuw: but you probably aren't including the cost of upskilling someone (i.e. you) to know how to deal with VMs, because you already know03:57
TohuwI theorize qman__ is the last professional Quake 2 Deathmatch host in the world.03:57
twbThat extra training can be a non-negligible cost03:57
twbqman__: yeah I wouldn't bother03:57
qman__counter strike was always more my thing03:58
twbqman__: esp. since the first two can run on your old laptop03:58
Tohuwtwb: Certainly a fair point: the clients in question use my employer for outsourced IT resources, so we're there in some capacity to help.03:58
qman__but this one actually runs minecraft03:59
twbI don't "get" minecraft03:59
Tohuwtwb: Though, to play devil's advocate, VMWare offers support as part of the licensure (yes, the support will make your eyes bleed sometimes, but it is there)03:59
TohuwYou and me both, twb03:59
twbThat's about as useful as normal upstream support03:59
qman__it's fun for a while, I actually haven't played it in quite some time, I've got a bunch of friends who do03:59
twbIs it a C4 game or what04:00
qman__I'm the only one with the experience and infrastructure to actually run a 24/7 server for it, though, so I do04:00
TenLetterstwb: "C4"?04:01
twbAnd why does a c4 game include a 16-bit CPU building block04:01
twbTenLetters: like civilization04:01
TenLetterstwb: Ah, no. Think interactive legos04:01
qman__not really04:01
qman__it's more of a sandbox04:01
twbqman__: so like GTA?04:01
TenLetterstwb: every played with legos?04:01
TenLetters:o04:01
qman__yes, except with less gangsters and more garry's mod04:01
qman__in cubes04:01
twbNever heard of garry's mode04:02
TenLettersPretty much do whatever you feel like twb. Build crap, blow it up...04:02
twbTenLetters: yes, but never with any that had electronics04:02
TenLetterstwb: You poor poor person...04:02
qman__it's part generic survival and part collaborative construction tool04:03
qman__in an infinitely large world of cubes04:03
TenLettersqman__: actually, there are problems if you go too far out04:03
qman__and that last part is why it needs a decent server04:03
qman__combined with the fact that it's written in java04:04
TenLetterspoorly written java04:04
TenLettershas hella memory leaks from what i've seen.04:04
twball java is poorly written java04:05
TenLettersgood point...04:06
twbThe only time I've seen anything on the JVM that wasn't poorly done, was because it was compiled down from something else04:06
Destreyflol i'm off reading, and you guys had a massive conversation04:26
DestreyfWe have the hardware infastruture to do a cloud, i just dont have all of it available now.04:27
Destreyfusing LXC to substitute the other "5-6" boxes i need.04:28
Destreyfuntil i can move stuff onto the virtualized enviroments.04:28
DestreyfSo anyone in here have any experiance with Juju?04:33
=== dduffey_afk is now known as dduffey
=== bicranial_ is now known as bicranial
=== Madkiss_ is now known as Madkiss
=== bicranial is now known as 16WAAJA84
Deathvalley122pdtpatrick: fixed it turns out that the install for ubuntu installed resolvconf package05:37
Deathvalley122its resolving fine now05:37
=== Lcawte|Away is now known as Lcawte
eagles0513875_hey guys i think there is an issue with the 12.04 server documentation when it comes to dovecot :( the configuration file is different compared to whats mentioned on the 12.04 server guide for it06:42
=== jtv1 is now known as jtv
SpamapSeagles0513875_: https://bugs.launchpad.net/ubuntu/+source/ubuntu-docs/+filebug07:19
eagles0513875_ty SpamapS07:20
=== jtv1 is now known as jtv
KM0201is there a command to list all installed packages?07:41
twbdpkg --get-selections07:42
KM0201thanks07:44
=== almaisan-away is now known as al-maisan
=== RoyK^ is now known as RoyK
=== cedr is now known as tcl
=== tcl is now known as cedr
=== al-maisan is now known as almaisan-away
=== Neal|ZNC is now known as Neal_
aboanybody here able to help with juju?12:17
Farithgood day peeps. Looking for some assistance on troubleshooting wakeonlan issues. Anyone?12:48
