2008/10/31/#ubuntu-server.txt

[01:38] <CarlFK> if I want to sample sound from a sound card using alsa, what .deb to i install?
[02:25] <mynous> how can you view your NIC settings on a remote server?
[02:26] <tonyyarusso> ssh to it
[02:27] <mynous> just ssh'ing to it doesnt give me the NIC settings
[02:27] <tonyyarusso> Then ifconfig, like you would locally.
[02:31] <mynous> tonyyarusso: im looking form the speed and duplex setting of my NIC
[02:35] <tonyyarusso> mynous: hmm, I'm not aware of a built-in way of getting that.  You may need to use a network monitoring utility of some sort.
[02:35] <mynous> ahh
[02:36] <mynous> i know ethtool to set it but i wanted to check and see what it was set at
[02:36] <tonyyarusso> can ethtool do a dump of current settings?
[02:37] <mynous> im not sure
[02:37] <tonyyarusso> sudo ethtool DEVICE maybe
[02:37] <mynous> would that be the EEPROM?
[02:38] <tonyyarusso> (although that gives me "Unknown!" in both fields here)
[02:38] <mynous> ahh, thats it
[02:38] <mynous> ty
[02:39] <mynous> for some reason it doesnt seem like its running full duplex
=== macd_ is now known as macd
[04:40] <CarlFK> dpkg-shlibdeps: failure: no dependency information found for /usr/lib/debug/libm.so.6 (used by debian/transcode/usr/lib/transcode/filter_levels.so
[04:40] <CarlFK> how did that happen?
[04:41] <CarlFK> trying to build transcode same way I have alwas done it.  and that magically happened
[04:41] <CarlFK> lets see if this changes it: unset LD_LIBRARY_PATH
[04:42] <Wavesonics> god i wish some one could help me... I'm setting up a RAID 10 array with 4 disks, each disk as 2 partitions on it, 1 encrypted (for swap) and 1 set up for RAID10 by mdadm, and 1 disk as a standard ext3 /boot partition. I installed just fine to the RAID10 md0 device, but when I try and boot, the first encrypted swap partition is decrypted fine, and then the md0 encrypted RAID10 device is "not found"... How can I make it recogni
[04:42] <Wavesonics> ze my multi-disk device?
[04:54] <antdedyet> Hmm.
[04:54]  * antdedyet is on the search to find out why apt can't authenticate some packages
[04:55] <antdedyet> ... in ibex
=== tonyyarusso is now known as anthony
[05:36] <Wavesonics> anyone who is interested, please take a look and vote up: http://brainstorm.ubuntu.com/idea/15043/
[06:20] <Wavesonics> in 8.10, how do I enable apache user directories?
[06:22] <tonyyarusso> Wavesonics: http://httpd.apache.org/docs/2.2/howto/public_html.html
[06:34] <Wavesonics> <tonyyarusso> thanks
[06:41] <Wavesonics> if I want to have a user able to edit /var/www what group should I put him in?
[06:56] <tonyyarusso> Wavesonics: www-data, and double-check that there are write permissions for the group.  (I think so, but haven't looked lately)
[07:10] <Adam_> <tonyyarusso> i added my user to that group: usermod -G www-data username
[07:10] <Adam_> but i still can't modify /var/www
[07:11] <tonyyarusso> what are the persmissions on /var/www?
[07:13] <Adam_> hhmm
[07:13] <Adam_> 1 moment, installing SVN :/
[07:14] <Adam_> drwxr-xr-x 3 root root
[07:16] <Weasel[DK]> is there a lilo guru in here ?    my server hangs at boot. probably after kernel upgrade. it says "LILO 22.8 Loading Linux" anda a lot of "." (dots), after 2lines it stops. Its Hardy with LVM
[07:16] <Weasel[DK]> can i enable debug output somewhere
[07:21] <soren> Weasel[DK]: Can you boot an older kernel?
[07:22] <Weasel[DK]> soren, nahh i tryed to edit the lilo.conf and set default kernel to "LinuxOLD", but the same result
[07:23] <soren> Weasel[DK]: How did you get to edit lilo.conf if you can't boot?
[07:23] <Weasel[DK]> soren, rescue boot from the CD... ;)
[07:23] <soren> Weasel[DK]: I see.
[07:23] <soren> Weasel[DK]: Could you try adding "large-memory" to your lilo.conf and try again?
[07:24] <soren> (of course you need to run lilo after changing the config)
[07:29] <Weasel[DK]> soren, just a line at the top ?
[07:29] <soren> Weasel[DK]: Something like that, yes.
[07:29] <soren> Weasel[DK]: As long as it's not in one of the kernel stanzas.
[07:30] <soren> And you're sure you're correctly rerunning lilo?
[07:31] <Weasel[DK]> soren, nahh i just noticed that inside lilo.conf there is a boot=/dev/sda
[07:31] <Weasel[DK]> i use lvm
[07:31] <soren> That's fine.
[07:31] <Weasel[DK]> ok
[07:31] <soren> boot=/dev/sda means that it'll install the bootloader on the mbr of /dev/sda
[07:32] <Weasel[DK]> soren, when runing lilo i get -> fatal device-mapper: mapped boot device cannot be on multiple real devices
[07:33] <soren> Weasel[DK]: And you didn't before you added large-memory?
[07:34] <Weasel[DK]> did't put in the memmory stuf yet... struggeling with vi... im used to vim ;)
[07:35] <soren> Ah :)
[07:35] <soren> So you're in a chroot? You have /proc mounted and all that?
[07:36] <soren> And /dev bind-mounted?
[07:38] <Weasel[DK]> i am booting from the install cdrom in rescue mode, runing mount shows that my LVM volume is the root
[07:39] <Weasel[DK]> had to be some kind of chroot env.
[07:39] <Weasel[DK]> has*
[07:40] <Weasel[DK]> soren, what do oyu mean by "And /dev bind-mounted"
[07:40] <soren> Weasel[DK]: Do you have e.g. /dev/mapper/control ?
[07:41] <Weasel[DK]> soren, nah i have /dev/mapper/root-lv_root which is the rootfs
[07:42] <soren> Ok, that's fine.
