UbuntuIRC / 2008 /04 /11 /#ubuntu-server.txt
niansa
Initial commit
4aa5fce
[00:05] <Kamping_Kaiser> 'doesnt work' doesnt help
[00:07] <Kamping_Kaiser> afk. going to work
[00:08] <n6rej> I don't know how to be more clear :(
[00:09] <n6rej> if I try to access the mail from my windows box it says plain text authentication can't be used, and If I say secure it says its not setup for secure, and if I telnet it says nothing at all just blank.
[00:33] <Centaur5> Could anybody that has owned or does own an Adaptec hardware raid card tell me if they have decent monitoring software that works in Ubuntu?
[04:03] <J-_> what file do I have to configure to use vhosts?
[04:18] <centaur5> I get access denied when trying to add windows xp to a samba domain. What would cause this when machine and user are added?
[05:12] <n6rej> can anyone help me create a user that can ftp into /var/www?
[05:23] <mactimes> Hi. I've just installed Ubuntu Server in a VM, using RAID1 for all partitions, to test it's behavior in case of a disk failure. I went through the same steps to install a debian server. Debian could still boot after removing one of the disks, but Ubuntu isn't. Could someone, please, help me out?
[05:52] <nijaba> mactimes: could you check if your problem is similar to https://bugs.edge.launchpad.net/ubuntu/+source/mdadm/+bug/125471 and eventually add your configuration and comment about debian on it?
[05:52] <ubotu> Launchpad bug 125471 in mdadm "Booting from a degraded array could be improved" [Medium,Confirmed]
[05:53] <mactimes> nijaba: This is the problem.
[05:53] <mactimes> nijaba: Exactly the same
[05:54] <mactimes> nijaba: I was afraid I was doing something wrong, but I reinstalled both systems 3 times each, just to make sure.
[05:55] <mactimes> nijaba: Debian won't fail, but Ubuntu will give me busy box shell.
[05:55] <nijaba> Could you add a comment to it, please, mentioning your experience with debian?
[05:55] <osmosis> if KVM is what is officially supported by Ubuntu now, how come no one has updated the doc?! https://help.ubuntu.com/community/KVM
[05:55] <nijaba> osmosis: the doc in progress is published on doc.ubuntu.com
[05:56] <nijaba> osmosis: it will be moved to help.ubuntu.com at release time
[05:56] <nijaba> osmosis: you will see that we have somehow done our homework
[05:56] <osmosis> oh.. well.. good to hear then. cheers.
[05:57] <mactimes> nijaba: Sure, I can. But I don't see the point. I've been talking to people at #debian a few days ago and, as per information provided by them, Ubuntu packages come from Debian unstable. Would there be a downgrade?
[06:00] <nijaba> mactimes: that is not completely true, and initramfs boot is one of the big differences
[06:01] <mactimes> nijaba: Hum. Ok. I'll add a comment there, but I guess people will start pursuing me for the comment...
[06:02] <nijaba> mactimes: nobody will force you to reply and you can unsubscribe from the bug at any time
[06:02] <mactimes> nijaba: Right.
[06:02] <nijaba> mactimes: thanks for your help
[06:03] <mactimes> nijaba: I'm writing the comment now.
[06:04] <nijaba> thanks a lot mactimes
[06:05] <mactimes> nijaba: I'll let you know when I'm done.
[06:07] <nxvl> does anyone has run ubuntu server on virtualbox?
[06:07] <nxvl> i can't run dapper
[06:09] <nxvl> i already run the installer
[06:09] <nxvl> but i can't run the installed one
=== macd_ is now known as macd
[06:27] <mactimes> nijaba: Just posted it there.
=== foo_ is now known as foo
[11:06] <MatBoy> hi guys !
[11:07] <MatBoy> is tehre a good tut to upgrade a debian etch box to a ubuntu-server box remotely ?
[11:07] <MatBoy> *there
[11:08] <henkjan> MatBoy: don't even try
[11:09] <henkjan> MatBoy: do a fresh install
[11:09] <henkjan> MatBoy: upgrading from debian to ubuntu is a pain
[11:11] <MatBoy> henkjan, I thought the same actually, but it seems that people have done it
[11:27] <ICU> That people have done it doesn't make it any better :)
[11:53] <_ruben> MatBoy: remote upgrade: far from recommended, remote reinstall: possible depending on configuration :)
[11:54] <MatBoy> _ruben, shut up :P no but why not ? I can try it remote, and after it I still can go to the DC :D
[11:55] <MatBoy> _ruben, are you anywhere ?
[11:56] * _ruben is here
[11:56] <MatBoy> _ruben, where ?
[11:56] <_ruben> MatBoy: why not upgrade? dependency hell awaits you
[11:56] * _ruben points
[11:57] <MatBoy> _ruben, if I need to reinstall it, I don't mind
[11:57] <MatBoy> better try it than never did it... same as with... youy know :d
[11:58] <_ruben> i wouldnt bother trying to upgrade .. i would give it a try using some debootstrap magic .. if you have a spare (temp root) partition available, and ideally a seperate /boot, then odds are looking good
[11:59] <MatBoy> yeah I better see a reinstall too, much faster... but I don't want to go to the DC really
[12:00] <_ruben> there's howto's available on how to install debian/ubuntu using debootstrap from any os for which debootstrap is avail
[12:01] <_ruben> https://help.ubuntu.com/6.10/ubuntu/installation-guide/hppa/linux-upgrade.html .. first hit in google ;)
[12:01] <MatBoy> _ruben, reinstall is easier ;)
[12:05] <_ruben> sure, but if you dont want to go to the DC than this'd be the way to go :)
[12:11] <henkjan> dist-upgrade from debian sarge to ubuntu dapper failed for me
[12:11] <henkjan> did it once in an vm
[13:55] <Folke> Hi, any thoughts about how to continue with a bugged out test install of hardy on dl360g4p?
=== blue-frog_ is now known as blue-frog
[13:56] <Folke> Do I fill out a bug or what do I do next?
[13:58] <mok0> Folke: did you start from a CD?
[13:59] <mok0> Folke: what's bugged about it?
[14:00] <Folke> mok0: Yes, it hanged on me on 2 different occasions. Both on the last thing "restarting now".
[14:01] <mok0> Folke: hmm, that has happened to me once... trying to remember what was wrong...