Farithwake from pm-hibernate seems to have disabled my eth012:50
henkjanethtool -s eth0 wol g12:53
henkjandoes that help?12:53
Farithif done heaps of stuff...including that...and it aint working....12:55
Farithwell...WOL works....12:55
Fariththe server wakes fine...but eth0 is disabled.12:56
Farithrunning sudo ethtool eth0 gives me a link down status12:56
Farithim on 12.04 lts btw12:57
Fariththis only happens after a reboot followed by a pm-hibernate13:01
Fariththe next pm-hibernate ran in the same session wakes the server perfectly13:01
Barbo91Hi guys! Need help with the new ubuntu 12, i need to configure dns on my ubuntu server. resolv.conf file change everytime i reboot and i don't know how to configure the resolvconf package. Someone can help me?13:19
tsimpsonBarbo91: have a look at http://www.stgraber.org/2012/02/24/dns-in-ubuntu-12-04/13:20
Barbo91tsimpson, thank you, i'm going to read it and i will come back if i cant fix it :)13:24
tsimpsonyou basically do the same edits, but in one of the files in /etc/resolvconf/resolv.conf.d/13:25
Barbo91yep just fixed it :) Thanks for help! Nice work with 12.04 i'm loving it! :D13:29
=== Ng_ is now known as Ng
=== matsubara-afk is now known as matsubara
=== soren_ is now known as soren
jsmith-argotecwhat is the best way to change the default i/o schedular back to deadline persistently in 12.04?14:22
jsmith-argotecs/schedular/scheduler14:22
jsmith-argotecnevermind found it - looks like adding to /etc/default/grub is probably best14:32
mgwis there a way to set the ldap rootpw from an already encrypted value?14:51
jsmith-argotecmgw: ldapmodify with replace: and use the old encrypted value14:56
jsmith-argotecmgw: what type of encryption?14:57
jsmith-argotecmgw: err should I say setup with whatever for a password then after replace with the correct value.  That is if I'm understanding your question properly14:57
philsfcan anyone please help me debug an annoying issue? I have a single lucid box that I can't get ssh public key authentication to work, and it refuses silently, asking for a password instead. I configured .ssh/authorized_keys correctly, and synced all ~/.ssh/* from another lucid box that accepts pubkey auth with the same user config files, and the same configs in /etc/ssh/sshd_config. Where else should I look? Here is the debug information from a15:05
philsfconnection attempt: http://paste.ubuntu.com/975923/15:05
patdk-wkdid you check permissions?15:09
patdk-wkand file ownership15:09
philsfpatdk-wk, yep, they are all the same. I synced with rsync15:09
patdk-wkall the same sounds like *broken*15:09
philsfhttp://paste.ubuntu.com/975949/15:16
philsfit's the same files, and permissions. both for files and the ~/.ssh dir itself15:16
philsfI use these same keys in several machines, but this is the only one that ignores them. And doesn't even give me an informative error15:17
philsfoh, wait. I see now the error in sshd log. bad permissions on home dir. :/15:19
philsfthanks anyway15:19
=== railsraider_ is now known as railsraider
eagles0513875hey guys i have nfs setup and exporting just fine on my local machine. My problem now is actually allowing the user to have write access to /mnt/iso to put some iso images in that directory. can someone help me and point me to what i am missing15:32
patdk-wkeagles0513875, nfs 3 or 4?15:37
eagles0513875patdk-lap_: this is the way i have the export setup /mnt/iso *(rw,sync,no_root_squash)15:39
eagles0513875i dunno if thats nfs2 3 or 415:39
eagles0513875patdk-lap_: i followed what was in the 12.04 server guide online15:39
jsmith-argotectrying to start up samba after upgrading to 12.04 and getting a failure on nmbd - not finding much in the logs15:42
jsmith-argotecroot@xx:~# service nmbd start15:42
jsmith-argotecstart: Job failed to start15:43
jsmith-argotecinit: nmbd pre-start process (19425) terminated with status 115:43
halvors1I'm trying to configure DNS Client on my server.15:49
halvors1But resolv.conf keep overwriting my config.15:50
halvors1What do i do?15:50