[07:42] <soren> I'm still a bit confused. You said you tried setting the default to LinuxOLD... So running lilo back then worked, but not now?
[07:42] <soren> Or did you forget to run lilo back then?
[07:44] <Weasel[DK]> soren, it allways where like that no matter what kernel i choose
[07:47] <soren> Weasel[DK]: What is root set to in your lilo.conf?
[07:47] <Weasel[DK]> opps
[07:47] <Weasel[DK]> soren, my notebook lost power :(  sorry
[07:48] <soren> No worries.
[07:48] <Weasel[DK]> im back now did i miss anything ?
[07:49] <soren> 07:47:34 < soren> Weasel[DK]: What is root set to in your lilo.conf?
[07:50] <Weasel[DK]> /dev/mapper/root-lv_root
[07:52] <soren> Weasel[DK]: Oh, that's not going to work.
[07:52] <soren> Weasel[DK]: I presume you have a /boot partition ?
[07:53] <Weasel[DK]> soren, no its LVM all way
[07:54] <soren> Has this ever worked?
[07:54] <Weasel[DK]> yeah sure ;)
[07:55] <soren> I didn't t hink the installer would let you do that.. :/
[07:55] <Weasel[DK]> until yesterday... i suspect the kernel update to be the reson but i dont know
[07:57] <Weasel[DK]> soren, well i also wondere why it was using LILO, the default is grub. maby LILO is better than grub for LVM ???
[07:58] <soren> grub will not work at all. lilo might have a chance.
[07:58] <soren> Do you have multiple disks?
[07:58] <soren> Is it possible that your root volume is spread over multiple pv's?
[07:59] <Weasel[DK]> yeah, 3 disks with 2 LVM partition om each 1 for root one for swap
[07:59] <soren> By "LVM partition" you mean physical volume?
[07:59] <Weasel[DK]> sda1 +sdb1 + sdc1 = root-lv
[07:59] <soren> Oh.
[08:00] <soren> That will *definitely* not work.
[08:00] <soren> :)
[08:00] <Weasel[DK]> it has worked ;)
[08:00] <soren> It might have worked at some point.
[08:00] <soren> If the kernel is now spread across multiple pv's, you're screwed.
[08:00] <soren> (or the initrd)
[08:01] <Weasel[DK]> ahh....
[08:01] <Weasel[DK]> crap
[08:01] <Weasel[DK]> so you are saying.... put an IDE drive in and put /boot and bootloader on that ?
[08:03] <soren> Either that or shrink one of your pv's on your /dev/sda and put it there.
[08:04] <soren> ...and next time do yourself a favour and make sure to have a boot partition.
[08:04] <Weasel[DK]> soren, ill do that for sure !
[08:06] <kraut> moin
[08:06] <Weasel[DK]> soren, ok, ill try to shrink sda1..
[08:07] <Weasel[DK]> soren, Thank you very much for the help :)
[08:14] <soren> Weasel[DK]: Oh, don't just shrink sda1.
[08:15] <soren> Weasel[DK]: You need to pvresize the data first, *then* shrink the partition.
[08:16] <Weasel[DK]> soren, yes ofcourse... orelse im in "real" trouble" ;)
[08:20] <soren> Weasel[DK]: Yeah. Just wanted to make sure you knew :)
[08:23]  * soren takes a break
[08:56] <mvo> does anyone has a idea what might cause "Setting up landscape-common (1.0.23-0ubuntu0.8.10.1) ...^M
[08:56] <mvo> Your account has expired; please contact your system administrator^M" (bug #291334)
[08:56] <uvirtbot> Launchpad bug 291334 in pam "package update-manager 1:0.93.32 failed to install/upgrade: ErrorMessage: SystemError in cache.commit(): E:Sub-process /usr/bin/dpkg returned an error code (1)" [High,Triaged] https://launchpad.net/bugs/291334
[08:56] <mvo> I have seen that in other reports as well
[08:56] <mvo> (not many though)
[08:57] <bugfixes> hello all
[09:04] <hads> mvo: Sounds like the root account has expired.
[09:06] <hads> I'm not sure of the correct way to do it but you can fix it by editing /etc/shadow
[09:12] <mvo> hads: thanks, I saw it in some bugreport while doing triage on udpate-manager and wonder what is the best way to avoid this kind of error and how it happend
[09:13] <hads> I'm not sure, I've seen it on VM's created with ubuntu-vm-builder
=== gaw is now known as gawin
[09:14]  * soren nods
[09:15] <soren> u-v-b for a while created locked root accounts.
[09:15] <soren> :(
[09:20] <mvo> oh, thanks. I will ask about that
[09:21] <soren> mvo: ...but it's a certainly a bug that a locked root account makes chfn fail, IMO.
[09:22] <mvo> ok, I reassign to pam
[09:24] <Weasel[DK]> soren, i reduced the filesystem on LVM then reduced with lvreduce and did a pvresize now fdisk still says the ther are no free sectors ?
[09:24] <Weasel[DK]> what did i forget ?
[09:25] <soren> Weasel[DK]: To resize the partition that holds the pv.
[09:48] <soren> mvo: Hm... It seems to happen in Hardy as well.
[09:49] <mvo> soren: with ubuntu-vm-builder? or in genral?
[09:50] <soren> mvo: I mean chfn failing if the root account is locked.
[09:50] <soren> I thought that was a side effect of the changes we made to the pam config in Intrepid.
[09:52]  * mvo nods
[10:13] <uvirtbot> New bug: #226369 in ubuntu "VFS Not responding error on shutdown (dup-of: 211631)" [Undecided,New] https://launchpad.net/bugs/226369
[10:15] <uvirtbot> New bug: #291186 in dnsmasq (universe) "disabling dnsmasq in /etc/default/dnsmaq disables init.d script" [Low,Confirmed] https://launchpad.net/bugs/291186
[10:31] <carbon_monoxide> any hardy backport repo for mono-runtime?
[10:31] <carbon_monoxide> i want newer version of mono-runtime
[10:40] <Wavesonics>  
[10:40] <Wavesonics> I think i somehow screwed up my account's permitions :( The main account I created during instal can no longer sudo, any ideas?
[10:49] <soren> Weasel[DK]: Any luck?