[14:02] <Folke> mok0: Then I did a cold reboot and the system started as it should. But when installting I hade some strange things. On the other tty's the text was borked up really good. Alot of @ and strange signs?
[14:02] <mok0> Folke: Is there enough space on the hard drive for the install?
[14:02] <Folke> Yes, It is a HP dl360g4p with 32G hwraid 1.
[14:02] <mok0> Folke: what's a dl360g4p?
[14:03] <Folke> Yes
[14:03] <mok0> Ah
[14:03] <Folke> Is there something problematic with them and hardy?
[14:03] <mok0> Folke: you can't have the boot partition on a raid AFAIK
[14:04] <Folke> mok0: But it is an ordinary hw raid. So that it presented to the system as /dev/ccsis01. So the raid is transparent for the system?
[14:05] <Folke> I will report it in the wiki, I did a test after the template in the serverteam pages.
[14:05] <mok0> Folke: we have an HP G5, but I didn't install it...
[14:05] <Folke> 380 g5?
[14:06] <mok0> Folke: In our machine, both / and /boot are on a /dev/cciss/ device
[14:06] <Folke> I planned to test on a dl380g4 and perhaps on a dl580g(?) next week.
[14:06] <mok0> Folke: yes, that's the newest of HP's blades
[14:06] <Folke> Ah.
[14:06] <mok0> Folke: Can't remember the actual model no
[14:07] <mok0> Folke: Oh, it's something with 580 too
[14:07] <Folke> mok0: We have no blades here :( Too sad, I want to play with one :)
[14:07] <mok0> Folke: or 380?
[14:07] <mok0> Folke: What version are you trying to install?
[14:08] <Folke> mok0: Just betatesting hardy. Not for production. Tried to help the serverteam with some testing on different hw.
[14:09] <mok0> Folke: ah, ok. I think you should fill out a bug report at Launchpad rather than the wiki
[14:09] <mok0> Folke: The good news is that it _should_ work, but hardy is in a pretty bad shape these days... (IMHO)
[14:10] <mok0> Folke: It's been rock-solid for months, but I have gotten problems this last week
[14:10] <Folke> mok0: When it entered beta?
[14:11] <mok0> Folke: I am not sure... but there's been around 4 kernel rebuilds during the last week
[14:11] <Folke> mok0: Auch. that's really what you not want when it goes stable :)
[14:11] <mok0> Folke: ... and I guess it takes a while for all the drivers to catch up
[14:12] <mok0> Folke: No!
[14:12] <mok0> ... fortunately there
[14:12] <mok0> is still a couple of weeks :-)
[14:12] <Folke> mok0: Yes :)
[14:13] <Folke> mok0: I cant seem to enter bugs at launchpad? Or am I looking at the wrong place? "https://bugs.launchpad.net/~ubuntu-server/"
[14:14] <mok0> Folke: Yes, you need to start at the main page, then choose project "Ubuntu", then file a bug which is aimed at Ubuntu (and not a particular package)
[14:14] <Folke> mok0: Ah, thanks!
[14:15] <mok0> Folke: when you've done with the bug, you can subscribe the ubuntu-server team, then the bug will appear on their task list
[14:17] <Folke> mok0: I am subscribing to the newslist. And asked for participation on the serverteam at launchpad. Is that the right way to go?
[14:17] <mok0> Folke, no
[14:18] <Folke> ack. I will never do right :D
[14:18] <mok0> Folke: what is the number of your new bug?
[14:18] <Folke> I filling out now.
[14:18] <mok0> Folke: when you've hit save, come back here
[14:19] <Folke> mok0: k.
[14:26] <Folke> mok0: #215686
[14:27] <mok0> bug 215686
[14:27] <ubotu> Launchpad bug 215686 in base-installer "Installation hang on dl360G4p and borked text on other consoles" [Undecided,New] https://launchpad.net/bugs/215686
[14:27] <Folke> ah, k.. I understand.
[14:27] <mok0> Folke: In the pink area under "Actions" find "Subscribe someone else"
[14:27] <Folke> mok0: aye
[14:28] <Folke> Shall I add ubuntu-server?
[14:28] <mok0> Folke: That seems reasonable
[14:28] <Folke> mok0: :)
[14:29] <mok0> Folke: if you click on the link "Ubuntu Server Team" you can now see that your bug has appeared on their list
[14:30] <Folke> mok0: Ah, then I understand how to assign bugs.
[14:31] <mok0> Folke: ... and if you go to your own LP page, it is listed on your Bugs tab
[14:31] <mok0> Folke: actually, it
[14:31] <mok0> it's pretty nifty, but you gotta know who to subscribe :-)
[14:32] <mok0> Folke: otherwise your bug gets lost among the 215685 others :-)
[14:33] <Folke> mok0: Now when vmware has certified ubuntu on their esx'es I thought that I could try to give something back to the community.. So I started to test hardy on some hw that we have here.. Then I guess that JeOS is the thing that seems most interesting to us :)
[14:33] <mok0> Folke: exactly
[14:34] <mok0> Folke: we are actually running several servers on JeOS guest machines using kvm
[14:34] <Folke> mok0: Ah, how does it do compared to "vanilla" server=
[14:34] <Folke> ?
[14:35] <mok0> Folke: you mean in terms of efficiency?
[14:35] <_ruben> minimal install (just enough, hence the name) and special kernel (-virtual flavour)
[14:35] <MagicFab> Folke, if you try JeOS beta today, be aware there is a bug in GRUB, although it's being fixed there is a workaround: https://bugs.edge.launchpad.net/ubuntu-jeos/+bug/215618
[14:35] <ubotu> Launchpad bug 215618 in ubuntu-jeos "Grub installation fails on JeOS Hardy on KVM Hardy" [Undecided,Confirmed]
[14:36] <mok0> Folke: It's a littlebit slower than running straight on the host. There is a bit of overhead in the internet bridge
[14:36] <mok0> Folke: but it's super stable
[14:37] <Folke> mok0: That sounds great! Thats what is needed. Stability is a sysadmins best dream :)
[14:37] <_ruben> hmm .. reminds me, lets upgrade my hardy jeos test install :p
[14:38] <Folke> _ruben: Are there alot of kernel switches that are different from vanilla?
[14:39] <_ruben> Folke: dont think im the right person to ask that, am just another user, hardly playing with hardy yet
[14:40] <Folke> Except from stripped and special kernel I guess JeOS is binary compatible with vanilla repo tree?