iunruhI have a reprepro-backed repository and I'm trying to add a 3rd party (signed by them) package and reprepro complains it can't find the key for said package15:50
eagles0513875jsmith-argotec:15:53
jsmith-argoteceagles0513875: yes15:53
eagles0513875oops sry bout that15:53
eagles0513875track pad on my netbook is very sensitive with no way to turn it off15:54
jsmith-argotec:-) you got me excited that you might have had an answer15:54
jsmith-argotechehe15:54
halvors1How add nameservers?15:56
halvors1I am using static ip configuration.15:56
halvors1But i can't set static nameservers in resolv.conf :(15:57
eagles0513875sigh15:58
Naglikhello everyone15:58
patdk-wkhalvors1, if you cant set static nameservers, you can't use static15:58
patdk-wkeagles0513875, dunno, either you mounted via nfs4, and didn't setup kerberos15:59
patdk-wkor you just have user mismatchs or something15:59
eagles0513875patdk-lap_: well friends server wasnt saving the name servers on reboot15:59
eagles0513875patdk-lap_: its probably users but is there an nfs user and group that the folder should be set to?15:59
ikonianfs is shared out on a per user basis15:59
ikoniayou set the permissions/user list when you export the file system from sharing15:59
ikoniait's not a "defualt user"15:59
patdk-wkby default, nfs works on uid and groupid16:00
patdk-wknfs4 adds all kinds of other stuff also16:00
ikoniait's the same as local permissions16:00
halvors1patdk-wk: The problem is that when i set the nameservers in resolv.conf, they are being overwritten when system reboot.16:00
eagles0513875ikonia: would you like me to show you what i have in my /etc/exports?16:00
ikoniahalvors1: are you aware of the new annoying dyndns config that is being used16:00
ikoniaeagles0513875: not really16:01
patdk-wkhalvors1, then you failed to read the 12.04, how to handle resolv.conf in the resolvconf world16:01
eagles0513875halvors1: i can confirm that and this is on a dedicated server im having the same issue16:01
eagles0513875seems like though adding dns-nameservers to /etc/network/interfaces solved the issue on my friends dedicated though16:01
Nagliki'v got a lil problem writing my thesis (Master) with peformance: nginx, apache, iis, can i rely on already existing articles covering this topic?16:01
patdk-wkdo people never read the upgrade notes?16:01
ikoniapatdk-wk: no16:01
ikoniaNaglik: how is that anything to do with #ubuntu-server ?16:02
Naglikhmm, good question16:02
Naglikbad search algorithm16:03
patdk-wkyou can install nginx/apache (and iis via kvm) on ubuntu-server? :)16:03
Naglikyep :)16:03
Nagliknvm then, though here are ppl who can help me :)16:03
Naglikbb16:03
* patdk-wk would ask your teacher16:03
terinjokesi'm running 11.04, can I still upgrade to 11.10?16:10
ikoniaeagles0513875: go on then, pastebin your exports file16:10
terinjokesnvm, looks like the standard do-release-upgrade will handle it16:10
patdk-wkterinjokes, only if you go to 11.10 -> 12.0416:11
patdk-wkoh, yes :)16:11
patdk-wkfor another 6months or so16:11
jsmith-argotecanyone know how I could troubleshoot an upstart job - it's pre-start process is terminating with status 116:14
halvors1eagles0513875: Should i use "dns-nameservers" on inet6 (ipv6) interfaces too?16:14
eagles0513875halvors1: sadly i dunno as we are using ipv4 only16:14
pdtpatrickjsmith-argotec, pastebin ur pre-start script16:14
=== Lcawte is now known as Lcawte|Away
eagles0513875ikonia: the exports file i only added this line the rest is commented out /mnt/iso *(rw,sync,no_root_squash)16:16
jsmith-argotecpdtpatrick: http://pastebin.ubuntu.com/976054/16:17
Davieystgraber: hey16:17
ikoniaeagles0513875: ok, so start setting some users/groups16:17
Davieyhallyn: hey16:17
Davieyhallyn / stgraber: you wanted a session re-scheduled?16:17
eagles0513875ikonia: set them to do what though? and how to do that16:17
ikoniaeagles0513875: set them to the permissions you want16:17
stgraberDaviey: we'd like to ensure the user namepsace session scheduled for tomorrow won't move16:18
eagles0513875if i want my user i use on this netbook to have access i would need to add that user account to the system in order for it to work?16:18