[10:53] <Weasel[DK]> soren, nahh the phone keeps ringing.... :(   ill let you know soon... i hope
[10:53] <soren> Weasel[DK]: :)
[11:12] <Weasel[DK]> soren, at last.... ;) do you have an advise for resizeing the physical partition on sda.... i have bootet an knoppix 5.1 with qtparted, but it does not seem to be up for the job
[11:18] <Weasel[DK]> soren, this parted does not know partitions types of  LVM ?
=== Abracadabra is now known as vendetta
=== vendetta is now known as Abracadabra
[11:57] <soren> Weasel[DK]: what I usually do is this:
[11:58] <soren> Say I have a pv of 100GB, and I want to downsize it to 80G. I start by resizing the pv to something well below 80G (say 75G).
[11:59] <soren> ...then I reduce the partition size to 80G. Depending on the tool at hand, this might involve removing the partition and creating it again. This feels scary, but make sure the starting cylinder is the same, and you'll be fine.
[11:59] <soren> Then I use pvresize again to grow the PV to the full size of the partition.
[11:59] <soren> Before that last step, I probably reboot, though.
[12:00] <soren> ...but that's pretty much the process I use. It's a bit more involved than it has to be, but I prefer to err on the side of caution.
[12:14] <Weasel[DK]> soren, ok, but i think i messed up, if i do a pvdisplay the sda1 i was working on is way too big. Data seems still to be ok.
[12:14] <Weasel[DK]> bigger than the entire disk
[12:16] <soren> Weasel[DK]: With pvdisplay? umm..  Do you still have the command line you used?
[12:17] <Weasel[DK]> soren, no but i should be able to set it at the same size at the other disks... the are identical
[12:18] <Weasel[DK]> well a bit smaller... verything should have been shrinked from the top
[12:18] <Weasel[DK]> *everything
[12:23] <soren> Weasel[DK]: Yeah, i'm jus curious how it happened.
=== Abracadabra is now known as Vendetta
[12:27] <Weasel[DK]> soren, i guess i typed the size of the whole lvm group instead of the disk. It matches the size. on the other hand that would be clumsy to do so ?
[12:30] <willwill> hello, I installed suphp but phpmyadmin does not work. I then enable mod_php5 (suphp still enabled) and phpmyadmin worked but my script does not run as my user
[12:31] <willwill> I tried to comment all lines in /etc/apache2/mods-enabled/php5.conf and it still assign itself to .php
=== Vendetta is now known as Abracadabra
[12:42] <stewils> hi.  i'm mid way through an ubuntu update.  ssh'd into the sever to do it and lost connecgtion mis way.  any way to get the session back
[12:42] <soren> No.
[12:43] <stewils> :(
[12:43] <stewils> bugger
[12:43] <stewils> so start the install process again?
[12:43] <henkjan> next time, start your upgrade in screen
[12:43] <stewils> yeah :)....live and learn
[12:43] <soren> Yes. It might pick up around where it left off, though.
[12:44] <stewils> ok
[12:44] <stewils> i'lll get off and do that then.  thanks
[13:39] <Frozen-Solid> thanks i thought it might be quieter in here - it's really hard to kepe an eye on the other chat while at work with distrcations
[13:39] <Frozen-Solid> okay so this alternate cd thing... i'm assuming that's just download latest version, burn iso, and mount like normal?
[13:39] <Frozen-Solid> i've been on an rpm-based distro at home for years, this is my first time dealing with apt
[13:40] <ScottK> Frozen-Solid: Are you installing a server?
[13:40] <Frozen-Solid> yeah
[13:40] <Frozen-Solid> running on a blade center
[13:40] <Frozen-Solid> ubuntu se blows my mind in how awesome it is on our new environments
[13:41] <ScottK> Generally you boot the CD on the machine you intend to install on.
[13:41] <Frozen-Solid> we gt the install done... i'm looking into how upgrading works
[13:41] <Frozen-Solid> we built our new systems on 8.04... and then 8.10 came out so i was charged with finding out hte best ways to upgrade with least amount of downtime and least invasive
[13:42] <ScottK> Oh.
[13:42] <ScottK> You don't need a CD for htat.
[13:42] <komotheit> What I usually do, is set the Alternate CD as a repository
[13:42] <ScottK> sudo do-release-upgrade should fire it off
[13:42] <komotheit> And do the apt-get distro upgrade
[13:42] <Frozen-Solid> komotheit was talking about using an alternate cd, so that's what i was asking about
[13:43] <mvo> make sure you edit /etc/update-manager/release-upgrades is set to "normal" - because hardy is a lts it will not upgrade to a non-lts release by default
[13:43] <mvo> upgrades via the alternate cd are possible as well, but the upgrade should work fine over the net
[13:43] <Frozen-Solid> is it generally recommended to stick to lts editions, rather than the fall reelases?
[13:43] <mvo> depends on your use-case :)
[13:44] <mvo> if you want the long support then lts is certainly a good choice, if you want (or need) the latest stuff then upgrading is probably a good idea
[13:44] <komotheit> Well, unless you need something only offered in a new version, its best to stick to LTS releases
[13:44] <Frozen-Solid> we're not sure our use case yet... we're migrating from an ancient novell netware 6.5 web server setup
[13:44] <ScottK> Frozen-Solid: ^^^ wrote the upgrade tools, so you are getting advice directly from the best expert.
[13:45] <Frozen-Solid> oh cool
[13:45] <komotheit> I didn't upgrade yet, I'm snooping around tech support chat/forums to get an idea of issues from upgrades
[13:46] <Frozen-Solid> we're not doing a support contract yet, but it's something we may look into in the future... depends on how well we manage this new setup ourselves. I was just getting a feel for how upgrades should be looked at, because noene of us have done linux in such a live environment before
[13:48] <Frozen-Solid> so way for april. apt-get distro upgrade should be safe - does ubuntu require a full restart after an upgrade like that because of the new kernel?
[13:48] <Frozen-Solid> s/way/wait
[13:49] <ScottK> Frozen-Solid: For going from release to release, do-release-upgrade is recommended because it often has special processing to deal with known issue.