[14:41] <_ruben> the repo's are identical .. just the base install differs between desktop/server/jeos
[14:41] <lilsyko> guys
[14:41] <lilsyko> i have an ubuntu 6.06LTS server
[14:41] <lilsyko> i wanna upgrade it ver 7.10
[14:41] <lilsyko> how do i go about performing the upgrade?
[14:42] <_ruben> lilsyko: step by step .. 6.06 -> 6.10 -> 7.04 -> 7.10
[14:42] <Folke> lilsyko: Is it a critical production server?
[14:42] <_ruben> once hardy is final you'll be able to go 6.06 -> 8.04
[14:42] <lilsyko> mail relay server
[14:42] <lilsyko> quite critical
[14:42] <lilsyko> but now in my country is 9.42pm
[14:42] <_ruben> for 6.06 -> 7.10 reinstall is usually safer and faster
[14:42] <lilsyko> go its a good time for an upgrade
[14:43] <Folke> lilsyko: Ah, then perhaps wait to the next LTS?
[14:44] <lilsyko> hmm
[14:44] <lilsyko> maybe i'll do step by step
[14:44] <lilsyko> as _ruben suggested
[14:45] <_ruben> lilsyko: be ready for troubles tho, it *might* go cleanly, but dont assume so
[14:45] <lilsyko> hmmm
[14:45] <_ruben> 8.04 should be out by the end of the month .. so if you can wait, then do so ;)
[14:45] <lilsyko> i mean 6.06 to 8.04
[14:45] <lilsyko> is such a big ver jump
[14:45] <lilsyko> wont it cause more issues?
[14:45] <_ruben> but it *is* supported
[14:46] <_ruben> 8.04 is also LTS
[14:46] <_ruben> LTS -> LTS will be supported .. for non-LTS versions you have to go step by step
[14:46] <lilsyko> ok
[14:48] <mok0> _ruben: "end of the month"... does that mean that the advertised release day of April 24th has been postponed?
[14:49] <lilsyko> weird
[14:49] <lilsyko> i tried 'sudo do-release-upgrade --devel-release' pn my 6.06LTS server but it says
[14:49] <lilsyko> No new release found
[14:50] <_ruben> mok0: 24th is close enough to the end of the month for me to call it "end of month" .. didnt know the exact date by head ;)
[14:50] <zul> lilsyko: did you follow the instructions in the release notes on the wiki for the beta?
[14:50] <mok0> _ruben: ah ;-)
[14:51] <lilsyko> yeah
[14:51] <_ruben> 24th does sound awfully close though :/
[14:51] <lilsyko> https://help.ubuntu.com/community/HardyUpgrades
[14:51] <_ruben> with the current ammount of "problems" that is
[14:55] <lilsyko> so any idea guys?
[14:55] <_ruben> lilsyko: we already gave several
[14:56] <lilsyko> i mean
[14:56] <_ruben> and i doubt you want to upgrade a production server to hardy beta now
[14:56] <Deeps> lilsyko: the page you linked suggests that the command you typed was for 7.10->8.04, rather than 6.06->8.04
[14:56] <lilsyko> why isnt it showing the beta stuff? :P
[14:56] <Deeps> lilsyko: see the section Upgrade from 6.06 LTS to 8.04 LTS
[14:57] <lilsyko> thats for desktop
[14:57] <lilsyko> i was reading the part that says
[14:57] <lilsyko> Network upgrade for Ubuntu servers (recommended)
[14:57] <Deeps> ah i see
[14:58] <Folke> Do anyone know why there is LILO instead of GRUB in jeos?
[14:58] <_ruben> Folke: you sure?
[14:58] * _ruben checks
[14:58] <Folke> I looked at the bug 215618
[14:58] <ubotu> Launchpad bug 215618 in ubuntu-jeos "Grub installation fails on JeOS Hardy on KVM Hardy" [Undecided,Confirmed] https://launchpad.net/bugs/215618
[14:58] <_ruben> Folke: my jeos install has grub, not lilo
[14:58] <_ruben> tho i aint using kvm
[14:59] <Folke> _ruben: Oh, nice.. I thought that it was general in jeos.. sry
[15:00] <_ruben> ah
[15:06] <lilsyko> damn
[15:06] <lilsyko> i tired the instructions as per the website
[15:06] <lilsyko> but still no go
[15:13] <lilsyko> _ruben
[15:13] <lilsyko> wanna ask you
[15:13] <lilsyko> it says to
[15:13] <lilsyko> #
[15:13] <lilsyko> enable the "dapper-proposed" repository
[15:13] <lilsyko> so i enable it like this
[15:13] <lilsyko> #
[15:13] <lilsyko> enable the "dapper-proposed" repository
[15:14] <lilsyko> deb http://us.archive.ubuntu.com/ubuntu/ dapper-proposed main restricted
[15:14] <lilsyko> is it correct?
[15:14] <_ruben> guess so
[15:14] <_ruben> looks ok atleast
[15:15] <lilsyko> hmm
[15:17] <lilsyko> maybe need to include universe n multiverse also ?
[15:18] <_ruben> doubt that, i'd expect the upgrade tools to be in main
[15:20] <lilsyko> the weird this is u cant seem to install update-manager-core
[15:22] <_ruben> i think i have a virtual 6.06 install somewhere, forgot on which server tho :/
[15:22] <lilsyko> this is my sources.list file
[15:22] <lilsyko> http://pastebin.com/d3343aa76
[15:22] <lilsyko> is there an error somewhere?
[15:23] <_ruben> looks ok at first sight, sudo apt-get update doesnt show anything odd ?
[15:23] <lilsyko> nope
[15:24] <_ruben> then the sources.list should be ok
[15:26] <_ruben> hmm ..packages.ubuntu.com seems to be down
[15:27] <_ruben> when browsing the dapper-proposed files i dont see any update related stuff either.. wouldnt know where to go from there
[15:28] <peterdv> lilsyko: as Folke pointed out 8.04 is currently in beta test. It has not been released yet. It is not advisable to install the current version on a production server. As _ruben pointed out, you will have troubles.