Davieystgraber: i can do that.16:18
ikoniaeagles0513875: you need to the users you want to assign permissions to to exist on the machine doing the auth16:18
stgraberDaviey: cool. I think the other LXC sessions are already locked so if you can lock that one too, we'll be all good16:18
eagles0513875ikonia: ok thanks will do that now :)16:18
jsmith-argotecpdtpatrick:  seems to work ok until I start smbd with pacemaker then nmbd wont start16:19
pdtpatrickjsmith-argotec, so what happens when u run those commands manually? btw I think upstart runs shell with -e so if there's an error it exits16:19
jsmith-argotecpdtpatrick: they work fine16:19
pdtpatrickjsmith-argotec, so pacemaker blows it up ?16:20
pdtpatrickjsmith-argotec, are there anything in /var/log/syslog ?16:20
jsmith-argotecpdtpatrick: with smbd running under pacemaker control: -f exits 0, install exits 0, loading NMBD variable exits 1 and loads No into it, NMBD test exits 116:21
jsmith-argotecpdtpatrick: seems that pacemaker does blow it up... however it worked fine under 10.0416:22
jsmith-argotecpdtpatrick: I have one server upgraded and one not.. still works on lucid but not yet on precise16:23
jsmith-argotecpdtpatrick: this is what I get in syslog: kernel: [ 4900.048887] init: nmbd pre-start process (25312) terminated with status 116:23
jsmith-argotecpdtpatrick: this is what I get when trying service nmbd start: start: Job failed to start16:24
pdtpatrickjsmith-argotec, interesting. So it is testing whether it is disabled. If it is then the script exits and the service never starts? why do u need to check whether it is disabled if ur exiting 0 right after it anyway ?16:24
pdtpatrickjsmith-argotec, seems like ur prescript exits before it gets to exec nmbd -D16:25
terinjokespatdk-wk: i'm going 11.04 to 11.10 to install a working libc PPA to then do 11.10 to 12.0416:25
jsmith-argotecpdtpatrick: this upstart job is default that comes when you install samba - I haven't modified it.16:25
jsmith-argotecpdtpatrick: I believe it checks if netbios is disable and if it (exit 0 on test) then there's the stop before exiting16:26
jsmith-argotecpdtpatrick: otherwise the prestart exits 016:26
=== matsubara is now known as matsubara-lunch
jsmith-argotecpdtpatrick: I agree it never gets to exec... don't get where it's getting the exit 1 from though16:27
pdtpatrickjsmith-argotec, add -x to ur script16:27
pdtpatrickand that should allow debug mode16:27
jsmith-argotecpdtpatrick: in the prestart part?16:28
pdtpatrickyup16:28
pdtpatrickjsmith-argotec, so u would add "set -x "16:28
pdtpatrickatop16:28
pdtpatrickjsmith-argotec, http://pastie.org/private/bez20phme9bmn14yevypka16:29
pdtpatricksomething like that16:29
pdtpatrickso change -e to -x16:29
jsmith-argotecpdtpatrick: no change. no more output in syslog or terminal16:30
pdtpatrickjsmith-argotec, does pacemaker log anything ?16:30
jsmith-argotecpdtpatrick: I've been trying to start nmbd manually to rule out the pacemaker side of it16:31
jsmith-argotecpdtpatrick: but the first time I tried with pacemaker controlling the start of nmbd it did...16:31
=== Lcawte|Away is now known as Lcawte
pdtpatrickjsmith-argotec, just curious - make sure there's no process already running and pacemaker is not trying to control it or is controlling it. "crm_mon --one-shot"16:33
jsmith-argotecpdtpatrick: http://pastebin.ubuntu.com/976093/ - output when pacemaker tried to start nmbd16:33
pdtpatrickjsmith-argotec, also you can add set +e to turn off set -e which is by default with scripts. And then add set -x under the set +e to see if it does any better.16:33
roaksoaxjsmith-argotec: is that with the upstart ra?16:34
jsmith-argotecroaksoax: yep16:34
roaksoaxivoks: ^^16:34
roaksoaxivoks: http://pastebin.ubuntu.com/976093/16:35
jsmith-argotecroaksoax: samba has been upstart since 10.0416:35
jsmith-argotecor earlier16:35
roaksoaxjsmith-argotec: ivoks is the man for that ;)16:35
jsmith-argotecroaksoax but it was fine under 10.0416:35