[13:49] <komotheit> Since you're running a server, I wouldn't know. I'm running a desktop environment, I probably would need a restart unless I would shut down X-server before upgrading
[13:49] <ScottK> Frozen-Solid: You do have to reboot.
[13:49] <ScottK> Because of the new kernel.
[13:49] <Frozen-Solid> okay thought so
[13:51] <Frozen-Solid> i need to remember do-release-upgrade. *makes note*
[13:52] <Frozen-Solid> while this hsouldn't happen, say network goes down... ubuntu's server's go down... something odd happens during an upgrade... what kinda safty procautions do i need to make?
[13:52] <ScottK> Frozen-Solid: You should always back up important data.
[13:52] <komotheit> Was going to say that too :P
[13:53] <Frozen-Solid> yeah i know that much, i meant specifically from a ubuntu se perspective
[13:53] <ScottK> If you lose power on a half upgraded system, it's possible that it could be in an odd state, but generally would be bootable/fixable.
[13:53] <Frozen-Solid> like, if the network goes down, can i just rerun do-release-upgrade again and it can recover? or what happens
[13:53] <ScottK> Should work.
[13:53] <ScottK> Just don't reboot a half-upgraded system.
[13:53] <Frozen-Solid> losing power would be pretty much impossible on this server
[13:54] <ScottK> Generally it's pretty straight fowared.
[13:54] <Frozen-Solid> cool
[13:55] <Frozen-Solid> i tend to do much more invasive upgrades on my home systems... full format cept for the /home partition, reinstall from scratch
[13:55] <Frozen-Solid> obviously i dont' want to do that here
[13:55] <ScottK> There's generally no need for that.
[13:56] <Frozen-Solid> i use Mandriva at home, it's not quite as "neat" of an upgrade as ubuntu seems to have
[13:56] <ScottK> I've only ever had to reinstall after doing experiments I knew were risky.
[13:56] <Frozen-Solid> and it feels nice starting fresh
[13:56] <ScottK> Upgradability is one of the key features of Debian based systems.
[13:56] <Frozen-Solid> yeah that's why we chose Ubuntu SE. so far i'm VERY impressed from a server perspective
[13:57] <Frozen-Solid> it's "easy" but it still expects you know what you're doing and doesn't hold your hand. i really appreciate how it's setup
[13:57] <Koon> mathiaz: I have been considering revisiting https://wiki.ubuntu.com/ImproveSSLCert for Jaunty... what is your position on that spec ? Something you want to do yourself, something you want someone to do...
[13:57] <Fenix|work> Good morning
[13:57]  * ScottK points Frozen-Solid to http://ubuntukids.org/blog/?p=117
[13:57] <Fenix|work> I have a question that I hope someone has an answer to
[13:57] <Weasel[DK]> soren, i changed my angle of attack a bit... i now removed one of my swap partitions, re-formated it for ext3 and copied the old /boot into the partition.
[13:58] <mathiaz> Koon: I don't really have strong opinion on doing it myself.
[13:58] <Frozen-Solid> haven't used ubuntu desktop yet... i'm kinda aattached to my mandrake i've used for 10 years
[13:58] <mathiaz> Koon: if you have more ideas, please add them to the spec.
[13:58] <soren> Weasel[DK]: Heh. That should certainly work, too.
[13:58] <Weasel[DK]> edited fstab, now i have trouble with lilo again
[13:58] <soren> Weasel[DK]: You need to adjust lilo.conf accordingly.
[13:58] <Koon> mathiaz: ok :)
[13:58] <Fenix|work> I'd like to make a bootable DVD of my install so if a HDD or two decide to screw the pooch, I can stick in this bootable DVD and restore my system to a working state, then restore the most up-to-date stuff from backup
[13:59] <Fenix|work> anyone with any ideas?
[13:59] <ScottK> Frozen-Solid: Right, well I've never actually used Ubuntu on the desktop for more than a minute or two myself.  Kubuntu (KDE) is my thing.
[13:59] <Frozen-Solid> lol excellent post scott
[14:00] <Frozen-Solid> I'm a big Gnome fan... but the fact that ubuntu seems to pick one distro to base it's installs around (kubuntu vs ubuntu) rather than give you abit of everything like i'm used to with mandriva keeps me off ubuntu/kubuntu
[14:01] <Weasel[DK]> soren, can you tell me how to set those options... i tried diffent things bu i always get : Fatal: Trying to map files from unnamed device 0x0010 (NFS/RAID mirror down ?)
[14:01] <Frozen-Solid> i use gnome, but a lot of KDE software on it (K3B, Amarok, Kaffeine) as opposed to the gnome counter-parts
[14:01] <ScottK> Frozen-Solid: They are both in the same repository, so you can certainly install whatever KDEish bits you want.
[14:01] <komotheit> That's what I did
[14:01] <Frozen-Solid> yeah... mandriva just does it all though, so it's hard for me to go for a switch
[14:01] <komotheit> I'm a fan of klipper :)
[14:01] <ScottK> Myself I like having a common core system so I only need to know one.
[14:01] <Frozen-Solid> i'm also attached to my root user and it feels rong not being able to log into my good ole root on ubuntu
[14:01] <soren> Weasel[DK]: Let me think..
[14:02] <ScottK> That's easy enough to enable if you want it.
[14:02] <Frozen-Solid> s/rong/wrong
[14:02] <Frozen-Solid> i'm sure i'll swithc and try ubuntu soonish. mandriva's starting to show it's bloat
[14:02] <Frozen-Solid> i need to look into ubuntu install on a Novell Netware network, i haven't checked if the novell client is on the apt-get repos yet
[14:03] <Frozen-Solid> thne i'll totally switch for work purposes
[14:03] <soren> Weasel[DK]: What is boot and root set to?
[14:04] <Frozen-Solid> thanks for all your help, i gotta meeting to go to so.
[14:05] <Weasel[DK]> soren, boot=/dev/sda      boot is only in the append line and says /dev/mapper/root-lv_root
[14:05] <Weasel[DK]> the new boot is on sda2
[14:06] <Weasel[DK]> soren, right now i have booted knoppix, mounted the rootfs and mount the new boot into that
[14:07] <komotheit> Frozen-Solid: Take care and good luck on your upgrades! :)
[14:09] <soren> Weasel[DK]: I wish I had a box with a similar setup so that I could be more sure..