[15:28] <Deeps> heh
[15:28] <_ruben> wtf .. im seeing unicode stuff or smth
[15:28] <Deeps> yeah me too
[15:28] <Deeps> lots of \ufeff
[15:29] <_ruben> hmm .. irssi on linux shows em fine, irssi on windows doesnt
[15:29] <Deeps> irssi on freebsd doesn't like it
[15:29] <_ruben> gotta love irssi's proxy
[15:29] <_ruben> guess the windows version isnt too keen on unicode ..lets check settings
[15:30] <_ruben> or perhaps its the irssi proxy that messes things up .. oh well
[15:32] <henkjan> _ruben: you are using irssi on windows? or just ssh with putty to an linux machine running irssi?
[15:33] <_ruben> henkjan: irssi on windows connecting to an irssi on ubuntu
[15:33] <_ruben> nasty setup, i know :)
[15:33] <henkjan> :)
[15:33] <_ruben> linux one's running in a screen with irssi proxy enabled
[15:35] <lilsyko> yeah i think i'm gonna abandon the project of goin for 8.04
[15:35] <lilsyko> maybe i'll stick with 7.10
[15:36] <peterdv> lilsyko: ... or just wait untill the release planned at the end of the month.
[15:53] <sommer> sweet
[16:32] * delcoyote hi
[16:44] <lilsyko> guys
[16:44] <lilsyko> will JeOS work on a physical system ?
[16:44] <lilsyko> instead of a VMWARE server?
[16:45] <zul> I dont think so but I have never tried it
[16:45] <nijaba> lilsyko: no it will not
[16:45] <lilsyko> hmmm k
[16:46] <lilsyko> just tot it may work
[16:46] <nijaba> lilsyko: it is meant to work only with vmware server, ESX (and KVM in 8.04)
[16:46] <nijaba> lilsyko: we have stripped the kernel clean of all unecessary drivers, so it should not
[16:47] <lilsyko> but JeOS is quite slow to install in a vmware session.
[16:47] <nijaba> lilsyko: you can do a minimal iso install instead
[16:47] <lilsyko> i'm installing 1 right now
[16:47] <lilsyko> but it stoped at the screen 'select and install software'
[16:47] <nijaba> lilsyko: try ubuntu-vm-builder on 8.04.
[16:47] <lilsyko> 90%
[16:48] <nijaba> lilsyko: the mini iso is stuck?
[16:48] <nijaba> lilsyko: which version?
[16:48] <lilsyko> JeOS
[16:48] <lilsyko> jeos-8.04-beta-jeos-i386.iso
[16:49] <nijaba> lilsyko: on real hw? surprised you got so far
[16:49] <lilsyko> nope
[16:49] <lilsyko> on VMware server 1.0.4 on gOS
[16:49] <nijaba> lilsyko: plus there is a (transitinal) problem today with the repositories for hardy, so installs will fail anyway
[16:50] <lilsyko> damn
[16:50] <lilsyko> looks like i have to download a new iso
[16:50] <lilsyko> but i didnt download the beta iso
[16:51] <lilsyko> sorry
[16:51] <lilsyko> i didnt download the daily iso, i downloaded the beta iso
[16:51] <nijaba> lilsyko: the issue is not with the iso, but with the repository
[16:51] <lilsyko> ok
[16:51] <nijaba> lilsyko: so keep your iso and wait til tomorrow
[16:51] <nijaba> lilsyko: https://bugs.launchpad.net/bugs/215618
[16:51] <ubotu> Launchpad bug 215618 in ubuntu-jeos "Grub installation fails on JeOS Hardy on KVM Hardy" [Undecided,Fix released]
[16:52] <lilsyko> hmmm so
[16:52] <lilsyko> JeOS downloads the packages from the repo as it installs?
[16:52] <lilsyko> intresting
[16:54] <nijaba> lilsyko: no, it updates them
[16:55] <nijaba> lilsyko: the ones that needs to, that is
[16:55] <lilsyko> ok
[16:55] <lilsyko> i mean can i do an offline intallation ?
[16:56] <nijaba> lilsyko: the objective is to provide the user with a secure system at first boot, but yes, an offline install should work today
[16:56] <lilsyko> ok
[16:56] <lilsyko> is the kernel compiled specifically for vmware?
[16:57] <lilsyko> is vmware-tools installed into jeos by default?
[16:57] <nijaba> lilsyko: yes (and kvm), it is named -virtual for that reason
[16:57] <nijaba> lilsyko: no, because there is one version of vm-tools per vmware type (and version)
[16:57] <nijaba> and we don't need it for kvm
[16:57] <lilsyko> ok
[16:59] <nijaba> lilsyko: there is an open-vm-tools version in the hardy repo, but it is not supported officially by vm-ware and does not provide all functionalities yet
[16:59] <lilsyko> k
[16:59] <lilsyko> i wanna transission frm gentoo to JeOS
[16:59] <lilsyko> thats why i'm intrested to get it working on vmware
[17:00] <nijaba> lilsyko: sure, sounds like a good plan
[17:01] <MatthewMetzger> Hello
[17:02] <MatthewMetzger> I'm using Hardy server and I'm having a problem with RAM being cached and not released. I ran the same services on a slower machine with Dapper LTS and did not experience this problem. I can give more specifics about services running and hardware, if anyone is interested.
[17:05] <nijaba> MatthewMetzger: do you have steps to reproduce your issue, or does it involve some sensitive code/data?
[17:06] <kirkland> MatthewMetzger: hi, i'd like to hear more about your issue
[17:06] <MatthewMetzger> nijaba: no sensitive code or data. I'm running all services off of packages installed from ubuntu's repositories.
[17:06] <MatthewMetzger> The only change in services is that I'm now running webdav extension of apache.
[17:06] <MatthewMetzger> I'm running squid and squidGuard, which are known to have memory issues
[17:07] * nijaba letting kirkland dive into this
[17:07] <MatthewMetzger> but I didn't have memory issues on Dapper LTS
[17:07] <kirkland> MatthewMetzger: i386? amd64? same on both?
[17:08] <mathiaz> MatthewMetzger: how do you know that RAM is cached and not released ?
[17:08] <MatthewMetzger> i386 (Mac mini Intel hardware)
[17:08] <MatthewMetzger> mathiaz: I'm watching it with "free -m"
[17:09] <MatthewMetzger> I also run "ps aux" and see that very little is actually actively being used by applications.