roaksoaxjsmith-argotec: so I'm guessing that if you start nmdb with upstart directly it doesn't fail but it does with pacemaker?16:36
jsmith-argotecroaksoax: no it fails either way when I have smbd under pacemaker control16:36
jsmith-argotecroaksoax: but when I migrate everything *samba* away then nmbd starts fine16:37
jsmith-argotecpdtpatrick: crm output: http://pastebin.ubuntu.com/976098/16:37
jsmith-argotecroaksoax: ^^^16:37
roaksoaxjsmith-argotec: isn't there any logs in /var/log/upstart/etcetc.log16:38
roaksoaxs/etcetc/by the job16:38
jsmith-argotecroaksoax: shouldn't matter that pacemaker won't start nmbd due to failcount when I'm trying to manually start nmbd should it?16:38
roaksoaxjsmith-argotec: for what I can read in the fail log before (The last reference on a connection was dropped without closing the connection), it seems that there's an open connection for the daemon on dbus16:39
roaksoaxand trying to open a new one, fails16:39
jsmith-argotecroaksoax: there is a nmbd.log in /var/log/upstart but I don't understand it16:39
pdtpatrickjsmith-argotec, so maybe "crm resource stop <servicename> && crm resource cleanup <servicename> "16:40
jsmith-argotecroaksoax: http://pastebin.ubuntu.com/976115/16:40
roaksoaxjsmith-argotec: http://upstart.ubuntu.com/cookbook/#debugging16:40
* roaksoax will brb16:41
ivoksjsmith-argotec: so, once smbd is started, nmbd won't start?16:42
jsmith-argotecpdtpatrick: I have a meeting in a couple minutes I just realized... thanks for the help!  I'll be back in a bit to try again16:42
jsmith-argotecroaksoax: I have a meeting in a couple minutes I just realized... thanks for the help!  I'll be back in a bit to try again16:42
jsmith-argotecivoks: once smbd is started under pacemaker yes16:42
ivoksrun nmbd -d 10 -D16:43
jsmith-argotecivoks: I've only tried to start nmbd with smbd stopped when pacemaker wasn't controlling smbd16:43
jsmith-argotecivoks: that worked16:44
ivokswhile smbd was managed by pacemaker?16:45
jsmith-argotecivoks: yes16:45
jsmith-argotecivoks: the upstart nmbd.conf has changed slightly in 12.04 vs 10.0416:45
jsmith-argotecivoks: I thought maybe the -f smb.conf test failed due to my using a pacemaker symlink for it but i tested that line in terminal and it worked fine16:46
jsmith-argotecivoks: because the -f smb.conf test in the nmbd.conf definition is new16:47
jsmith-argotecivoks: you going to be around in about 45 minutes?16:47
jsmith-argotecpdtpatrick: btw I tried the stop/cleanup and then manually start nmbd with no change16:48
pdtpatrickjsmith-argotec, looks like ivoks fixed :) Learned something myself.16:48
jsmith-argotecpdtpatrick: well yes and no... that made it start but didn't fix pacemaker/manual starting of nmbd16:50
pdtpatrickoh16:50
jsmith-argotecivoks: when you asked "while smbd was managed by pmkr" you where referring to when I ran nmbd -d 10 -D right? 'cause that's what I answered16:50
ivoksjsmith-argotec: yes; it works for me16:51
ivoksjsmith-argotec: my guess is that starting nmbd is racing with your linking16:51
jsmith-argotecivoks: links are up before smbd or nmbd...16:52
ivoksthen it's something else16:52
ivoksstarting smbd and nmbd works for me16:52
ivoksprimitive nmbd upstart:nmbd16:52
ivoksprimitive smbd upstart:smbd16:52
ivoksclone cloneS smbd16:52
ivoksclone cloneN nmbd16:52
ivoksboth start16:52
jsmith-argotechmm16:52
hallynDaviey: thanks, we should be all set (I asked Michelle for help yesterday)16:53
jsmith-argotecthanks for testing.... I have to go to a meeting but when I get out I'll pick up where I left off16:53
hallynDaviey: http://summit.ubuntu.com/uds-q/meeting/20363/servercloud-q-lxc/ is rescheduled and should be (i hope) locked to the 9am friday spot16:54
ivoksjsmith-argotec: works as a group too16:58
jsmith-argotecivoks: tested starting smbd and nmbd manually on that node without any pcmkr and worked fine... grrrr :-)16:59
jsmith-argotecivoks: I gotta run - be back in 30 or so16:59
=== Arc__ is now known as a5m0
roaksoaxDaviey: ping?17:03