[14:09] <soren> Weasel[DK]: I keep forgetting how much stuff lilo works out on its own.
[14:10] <soren> Weasel[DK]: If you run lilo now, what happens?
[14:10] <Weasel[DK]> soren, i run lilo -C /mnt/root/etc/lilo.conf
[14:11] <Weasel[DK]> and it responds with     Fatal: Trying to map files from unnamed device 0x0010 (NFS/RAID mirror down ?)
[14:11] <soren> Weasel[DK]: oh, do this:
[14:11] <soren> Weasel[DK]: lilo -r /mnt/root
[14:12] <Weasel[DK]> soren, Fatal: raid_setup: stat("dev/sda")
[14:14] <soren> Weasel[DK]: sudo mount --bind /dev /mnt/root/dev
[14:14] <soren> Weasel[DK]: And try again.
[14:16] <Weasel[DK]> soren, i get some complaints but at the end i says Added Linux and LinuxOLD
[14:17] <Weasel[DK]> should i give it a try
[14:17] <soren> Go for it :)
[14:20] <Weasel[DK]> soren, IT IS BOOTING.... !
[14:20] <Weasel[DK]> soren, you are my hero
[14:21] <soren> Weasel[DK]: I accept beer deliveries. http://www.øltorvet.dk/ for your clicking pleasure :)
[14:21] <soren> Weasel[DK]: Make sure you check that lilo runs as expected in the running system.
[14:29] <Weasel[DK]> soren, seems to be ok
[14:29] <soren> Fantastic.
=== Abracadabra is now known as Vendetta
[14:41] <Technoviking> hello
[14:42] <Technoviking> I'm using 12 GB of ram on a 32bit version of Ubuntu 8.04.1 Ubuntu Server
=== Vendetta is now known as Abracadabra
[14:43] <Technoviking> Someone is worried that even with PAE you can't assign 9-10GB to one process, because that
[14:43] <Technoviking> "overflows size_t"
[14:44] <soren> Technoviking: Well..
[14:45] <soren> Technoviking: The explanation is not entirely accurate, but the problem is real.
[14:45] <soren> size_t is mandated by the hardware, and it happens to be 32 bits, which limits you to 4 GB of addressable memory.
[14:47] <Technoviking> so even with PAE, you can not use more than 4 GB for a process/app PostgreSQL for example
[14:47] <soren> Technoviking: That's correct.
=== jkerihuel_ is now known as kerihuel
[14:50] <Technoviking> soren: thank for clearing that up for me
[14:51] <Weasel[DK]> soren, PM me your delivery address or do you live near by ?
[15:06]  * Weasel[DK] promised his family a trip to Bilka... so c u later.
[15:13] <soren> Weasel[DK]: Other end of the country, unless I'm much mistaken. Aalborg, specifically.
[16:08] <CRELOUR>  hi
[16:21] <jtmoney> hello, guys... 2 x 1 TB drives just got here, and i'm reading up on setting up RAID-1 in ubuntu server... can someone tell me how i can determine if dmraid now supports error handling? the fakeRAIDhowto isn't very clear... https://help.ubuntu.com/community/FakeRaidHowto#RAID-1%20Notes
[16:24] <Rubin> jtmoney, i would guess that it is not in there
[16:25] <jtmoney> heh okay
[16:25] <jtmoney> but how can i find out definitively?
[16:30] <Rubin> jtmoney, i'd try to find a site for the author of the patch that enables it, and see if they have any kind of info about it
[16:30] <Rubin> or you could go through the kernel package and try to figure out what patches they have applied etc
[16:34] <lukehasnoname> I see Ubuntu Server has an ad running on the Ubuntu home page
[16:34] <lukehasnoname> props to the person who got that ball rolling
[16:35] <Rubin> wasnt there something on slashdot yesterday about ubuntu loosing money but if server takes off for a couple years they will be in the black?
[16:36] <lukehasnoname> I wouldn't doubt CANONICAL is losing money right now, but they have a lot of it, so their demise is not imminent. A takeoff of their server OS and the support that businesses would buy for it would really help the company, though.
[16:38] <Rubin> thats what the article said, almost exactly.
[16:38] <Fenix|work> can I resize an ext3 partition on the fly?
[16:38] <ScottK> IIRC Canonical has not made a secret of being in the red nor not being particularly worried about it.
[16:38] <Fenix|work> Greetings!
[16:39] <Rubin> i think he said he'd continue to fund it for 5 years
[16:40] <Rubin> Fenix|work, no
[16:40] <Rubin> Fenix|work, that kind of thing is reserved for 'real' filesystems, like advfs :P
[16:40] <Fenix|work> How long does an offline resize take?
[16:40] <Rubin> if its not heavely fragmented i'd guess it doesn't take very long
[16:41] <Rubin> course depends on your hardware
[16:41] <Rubin> are you making it bigger or smaller?
[16:41] <lukehasnoname> 5 years for ubuntu losing money before MarkS turns off the tap? It sounds like a long time, but it will be here before we know it. I'm sure profitability is in their sights.
[16:41] <Fenix|work> I have a 458GB partition that I'd like to lop off 58GB and make another partition
[16:41] <Rubin> lukehasnoname, 13.04 could be the last version :P
[16:41] <Rubin> Fenix|work, that could take quite a while
[16:42] <lukehasnoname> Rubin: heh
[16:42] <Rubin> i did that a while ago to break /home into its own, but i didnt time it or anything
[16:42] <Fenix|work> Rubin, it's a production web/ftp server and the previous admin failed to make web and ftp separate
[16:42] <Rubin> completed it in an evening if i recall
[16:42] <Rubin> Fenix|work, clone it, do the change on the other one, swap the IP over
[16:43] <Rubin> or add another hard disk :P
[16:43] <Fenix|work> for that matter, he failed to move web and ftp out of the system partition
[16:43] <Fenix|work> so everything is lumped into /dev/sda1
[16:43] <Rubin> yeah lame.