[17:09] <lilsyko> yay~ i managed to upgrade my 6.06LTS to 8.04beta
[17:09] <lilsyko> with no issues so far
[17:09] <kirkland> MatthewMetzger: how much total memory?
[17:09] <MatthewMetzger> 1 GB
[17:10] <kirkland> MatthewMetzger: can you paste the output of "free" ?
[17:10] <MatthewMetzger> sure
[17:11] <MatthewMetzger> total used free shared buffers cached
[17:11] <MatthewMetzger> Mem: 979 951 28 0 81 789
[17:11] <MatthewMetzger> -/+ buffers/cache: 80 899
[17:11] <MatthewMetzger> Swap: 2870 25 2845
[17:11] <Deeps> lilsyko: other than all the issues you had trying to do the upgrade? what was the problem in the end?
[17:11] <MatthewMetzger> I have run "sync; echo 3 > /proc/sys/vm/drop_caches" to clear the caches (I know it is not recommended)
[17:12] <lilsyko> only had a few issues before the upgrade
[17:12] <lilsyko> it could locate the upgrade
[17:12] <lilsyko> so what i did was removed update-manager-core
[17:13] <lilsyko> del /var/lib/update-manager
[17:13] <lilsyko> enable update-proposed in sources.list
[17:13] <kirkland> MatthewMetzger: you wouldn't happen to have a similar dump of "free" from dapper, huh?
[17:13] <MatthewMetzger> After restarting the machine or clearing the RAM cache, free shows that the cache slowly builds up after a period of 8 hours or so.
[17:14] <lilsyko> enable also multiverse & universe for dapper-proposed
[17:14] <MatthewMetzger> kirkland: not with the same services. no.
[17:14] <Deeps> heh, quite a few steps then
[17:14] <lilsyko> then reinstall update-manager-core
[17:14] <lilsyko> then off you go
[17:14] <Deeps> hehe
[17:14] <Deeps> i wont be upgrading my 6.06lts then
[17:14] <Deeps> although it should be out of commission by that time
[17:15] <Deeps> by the time 6.06 is no longer supported, that is
[17:15] <MatthewMetzger> Deeps: I had to use Hardy instead of Dapper LTS because I was moving to Mac mini hardware
[17:15] <Deeps> MatthewMetzger: um, ok?
[17:15] <MatthewMetzger> sorry, crossed conversations, maybe :)
[17:16] <Deeps> i think so!
[17:16] <Deeps> :)
[17:17] <MatthewMetzger> I think my problem may be related to squid, but I'm not sure how to isolate it to make sure.
[17:17] <MatthewMetzger> And, like I said, it ran fine with the same configuration on Dapper
[17:18] <MatthewMetzger> I suppose I could set up another machine with Hardy, set up squid with the same configuration and see if it has the same memory issues. That would be a lot of work, though.
[17:18] <kirkland> MatthewMetzger: can you just turn off squid?
[17:18] <lilsyko> hmmm so far my hardy is good
[17:18] <kirkland> MatthewMetzger: /etc/init.d/squid stop
[17:18] <lilsyko> will test it out a while and let u guys know k
[17:18] <lilsyko> btw this is a heavyly used production email relay server
[17:18] <MatthewMetzger> kirkland: I can turn it off for testing, of course, but it is a necessary service.
[17:19] <lilsyko> relaying about 1000+ mails a day
[17:19] <lilsyko> :P
[17:19] <nijaba> lilsyko: postfix, no dovecot?
[17:19] <kirkland> MatthewMetzger: sure, i just meant to isolate it
[17:19] <kirkland> MatthewMetzger: perhaps turn it off for one boot, reboot
[17:19] <kirkland> MatthewMetzger: give it a fixed amount of time
[17:19] <kirkland> MatthewMetzger: 10 minutes? 30 minutes? 60 Minutes?
[17:20] <kirkland> MatthewMetzger: do the same with a fresh boot with/without squid
[17:20] <MatthewMetzger> I can't turn it off right now as people are actively using the proxy as we speak, but I can test it over the weekend perhaps. I can maybe do it this evening.
[17:20] <kirkland> MatthewMetzger: measure "free" at regular intervalas
[17:20] <kirkland> intervals
[17:20] <MatthewMetzger> Thanks kirkland
[17:20] <kirkland> MatthewMetzger: one more question....
[17:21] <kirkland> MatthewMetzger: is the high cache usage causing noticeable performance problems, or are you just a memory tuning "nut"? :-)
[17:21] <lilsyko> postfix only
[17:21] <nijaba> lilsyko: thanks
[17:21] <MatthewMetzger> kirkland: :) It's causing problems. Squid stopped on its own because it ran out of memory. I had to restart it. Then later, the whole machine became unresponsive and had to be rebooted.
[17:22] <kirkland> MatthewMetzger: arg, that is bad
[17:22] <kirkland> MatthewMetzger: how much disk cache have you given Squid?
[17:22] <MatthewMetzger> Very different performance than from Dapper.
[17:23] <MatthewMetzger> very little, as I'm mainly using squidGuard as a proxy filter and don't really care about the cache. I think it is actually only 1 MB.
[17:23] <kirkland> MatthewMetzger: i'm going to install Hardy and Dapper + squid in a pair of VM's right now
[17:24] <kirkland> MatthewMetzger: will you be around a bit, such that I can duplicate your configuration reasonably?
[17:24] <MatthewMetzger> kirkland: I have to do lunch and will be gone for an hour or so after. I can hover on the channel a bit later today.
[17:24] <MatthewMetzger> by the way: cache_mem 1 MB
[17:25] <kirkland> MatthewMetzger: in any case, could you open a bug in Launchpad against squid?
[17:25] <MatthewMetzger> from the squid.conf
[17:25] <kirkland> MatthewMetzger: subscribe me to it (kirkland@canonical.com)
[17:25] <kirkland> MatthewMetzger: and we can communicate that way if you're not in IRC
[17:25] <MatthewMetzger> kirkland: I haven't opened a bug before, but it'd be a great learning experience for me. I'll try to do it later this afternoon :)
[17:26] <kirkland> MatthewMetzger: thats fine, ping me here if you need help opening the bug
[17:26] <MatthewMetzger> At this point, I'm just guessing it's squid.
[17:27] <nijaba> MatthewMetzger: go to https://bugs.launchpad.net/squid/ and click on 'report a bug'
[17:27] <MatthewMetzger> thanks nijaba. I'll be back later.