some1_2Hi, trying to install 10.04 64bit as a guest on a Xen (XCP 1.1) host, but during install the guest doesn't receive DHCP. The vif seems to be assigned to xenbr0 correctly (and xenbr0 is associated with eth0, which is the only network device). Any ideas what I might try?17:15
=== matsubara-lunch is now known as matsubara
* jkyle is at the cloud summit18:03
jkylesome cool talks18:04
=== Ursinha` is now known as Ursinha
=== Ursinha is now known as Guest66718
=== Guest66718 is now known as Ursula
=== Ursula is now known as Ursinha
CowBoyVVhello..18:27
CowBoyVVthis place help me with my ubuntu server?18:27
PiciYes.18:28
CowBoyVVhello everyone..hi Pici..18:28
CowBoyVVI am new to this ubuntu server..18:28
CowBoyVVI need to learn from very begginning..18:28
CowBoyVVhelp me at much as you can...18:28
bluefrog_CowBoyVV, http://doc.ubuntu.com/ubuntu/serverguide/C/18:29
CowBoyVVI need to set up my website...where I can access all the features of ubuntu server...how to set up ...18:29
CowBoyVVlike chatroom18:29
CowBoyVVlike security network18:29
CowBoyVVthanx bluefrog..I check it out18:30
CowBoyVVyes..18:30
CowBoyVVI do all of that already..but I need more how to set up from my own website18:33
iggiI'm having a very odd error. After reboot one of my virtual machine hosts no longer responds to ping (IP or hostname) and SSH connections. The virtual machines it is running are accessible, the router sees it's entry in the ARP table (even the router can't ping it though). I've turned off ufw and flushed iptables rules, but it's still not accessible, any ideas?18:36
iggiI've also checked netstat and it is listening for SSH connections18:39
bobweaverHello there I am making a package for ubuntu server and the upsource code was real bad re-righting apache ect I have now re-wrote the files and what too see if there is a better way to write this into the package. the bug is located here https://bugs.launchpad.net/zpanelcp/+bug/996282   Code is located here http://bazaar.launchpad.net/~josephjamesmills/zpanelcp/zpanelcp/files/head:/etc/zpanel/configs/apache/      I will be trying too18:40
bobweaverpush to /var/www/ and make avilble during build but I do not know the steps too link sites-enabled How too link ? any tutorial would be great thanks for your time18:40
uvirtbotLaunchpad bug 996282 in zpanelcp "Turning on zpanel should not re-write /etc/apache2/apache2.conf" [High,Triaged]18:40
Captain_ProtonNeed some help I upgrade my 10.04 server to 12.04 running postfix + dovecot w/managesieve now none of the filter work watching the logs it looks like it getting passed to lda which dumps it right in the Inbox18:52
bobweaverI am making a package for ubuntu server and the upsource code was real bad re-righting apache ect I have now re-wrote the files and what too see if there is a better way to write this into the package. the bug is located here https://bugs.launchpad.net/zpanelcp/+bug/996282   Code is located here http://bazaar.launchpad.net/~josephjamesmills/zpanelcp/zpanelcp/files/head:/etc/zpanel/configs/apache/      I will be trying too18:56
bobweaver push to /var/www/ and make avilble during build but I do not know the steps too link sites-enabled How too link ? any tutorial would be great thanks for your time18:56
uvirtbotLaunchpad bug 996282 in zpanelcp "Turning on zpanel should not re-write /etc/apache2/apache2.conf" [High,Triaged]18:56
bobweaversorry wrong channel18:56
Captain_Protonnm I found the problem Thanks anyway19:00
terinjokesanyone have suggestions for how to upgrade to 12.04 on an OpenVZ host correctly?19:00
terinjokesI read this: http://askubuntu.com/questions/125742/upgrade-to-ubuntu-server-12-04-brings-error-glibc-not-found19:00
terinjokesI haven't started the upgrade, but I also don't want to screw everything completely up19:01
terinjokesdo I upgrade, fail, then add the PPA?19:01
terinjokesi added the ppa here on 11.10, but it gets disabled during the upgrade19:02
Captain_Protonterinjokes, unless there is a real good reason I would not.19:02
Captain_Protonterinjokes, most of managed by pearl and you will run into all cans of problem when the perl lib upgrade19:03