[16:43] <Rubin> hard disks are cheap though. juts add another
[16:43] <Fenix|work> hard disks may be cheap, but servers aren' t:)
[16:44] <Fenix|work> I don't have a spare to swap to
[16:44] <Rubin> psh. real production setups have hot spares or failover pairs :P
[16:44] <Fenix|work> Rubin, real servers use hardware raid and the OS can't see hot spares :)
[16:44] <Rubin> no i mean whole spare systems
[16:45] <lukehasnoname> BTW I just read about the installation-time server profile of "Virtualization Host"
[16:45] <lukehasnoname> totally awesome
[16:45] <Fenix|work> ooh, well in that case I don't have a real production server :)
[16:45] <Rubin> Fenix|work, then they can handle some down time :P
[16:46] <Fenix|work> corporate website may not handle downtime well with marketing and execs
[16:46] <lukehasnoname> five nines
[16:47] <Fenix|work> how does one check the fragmentation of a partition?
[16:47] <Rubin> Fenix|work, then they should spring for a real setup with failover and/or hot spare servers :P
[16:47] <Rubin> uptime costs money
[16:47] <Fenix|work> yes, yes it does
[16:48] <Fenix|work> but that still doesn't solve my more immediate problem :)
[16:48] <lukehasnoname> Rubin: What's the problem with software raid? For a SMB server, wouldn't software raid do alright?
=== lamont` is now known as lamont
[16:48] <Rubin> lukehasnoname, there are several problems with it, but it does alright yes.
[16:49] <Rubin> it doesnt handle errors very well, as you noticed. its slower than hardware raid, etc
[16:49] <Rubin> but if a disk full out fails, your golden after a reboot (and maybe unplugging it worst case)
[16:49] <Rubin> i run a mirror lvm setup on my home server
[16:56] <lukehasnoname> gentoo4life, lol
[17:05] <r00tintheb0x> Hello all, if I have 6x147GB SCSI drives in a RAID6 array... what would be my total available space?
[17:06] <r00tintheb0x> Or could someone give me a rule of thumb to calculate RAID space in general?
[17:11] <r00tintheb0x> Sorry, my client crashed.
[17:11] <r00tintheb0x> If anyone responded to my question, could you please re-respond.
[17:11] <r00tintheb0x> :)
[17:49] <jtmoney> hey guys, recently purchased 2 x 1 TB drives for a file server... it seems like fakeraid (through dmraid) is in its infancy and has poor error reporting... i'm too paranoid that i would screw something up if i used softraid... would it make sense at this point just to use rdiff or something similar to copy files from one partition to another?
[17:49] <jtmoney> seems like there's no good answer
[18:02]  * antdedyet dwindles in the fun that is email delivery
[18:30] <trashguy> i has 8.10
[18:50] <r00tintheb0x> Hello all, if I have 6x147GB SCSI drives in a RAID6 array... what would be my total available space?
[18:50] <r00tintheb0x>  Or could someone give me a rule of thumb to calculate RAID space in general?
[18:54] <Deeps> raid6 - you 'lose' 2 disks worth of space for parity data
[18:54] <Deeps> raid5 - you 'lose' 1 disks worth of space for parity data
[18:56] <kirkland> ScottK: hi, i have a couple of dumb mail server questions ...
[18:56] <ScottK> OK
[18:57] <maswan> raid10, you lose half of the space for parity data
[18:57] <Deeps> r00tintheb0x: you can potentially 'lose' more disks if you factor in that it's sometimes smart to have a 'hot spare' as well
[18:57] <kirkland> ScottK: i simply want an ubuntu system to accept incoming mail, and if there's a local user matching the address, use his .forward file
[18:58] <kirkland> ScottK: so machine, "foo", dustin@foo address, with a .forward file pointing to gmail or some such
[18:58] <ScottK> OK, so no actual local delivery?
[18:58] <kirkland> ScottK: i've got mail to dustin@localhost actually working
[18:58] <ScottK> OK.
[18:58] <kirkland> ScottK: right, no local delivery
[18:59] <kirkland> i think i just need some magic to accept traffic from outside, perhaps?
[18:59] <ScottK> Do you need to use actual .forward files or is that just the effect you want?
[19:00] <ScottK> kirkland: I assume you're using Postfix?  Please pastebin the output of postconf -n
[19:00] <kirkland> ScottK: yes, postfix
[19:01] <kirkland> ScottK: .forward is the effect i want, if it can be configured differently, that's cool too
[19:01] <ScottK> I do this with virtual alias domains.  It's pretty easy.
[19:02] <kirkland> cool, let me pastebin that for you
[19:03] <kirkland> ScottK: http://pastebin.ubuntu.com/65346/
[19:03]  * ScottK looks
[19:04] <ScottK> kirkland: foobar.com is your actual domain you want mail from or the machine's hostname?
[19:05] <kirkland> ScottK: i just sed-replaced that
[19:05] <ScottK> Yeah.  I figured.
[19:05] <ScottK> For virtual alias domains, here's what I have ...
[19:06] <ScottK> virtual_alias_domains = example.com (note that it should not be listed in mydestination.
[19:06] <ScottK> Then I have virtual_alias_maps = hash:/etc/postfix/virtual
[19:06] <ScottK> This is all in main.cf BTW
[19:06] <kirkland> right
[19:07] <ScottK> That map has entries like example.com Whatever (need that, it doean't matter what's on the right side)
[19:07] <ScottK> scott@example.com scott@example.org
[19:07] <ScottK> etc.
[19:08] <ScottK> Then in smtpd_recipient_restrictions I have check_recipient_access hash:/etc/postfix/somename
[19:09] <ScottK> somename has postmaster@example.com permit
[19:09] <kirkland> gotcha
[19:09] <ScottK> That can be permit, reject, or dunno
[19:09] <ScottK> So the virtual table says where it goes.
[19:09] <kirkland> right
[19:09] <ScottK> the check_recipient_access does your recipient validation.
[19:10] <kirkland> let me give that a shot
[19:10] <ScottK> Don't forget to postmap or postalias your table.
[19:10] <ScottK> s
[19:10] <kirkland> so no dovecot required at all for this, then?
[19:10] <ScottK> Nope.
[19:10] <kirkland> good
[19:10] <ScottK> You do this all in the MTA.
[19:10] <kirkland> right, i was hoping so
[19:11]  * ScottK needs to run out.  Back later if you have questions.