[17:31] <chimp> Hey ive just set up ubuntu server on a new system, but strangely i cant get it to connect to my local network
[17:32] <chimp> it has 2, gigabit lan connectors, but neither seem to connect
[17:33] <chimp> ive tried turning off one of the lan connectors in the bios, but that didnt help
[17:35] <chimp> is there an automated way i can try connecting it, ive set up the /etc/network/interfaces file identical (cept for its ip) to another ubuntu server i have running, and it connects the other server fine
[17:45] <dthacker-work> chimp: does 'sudo ifconfig' show a network address set?
[17:45] <dthacker-work> chimp: and I have to ask this :) If you have a dual nic, are you plugged in to the right port?
[17:46] * dthacker-work wouldn't ask if he hadn't made the mistake himself......
[17:52] <chimp_> argh
[17:53] <chimp_> dthacker-work: for sudo ifconfig, using both ports it shows inet addr as the one i set statically
[17:53] <chimp_> if i set dhcp, then it tries a few times, then says no offers received
[17:53] <chimp_> this is trying on both ports
[17:54] <chimp_> Its a gigabit ethernet port, but the router its connect to is only 100mbit, could this cause the problem?
[17:54] <chimp_> And is there any other config files for networking other than /etc/network/interfaces that i should worry about
[17:54] <dthacker-work> chimp_: does ifconfig show packets being received and sent? If so, can you ping the device you are connected to?
[17:55] <chimp_> pinging the router does nthing, and it doesnt receive packets me things, it has many dropped packets
[17:55] <dthacker-work> chimp_: can you pastebin your /etc/network/interfaces?
[17:56] <chimp_> hard to do considering its not connected to the net :P
[17:56] <chimp_> Any specifc part that would be interesting?
[17:56] <dthacker-work> ummmm yeah
[17:57] <dthacker-work> what's your server ip address and router ip address?
[17:58] <chimp_> inet addr: 192.168.1.42 Bcast:192.168.1.255 Mask:255.255.255.0
[17:59] <dthacker-work> that's the server, right?
[17:59] <chimp_> yep
[18:00] <chimp_> the router is on 192.168.254
[18:00] <chimp_> Its really odd as i said, another ubuntu server is connect absolutely fine
[18:00] <dthacker-work> what's your default route?
[18:02] <dthacker-work> chimp_: do you have a gateway set?
[18:02] <chimp_> 192.168.1.254
[18:02] <dthacker-work> is it showing up in netstat -rn?
[18:03] <chimp_> never done that before
[18:03] <chimp_> it gives 2 sets
[18:03] <dthacker-work> look for the one that starts with 0.0.0.0
[18:04] <dthacker-work> 0.0.0.0 10.0.35.250 0.0.0.0 UG 0 0 0 eth0
[18:04] <dthacker-work> that's mine
[18:04] <chimp_> 0.0.0.0 192.168.1.254 0.0.0.0 ug 0 0 0 eth0
[18:04] <chimp_> thats the second one
[18:04] <dthacker-work> looks good
[18:04] <chimp_> the first one has destination of 192.168.1.0
[18:05] <dthacker-work> we could look at the link
[18:05] <dthacker-work> 'sudo ethtool eth0' or whatever number you are working with
[18:06] <chimp_> ahh
[18:07] <chimp_> it says its speed is 1000Mb/s
[18:07] <chimp_> it shouldnt be
[18:07] <chimp_> the router cant handle gigabit
[18:08] <dthacker-work> that might be it, but most newer cards auto-negotiate.
[18:08] <chimp_> erm wtf
[18:08] <chimp_> its working now
[18:08] <dthacker-work> \o/
[18:08] <chimp_> found out why
[18:09] <dthacker-work> .....and?
[18:09] <chimp_> my housemate just plugged it all into a gigabit swithc
[18:09] <chimp_> so it really didnt like being plugged into 100mbit
[18:10] <dthacker-work> hehe. cool, glad you found the trouble.
[18:11] <chimp_> lucky we had some gigabit stuff lieing around
[18:14] <chimp_> How does ubuntu server compare to other linux servers, btw?
[18:16] <leonel> the differences are support and software updates
[18:16] <leonel> is the same kernel and same server programs
[18:20] <infinity> (more or less)
[18:20] <infinity> We do put some effort into integration of said applications.
[18:24] <leonel> infinity: and very good integration
[18:24] <leonel> infinity: and that's one of the ubuntu features that made a big difference JUST WORKS !
[18:25] <leonel> chimp_: and another is the great community supporting ubuntu
[18:26] <chimp_> This is very true
[18:26] <chimp_> This is annoying, i rebooted the server and it will no longer connect again
[18:26] <chimp_> i didnt even change anything
[18:26] <chimp_> :S
[18:26] <chimp_> Is there a way for me to configure it to only use 100mbit
[18:27] <Deeps> as in, force your link speed to be 100mbit?
[18:27] <Deeps> rather than autoneg to something else like 10mbit half duplex?
[18:28] <chimp_> yes
[18:28] <Deeps> could do via the interfaces file
[18:29] <pr0le> mii-tool, but it's probably not persistent
[18:30] <Deeps> yeah, i was gonna suggest adding a post-up line to your interface configuration
[18:31] <chimp_> before when i did sudo ethtool eth0 it gave me a list of possible connection speeds ie. 10/100/1000
[18:31] <Deeps> post-up ethtool -s <int> autoneg off speed 100 duplex full
[18:31] <chimp_> now it doesnt show them
[18:31] <Deeps> sudo ethtool <interface>
[18:32] <Deeps> eg, sudo ethtool eth0
[18:32] <chimp_> yer, what im saying is that before i rebooted it listed possible speeds, doesnt anymore
[18:32] <Deeps> is the cable plugged in?
[18:32] <Deeps> what output does it give?
[18:35] <chimp_> supported ports: fibre, link modes: 1000baseT/Full, auto-nego: yes Advertised link modes: not reported, advertised auto-negotiation: yes speed: 1000Mb/s, Duplex full, port fibre
[18:35] <chimp_> and a few others after that
[18:35] <chimp_> Its very odd that it worked then stopped working
[18:36] <Deeps> advertised link modes: not reported, suggests an issue on the other side
[18:36] <chimp_> It gave speeds before i rebooted tho
[18:37] <chimp_> turning off router/switch etc.