Captain_Protonterinjokes, if you are hell bent on 12.04 start a new vm with 12.04 and move your data19:04
terinjokesCaptain_Proton: well I can't, the provider won't upgrade the kernel19:05
Captain_Protonterinjokes, will they create a vm with 12.0419:05
terinjokesCaptain_Proton: maybe. would have to install it somewhere else, switch out the libc then package it up19:07
Captain_Protonterinjokes, if they will not help you move vps providers are cheap and everywhere. Are you running just web apps ie.. Lamp or a custom app19:08
terinjokesrunning some lamp stuff, but also some custom stuff19:09
=== railsraider_ is now known as railsraider
Captain_ProtonI would ask them I do not see why that would not. I have down my whole VPS trying to upgrade it, will not do that again  lol19:12
ZanzacarI am trying to know when someone is connected to my ftp server,sftp,or ssh but cant figure it out.20:30
Zanzacarlastlog will output ssh. auth.log might show me who tried to login. anyways i dont know where to go from here.20:30
jsmith-argotecpdtpatrick: figured out that nmbd startup issue (with lots of help from ivoks) - my smb.conf had warnings when running testparm which resulted in exit 1 on the line checking for NMBD disabled21:07
jsmith-argotecpdtpatrick: default smb.conf tested fine so it would work fine when it wasn't trying with pacemaker21:08
jsmith-argotecpdtpatrick: thought you might like to know the final outcome21:08
=== dendro-afk is now known as dendrobates
RichardRaseleyHello all. I am interested in setting up a test Apache Cassandra box on Ubuntu 12.04 but am having difficulty figuring out how to get Java installed (sun-java6-bin doesn't appear anywhere in the repositories). Can anyone assist with this?22:11
lifelessjuju would be a good way to get cassandra up and runing22:12
lifelessit can run against lxc on your local machine22:12
RichardRaseleyI am not familiar with LXC22:14
RichardRaseleyI thought JuJu could only run against MaaS and AWS compatible services?22:14
lifelesshttp://askubuntu.com/questions/65359/how-do-i-configure-juju-for-local-usage22:14
FireboltRichardRaseley, Sun has forced Canonical to remove sun-java6* from the repos22:14
lifelessthats includes info for oneiric, precise should be similar22:14
lifelessRichardRaseley: juju has an LXC backend for local development22:15
RichardRaseleyFirebolt: Can you suggest the best way to get it at this time?22:15
FireboltThere's a ppa for it22:15
RichardRaseleylifeless: I appreciate your suggestion, but I would like to just run through the standard setup on this particular box (learning experience).22:16
lifelessRichardRaseley: juju *is* the recommended standard setup, FWIW.22:16
RichardRaseleylifeless: Recommended by Canonical or the Apache Cassandra project?22:17
lifelessUbuntu22:17
FireboltRichardRaseley, if you have your heart set on using sun-java6, sudo add-apt-repository ppa:ferramroberto/java && sudo apt-get update && sudo apt-get install sun-java6-bin22:17
lifelesscassandra is extremely hard to package22:17
=== dendrobates is now known as dendro-afk
lifelessbecause of intricacies around java local library versions and so forth22:17
RichardRaseleyFirebolt: You say that like I should be using something else?22:17
FireboltI don't know if you *should* be using something else22:18
RichardRaseleylifeless: When you say package, you mean like a .deb? I was going to just download the source from their site.22:18
FireboltI said it in such a manner because lifeless has proposed that22:18
RichardRaseleydddddddddddddAh22:19
RichardRaseleyAh*22:19
RichardRaseleylifeless: The LXC method will not work the best for me as I'd like to deploy it across several physical nodes as a test.22:20
lifelessRichardRaseley: you could use the lxc approach on each node and then hand configure them22:20
lifelessor you could bring the nodes up via maas, or openstack, and let juju do it all22:21
RichardRaseleyHmm...22:21
RichardRaseleyIf I use MaaS though I have to mess around with the DHCP infrastructure in my environment, correct?22:21