[19:11] <ScottK> There's a good REAME on virtual stuff on postfix.org too.
[19:11] <ScottK> That's pretty much straight out of it.
[19:14] <kirkland> ScottK: awesome, thanks for your help
[19:14] <ScottK> You're welcome.
[19:18] <ShawnR> for a single user (PC sitting in my house) is there any real benefit to using a non-root login? (it seems like a pain to have to sudo everything)
[19:19] <trashguy> i su in to root
[19:20] <trashguy> im used to freebsd so it seems sane to me
[19:20] <ScottK> trashguy: If you want to do that, use sudo -i.  It gives you a slightly more sane environment.
[19:21] <ShawnR> on my debian install, i just logged in as root all the time, i know for multi-admin users, it's better to not have one ROOT that everyone/anyone uses, but for a home/personal use, i didn't know if there were any real security flaws
[19:21] <trashguy> ScottK, root is sane for me
[19:21] <kirkland> ShawnR: i find it keeps me from stepping on my own ... feet
[19:21] <ScottK> trashguy: sudo -i gets you a root prompt.
[19:21] <trashguy> ShawnR, there are program like irc i wouldnt run as root
[19:22] <ShawnR> yeah, that is true
[19:22] <ShawnR> but for administration
[19:23] <ShawnR> i'll run many programs as not root, but for configuring everything... i mean, i couldn't even make /downloads without having to sudo... then i had to sudo chmod it to allow myself write access
[19:23] <trashguy> these are good things
[19:24] <ShawnR> oh?
[19:24] <trashguy> well
[19:24] <trashguy> im all for having to do extra steps to promote secure methods ^^
[19:25] <ShawnR> i would be, too... but is it really any more secure than me having a password on the root account?
[19:25] <trashguy> if you need to do a bunch of admin stuff you can sudo -i or su
[19:25] <ShawnR> i won't be running services as root, just configuring things as root
[19:25] <ShawnR> hmm
[19:25] <trashguy> then use sudo -i
[19:25] <ShawnR> maybe i'll try and get in the habit of doing that
[19:25] <ShawnR> i guess part of me just loves the idea of logging in as "root"
[19:25] <trashguy> it just t hrows you in a root prompt
[19:26] <ShawnR> root@WOPR:~# just looks so sexy
[19:26] <trashguy> i only log in as root on most machines if im local
[19:26] <trashguy> avonders@avonders-desktop:~$ sudo -i
[19:26] <trashguy> [sudo] password for avonders:
[19:26] <trashguy> root@avonders-desktop:~#
[19:26] <ShawnR> 9/10 times i will be local
[19:26] <ScottK> ShawnR: Up until you rm -rf * and aren't in the dir you thought you were, sure.
[19:26] <trashguy> or umount var
[19:26] <trashguy> ^^
[19:26] <uvirtbot> trashguy: Error: "^" is not a valid command.
[19:26] <ShawnR> that'll be my own problem... but i don't do those things
[19:27] <ScottK> Of course.
[19:27] <ShawnR> this is just going to be a personal file/media streaming server with LAMP things just for when it's convenient
[19:27] <trashguy> smash buttons
[19:27] <trashguy> RAWR
[19:27] <ScottK> ShawnR: If it's exposed to the internet be sure to turn off SSH root login.
[19:28] <ShawnR> yeah
[20:20] <CarlFK> how can I turn ono the cpu fan?
[20:22] <ShawnR> put a hairdryer on it, that should trigger the thermal sensors
[20:23] <ShawnR> it's not on currently?
[21:18] <lukehasnoname> the server guide needs to be moved from doc.ubuntu.com to help.ubuntu.com as the 8.10 guide
[21:28] <Appl3Kork> so I've got Ubuntu Server installed, and I got it all setup in webmin, but I can't get ppl outside the home network view it?
[21:30] <ScottK> Appl3Kork: Webmin isn't in the Ubuntu repositories, so it's not supported here.  Ask the webmin people.
[21:31] <Appl3Kork> well I'm not using webmin to configure it. sorry.  I'm trying to get the ddclient working through putty
[21:32] <ShawnR> appl3kork ppl from outside can't view what? can't access webmin or what services?
[21:33] <Appl3Kork> can't access my server
[21:33] <Appl3Kork> it's running apache
[21:34] <ShawnR> and you cna view it locally, right?
[21:34] <Appl3Kork> yes
[21:34] <Appl3Kork> I've forwarded ports
[21:34] <ShawnR> can you access any other service from outside?
[21:34] <Appl3Kork> no
[21:34] <Appl3Kork> nothing
[21:35] <ShawnR> what error do ppl get?
[21:35] <Appl3Kork> well I just give them the url, and it just says cannot be found
[21:35] <ShawnR> what url?
[21:35] <Appl3Kork> g3rm.dontexist.com
[21:36] <Appl3Kork> I'm just trying to do the basics right now, of them seeing it say "It Works!"
[21:36] <ShawnR> on port 80, right?
[21:36] <Appl3Kork> I think port 80 is blocked by my ISP, so my buddy said to use 81
[21:37] <ShawnR> and you forwarded ports 81?
[21:37] <Appl3Kork> yes
[21:37] <Appl3Kork> both router and modem
[21:37] <ShawnR> and when you access it locally, you use http://g3rm.dontexist.com:81
[21:37] <ShawnR> right?
[21:37] <Appl3Kork> no actually
[21:37] <ShawnR> ....
[21:37] <Appl3Kork> I just typed in g3rm.dontexist.com
[21:37] <ShawnR> http:// will default to port 80 unless otherwise specified
[21:37] <ShawnR> so you're still running on port 80
[21:37] <Appl3Kork> o
[21:38] <Appl3Kork> k let me change that really quick
[21:38] <ShawnR> have you restarted apache?
[21:38] <TANATHOS> either port 80 nor port 81 works
[21:38] <ShawnR> 80 works locally, and 80 is blocked by ISP
[21:39] <Appl3Kork> well I don't think on dyndns.org will allow me to make the site say g3rm.dontexist.com:81
[21:39] <TANATHOS> what kind of isp block port 80?