[18:37] <chimp_> brb
[18:44] <chimp> well that didnt work
[18:50] <chimp> this is driving me up the wall
=== joerlend_ is now known as XiXaQ
[19:11] <chimp> Its basically dropping all packets it receives
=== MagicFab is now known as MagicFab_brb
[19:47] <chimp> what command would show me all ethernet connections possible?
[19:50] <chimp> lsbvwp11
[19:50] <chimp> oops
[19:51] <zul> ifconfig -a
[19:51] <MatthewMetzger> kirkland: the problem wasn't squid. At least, I don't think so. Did you find any differences between Dapper and Hardy concerning squid?
[19:52] <MatthewMetzger> I think I have a memory leak in a perl website cgi script.
[20:00] <frame45> does anyone know how I can get an ftp domain name to work? ex: ftp.mysite.com
[20:02] <m11> hello
[20:03] <m11> i have problem with my DHCP server not starting. i added interface to /etc/default and added all needed information to /etc/dhcp3/dhcp3-server , where can i check more why is failing to start ?
[20:04] <lilsyko> guys
[20:05] <leonel> frame45: take a look at : https://help.ubuntu.com/ubuntu/serverguide/C/ftp-server.html and https://help.ubuntu.com/ubuntu/serverguide/C/dns.html
[20:05] <lilsyko> how do i remove unused packages on a debian server?
[20:05] <leonel> m11: check in /var/log/syslog
[20:05] <lilsyko> i have removed a few core packages
[20:05] <m11> tnx leonel
[20:05] <lilsyko> now i wanna remove the additional packages that were installed with it and are no longer used
[20:06] <lilsyko> hwo do i do that?
[20:07] <frame45> i think it's: sudo apt-get remove (name of package)
[20:08] <lilsyko> yeah i have done that
[20:08] <lilsyko> i'm running this command now
[20:08] <lilsyko> dpkg -P $(dpkg --list | grep ^rc | awk '{ print $2; }')
[20:08] <frame45> thanks leonel
[20:08] <lilsyko> is it ok ?
[20:10] <m11> it says:no subnet declaration for eth0 (0.0.0.0) , but that is first setting in my config:subnet 192.168.1.0 , what i am doing wrong ?
[20:11] <m11> can i add interface to dhcpd.conf somehow ?
[20:25] <mathiaz> sommer: Have you seen the mentoring request on ubuntu-doc ?
[20:27] <sommer> mathiaz: yep
[20:27] <sommer> are you referring to the most recent one?
[20:31] <mathiaz> sommer: yes - grab him !
[20:31] <sommer> mathiaz: does he live in north carolina too?
[20:31] <sommer> heh, yep I was plannig on responding
[20:32] <sommer> got caught up with some sql stuff this afternoon
=== MagicFab_brb is now known as MagicFab
[20:41] <sommer> mathiaz: he's grabbed... heh
=== blue-frog_ is now known as blue-frog
[20:41] <mathiaz> sommer: kwel - I was wondering what he could work on
[20:42] <mathiaz> sommer: since the serverguide is almost done for hardy
[20:42] <sommer> mathiaz: ya, I asked that myself... I pointed him toward the Server Idea Pool page with the new content ideas
[20:43] <sommer> mathiaz: if he replies back I'll suggest the help wiki as well :)
[20:54] * faulkes- grumbles
[20:55] <sommer> faulkes-: it'll be okay
[20:56] <MatthewMetzger> mathiaz: does the serverguide for hardy include info on running virtual machines so that they're on the same local network as other real machines (not a NAT creating a virtual network)?
[20:56] <MatthewMetzger> I'm interested in looking over the serverguide if there's a url for it
[20:57] <sommer> MatthewMetzger: sure does: http://doc.ubuntu.com/ubuntu/serverguide/C/index.html
[20:58] <MatthewMetzger> sommer: Thanks!
[20:58] <sommer> np
[21:10] <faulkes-> sommer: people who don't read email and wait until the last second to do things, well, they make me want to drink more than I do anyways
[21:11] <sommer> heh, I know the feeling
[21:35] <faulkes-> there is a terrible shortage of alcohol in the office currently, it is most distressing
[21:38] <sommer> you should get a job at a brewery or distillery :)
[21:57] <balzac> hello
[21:57] <balzac> anyone not idling ?
[21:58] <saltedlight> hi everyone. i need some advices to setup an dns auto-updater on a server with a dynamic ip... anyone have any ideas?
[21:59] <balzac> helo
[21:59] <saltedlight> :))
[21:59] <saltedlight> hi again :D
[21:59] <balzac> I'd like to help but I don't know anything about it
[22:00] <balzac> saltedlight: are you using eBox?
[22:00] <saltedlight> no
[22:01] <balzac> it's going to become a part of ubuntu-server, I think
[22:01] <saltedlight> what i realy need is a script or something to check my ip and if is changed to auto-update the ip on the dns server...
[22:01] <balzac> http://digg.com/linux_unix/eBox_slated_to_be_the_official_Ubuntu_server_management_tool
[22:01] <saltedlight> never used eBox 'till now...
[22:02] <balzac> it gives one centralized administrative interface for lots of possible networking configurations
[22:02] <balzac> not sure if it would help you though
[22:02] <balzac> I'm pretty ignorant of networking. It's very complex.
[22:04] <saltedlight> the only thing that i need is to have a dns instead of a ip... and my isp has the ridiculous idea to change the ip's when tey think will be 'good' to have some changes...
[22:05] <saltedlight> i dont realy know how to search this on manuals or any other kind of docs related to networking...
[22:06] <balzac> All the documentation I find on eBox assumes the reader knows networking already
[22:07] <balzac> NAT, DHCP, all kinds of other crap I don't know much about
[22:07] <balzac> it's a bloody mess of different networking protocols, none of them simple and they can interfere with each other
[22:08] <saltedlight> ya right... some one said long time ago about ubuntu: "linux for human kinds? well i need linux for dummies!"
[22:08] <saltedlight> ;))
[22:09] <balzac> I should be a software designer because I could build applications which can be used without a PhD
[22:09] <balzac> most software only gives you hints
[22:10] <balzac> most writers of software documentation write from a subjective point of view
[22:10] <saltedlight> i know... there are tons of scripts that do auto-update... but none is doing the job by himself... it has to be started manualy...