RichardRaseleyThat would be a mess (different team)22:21
RichardRaseleyOr can I manually enroll to a MaaS server?22:22
lifelessyou can manually enroll using avahi, though I don't know how much heavy-duty work that that has had22:23
RichardRaseleyI am not familiar with avahi...22:24
RichardRaseley(Reading Now)22:25
RichardRaseleyIs apt-add-repository not part of 12.04 server?22:33
slideRichardRaseley, its like in some python utils package22:38
slidewhich is really weird lol22:39
RichardRaseleyYeah, I see that now.22:39
RichardRaseleyHmph22:39
aboSamoorhow can I setup the default kernel logging level?22:43
ninjaihey guys... so I ran usermod -a -G group2 user.  When I run groups user, it shows groups: user group2.  When I try to modify files in a folder owned by the user and group (group2:group2), where permissions are set 770, I get permission denied.  What gives?22:47
=== bobweaver is now known as josephmills
=== josephmills is now known as bobweaver
rocketsPsi-Jack, hey dude. So I've got a server now running with kvm23:13
rocketseth0 = 192.168.1.4923:13
rocketsvirbr0 = 192.168.122.123:13
rocketsbut theres effectively nothing in my /etc/network/interfaces23:14
rocketsany thoughts on "transforming" virbr0 to bridging instead of NAT?23:14
Psi-Jackrockets: You don't.23:15
Psi-Jackrockets: You replace eth0 and put that into a bridge, like vmbr023:15
rocketsPsi-Jack, so how does one get a bridged network so that I can place VMs on my local lan23:15
Psi-Jackvirbr0 is a virtual bridge, not actually attached to a real NIC.23:15
rocketsok23:15
rocketsPsi-Jack, I have a dual-nic system. Would it be possible to configure my second NIC as the bridged-nic23:32
Psi-JackOf course.23:32
Psi-JackAs long as it'll be the NIC that has the connection you need.23:32
rocketsPsi-Jack, does the ethernet card im bridging to need to have a real IP of its own?23:33
rocketsid rather it didn't23:33
Psi-JackNo23:33
Psi-Jackbut it has to be connected to something.23:33
rocketsPsi-Jack, of course23:34
rocketsthanks!23:34
=== Lcawte is now known as Lcawte|Away
rocketsPsi-Jack, the official Ubuntu 12.04 server guide actually has some nice docs on this23:37
Psi-JackThey actually have 12.04-specific server guides?23:38
rocketsPsi-Jack, https://help.ubuntu.com/12.04/serverguide/serverguide.pdf23:39
rocketsyessir.23:39
rocketsMaybe they only do them for LTS releases?23:39
Psi-JackNo, they don't..23:40
rocketsPsi-Jack, the above link is a 12.04 specific guide sir,.23:44
jvargasI need to configure a network interface (eth1) neither in dhcp neither in static mode.23:47
jvargasI just want it to be always UP but with no IP address configured. How can I do that using /etc/network/interfaces?23:47
rocketsPsi-Jack, would something like this be appropriate? http://pastie.org/388168823:47
rocketsjvargas, I don't know the answer, but why, out of curiosity?23:48
jvargasrockets: I have a virtualbox machine that is reserved to use a unique IP in that address because is it a 255.255.255.252 netmask.23:49
rocketsjvargas, so?23:49
rocketsso you DO want it to have an address23:50
jvargasrockets: so, the host must not be configured with an IP, but have the interface UP so that it could be bridged to vbox machine.23:50
jvargasNo, I do not want it to have an address, just to keep UP. I can do it, but manually using ifconfig eth0 up. I wondered how to do it automatically at startup23:50
rocketsjvargas, ifconfig eth0 auto23:51
rocketswont do dhcp23:51
rocketsor rather23:51
rocketsauto eth123:51
rocketsbut no iface blah blah23:51
rocketsjust auth eth1 in /etc/network/interfaces should do it23:51
rocketsit will bring it up but set no address23:51
jvargasit is in auto mode, but it remains OFF.23:51
rocketsthat's odd.23:51
rocketsjvargas, why not just assign it a dummy private IP23:52
rockets172.31.0.1 or somethign23:52
rocketsno gateway23:52
jvargasrockets: Will try.23:54
rocketsjvargas, I just tried auto eth1 on my server23:56
rocketswith no IP23:56
rocketsit works23:56
rocketshttp://pastie.org/388172323:57

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!