[22:11] <balzac> can you make a cron job to call one of those scripts?
[22:12] <balzac> or a cron job to call a shell script to call your update script?
[22:12] <saltedlight> and will that crontab or what you may call it will do the job if the ip will be changed by the @#$%^& ISP?
[22:13] <balzac> maybe have a variable which gets the current ip? easier said than done...
[22:13] <saltedlight> they changed my ip twice in a half hour... and this still being connected... dont know how...
[22:14] <balzac> ok, your computer is a client of their DHCP server, in order to follow their IP changes, right?
[22:14] <saltedlight> yes
[22:14] <saltedlight> i have to connect with pppoe
[22:15] <saltedlight> and then tey give me wan ip
[22:15] <balzac> is there any way for your update script to query their DHCP server or your DHCP client to get the latest IP?
[22:16] <m1r> i have set eth0 as default interface in /etc/default/dhcp3-server for DHCP server, but it constanly going out on wlan0, anyone have idea what is going on ?
[22:16] <balzac> it's easy to conceptualize, much more difficult to do.
[22:17] <CannibalM> oh thank god, there is a server room!
[22:18] <balzac> yeah, but it's very small
[22:18] <CannibalM> Ok, Proliant 1600, network cards are not functioning. I am newer with linux so have yet to get a non functioning network card to work. So, any ideas?
[22:18] <balzac> I think Canonical needs to sell more LTS licenses
[22:18] <CannibalM> I already have 5 other boxes running server edition hosting stuff, but yet can't figure this small thing out.
[22:19] <kirkland> CannibalM: what version?
[22:19] <CannibalM> It's server 7.10
[22:19] <kirkland> CannibalM: what kind of network card?
[22:19] <CannibalM> shows up during the configurations, and shows the card type, when I go to run from there, ifconfig shows nothing.
[22:19] <CannibalM> one sec, gonna pull the exact model.
[22:19] <saltedlight> so will this be possible? boot > connect pppoe > update dns on no-ip.org or something > 'fill' if the ip changes > if ip change then re-update dns ??
[22:20] <balzac> that question is over my head
[22:21] <balzac> I was just hoping I could get some help with eBox, but their channel is only a handful of people
[22:21] <balzac> I know jack-squat about networking
[22:21] <saltedlight> anyone else please?
[22:22] <CannibalM> So exact model not sure, it just reads "Compaq Netelligent Integrated 10/100 T"
[22:22] <m1r> saltedlight: no-ip should work automaticly
[22:22] <CannibalM> thats all ubuntu shows, and thats all I know. Integrated.
[22:22] <CannibalM> As well, I have an expansion card with duel ports, if those would be easier I can pop the card and check the model as well.
[22:23] <saltedlight> and if i get disconnected will the no-up client re-update the ip or i'l have to rerun the command manually?
[22:24] <m1r> saltedlight: update on every conection
[22:25] <CannibalM> Any thoughts anyone?
[22:26] <saltedlight> and how exactly i have to do that? using a cron job or what?
[22:26] <m1r> saltedlight: http://no-ip.org
[22:27] <kirkland> CannibalM: you can check the hardware much easier than popping the box
[22:27] <kirkland> CannibalM: use "lshw" and "lspci -v"
[22:27] <balzac> CannibalM: http://ubuntuforums.org/showthread.php?t=698747 <-- this thread might help
[22:29] <balzac> https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/99821
[22:29] <ubotu> Launchpad bug 99821 in linux "Compaq Netelligent 10/100 TX PCI TLAN driver missing" [Medium,Fix released]
[22:29] <balzac> "Compaq Netelligent 10/100 TX PCI TLAN driver missing"
[22:30] <CannibalM> I'm reinstalling ubuntu now.
[22:30] <CannibalM> I was thiiiiiiiiiiis close to using MS server 03
[22:30] <CannibalM> but sc#@$@ that.
[22:30] <CannibalM> lol
[22:30] <balzac> drivers is the only thing winblows has going for it
[22:30] <CannibalM> exactly.
[22:30] <balzac> I think M$ subsidizes hardware companies to keep drivers out of our hands.
[22:31] <CannibalM> so the question is, where is that info on the missing driver in linux?
[22:31] <CannibalM> the launchpad section:
[22:31] <balzac> that is the question
[22:32] <balzac> maybe an older kernel would help
[22:32] <balzac> older ubuntu v6.1 perhaps
[22:32] <balzac> but that's no fun
[22:32] <balzac> the linux kernel shouldn't need to have driver compatibility built in
[22:32] <balzac> it should all be modular
[22:34] <CannibalM> perfect answer,
[22:34] <CannibalM> hopefully you guys can explain this to me then...
[22:35] <CannibalM> how can I check the kernal on my server? Like the actual kernal version?
[22:35] <CannibalM> *kernel
[22:35] <m1r> uname -a
[22:36] <CannibalM> hmm ok cool thanks,
[22:36] <CannibalM> so in theory It should install the driver automatically then
[22:37] <m1r> for ?
[22:37] <CannibalM> But once the install is finished and I log in, ifconfig shows nothing but a loopback
[22:37] <CannibalM> as well, during install I can't get an IP.
[22:37] <m1r> for ethernt card?
[22:38] <CannibalM> exactly.
[22:38] <CannibalM> pulls nothing it seems.
[22:38] <CannibalM> Meanwhile, I have like onboard, a dual port ethernet card, and a third single card.
[22:38] <CannibalM> nothing can get an IP during install. So I'm just trying to get any of them to function and install
[22:39] <CannibalM> I figure, one port functioning is all I need for now to start developing on it and ssh'ing in.
[22:39] <m1r> <sugests realtek
=== jjesse_ is now known as jjesse
[23:36] <Zylogue> hello, all...I used Ubuntu server 5.x a while back. I recall in it was a web fronted mp3 player. This app would be great for a multimedia server I'm looking at building. Does anyone recall the name of this app, or similar app? I would like the music to play locally to that machine, out the soundcard audio jack and into the stereo. Thanks
[23:51] <mathiaz> kirkland: https://bugs.launchpad.net/bugs/215998
[23:51] <ubotu> Launchpad bug 215998 in squid "possible memory leak in Hardy's squid" [Undecided,New]
[23:51] <mathiaz> kirkland: were you able to reproduce it ?