[21:39] <Appl3Kork> gay ones
[21:39] <ShawnR> comcast used to
[21:39] <ShawnR> some still do, they don't want you using your home ISP connection as a server
[21:39] <ShawnR> no, dyndns will not
[21:39] <TANATHOS> I don't really think you can forward
[21:39] <ShawnR> you just gotta tell ppl to use the :81 at the end
[21:39] <TANATHOS> I was about  to write that
[21:40] <TANATHOS> not necesarily
[21:40] <Appl3Kork> whats the command to restart apache?
[21:40] <TANATHOS> ubuntu?
[21:40] <Appl3Kork> ya
[21:40] <TANATHOS> ./etc/init.d/apache restart
[21:40] <TANATHOS> as I remember it
[21:40] <ShawnR> tanathos what would be the alternative to telling ppl to include port #?
[21:41] <TANATHOS> make a redirect from another page
[21:41] <ShawnR> lol
[21:41] <ShawnR> true
[21:41] <TANATHOS> ?
[21:41] <TANATHOS> oh I thought you were laughing at me
[21:41] <ShawnR> but that'd be a lot of work
[21:41] <ShawnR> no, just at the backbending
[21:41] <Appl3Kork> well in webmin it actually asks what port for the server
[21:41] <TANATHOS> he could make a redirect from another page
[21:41] <Appl3Kork> and I put 81
[21:41] <TANATHOS> Appl3Kork yes
[21:41] <TANATHOS> but that is locally
[21:42] <Appl3Kork> o
[21:42] <TANATHOS> you need to redirect from outside your isp
[21:42] <ShawnR> yeah, but if he has a hosted page somewhere, why not just put the website there?
[21:42] <TANATHOS> true
[21:42] <Appl3Kork> i didn't think I had a hosted page
[21:42] <Appl3Kork> or is that the g3rm.dontexist.com
[21:42] <TANATHOS> wait
[21:42] <ShawnR> by hosted i mean on someone else's computer... like at a website hosting company
[21:42] <TANATHOS> do you have another hosting account somewhere
[21:43] <Appl3Kork> no
[21:43] <TANATHOS> even better
[21:43] <Appl3Kork> not that I know of
[21:43] <TANATHOS> do you own dontexist.com
[21:43] <Appl3Kork> no it was just a free one with dyndns.org
[21:43] <ShawnR> nope, that's a dyndns.org name
[21:43] <ShawnR> all they do is redirect the subdomain to his IP at home
[21:43] <TANATHOS> I could give you a hosting package at my comapny
[21:43] <ShawnR> well, this looks like a match made in IRC heaven
[21:43] <Appl3Kork> ya but how much is that?
[21:44] <ShawnR> lol
[21:44] <TANATHOS> but still you have to own a domain
[21:44] <TANATHOS> Appl3Kork for free
[21:44] <TANATHOS> I own the company
[21:44] <Appl3Kork> so if I own a domain you can host it for free?
[21:44] <TANATHOS> I got about 16 working servers
[21:44] <ShawnR> i want a free hosting package!
[21:44] <TANATHOS> :))
[21:44] <TANATHOS> hold your horses
[21:44] <ShawnR> i'm just playin
[21:44] <TANATHOS> Appl3Kork is in trouble not you
[21:45] <ShawnR> i've got one hosted now, i'd hate to keep switching... unless you can get me exchange for free :P (which i highly doubt you'd do for me)
[21:45] <TANATHOS> ShawnR how much do you pay for that package
[21:45] <TANATHOS> I'll make it half:))
[21:46] <ShawnR> well, i don't have exchange right now
[21:46] <TANATHOS> and try not to lie
[21:46] <TANATHOS> :))
[21:46] <ShawnR> it's ike USD $80/yr
[21:46] <ShawnR> i think 84 to be exact
[21:46] <TANATHOS> when does it expire?
[21:46] <ShawnR> i just paid for another year... heh
[21:47] <ShawnR> but all i want is shell access to do file management and backup and i'd REALLY like exchange
[21:47] <TANATHOS> well your los
[21:47] <TANATHOS> loss
[21:47] <TANATHOS> exchange server?
[21:47] <ShawnR> yup
[21:47] <TANATHOS> Micro shit?
[21:47] <TANATHOS> I only have 2 of those
[21:47] <ShawnR> exchange being that i could sync up my phone contacts with outlook and not have to worry about when i get a new phone
[21:48] <TANATHOS> sorry I can't give you on MES
[21:48] <ShawnR> but it's not worth me paying like $100/yr more or whatever it is, just to have that feature
[21:49] <ShawnR> it's ok
[21:49] <TANATHOS> are rented
[21:49] <TANATHOS> and can't really mess them up
[21:49] <Appl3Kork> i think I own a domain name
[21:49] <Appl3Kork> had one before
[21:49] <Appl3Kork> lemme check my email really quick
[21:49] <ShawnR> well, i'm headed out now... good luck appl3
[21:49] <Appl3Kork> thanx
[21:51] <Appl3Kork> so TANATHOS if I've got a domain name and you can host it, does that make it so people can see my server?
[21:52] <TANATHOS> yes
[21:52] <Appl3Kork> ok
[21:52] <Appl3Kork> im just checking to see where my domain name is
[21:53] <TANATHOS> ok
[22:04] <TANATHOS> Appl3Kork?
[22:04] <Appl3Kork> yup
[22:05] <TANATHOS> domain?
[22:06] <Appl3Kork> i was just going to get a new one cause it was 1.99
[22:06] <Appl3Kork> just trying to think what it should be called
[22:06] <TANATHOS> o lol
[22:06] <TANATHOS> only 1.99
[22:06] <TANATHOS> well then get me tanathos.org
[22:06] <TANATHOS> :))
[22:07] <Appl3Kork> does it matter if it's .com .net and whatnot?
[22:08] <Appl3Kork> for your hosting
[22:08] <TANATHOS> no
[22:28] <mindframe> so my update to ibex failed during the download and now update-manager -d wont show it as an option.  how can i upgrade my system?
[22:32] <mvo> mindframe: if that happend during the download (and not during the install then its odd) - should work
[22:32] <Appl3Kork> but TANATHOS when I have a domain and a host, how does it all setup with the server?
[22:32] <mvo> mindframe: could you put your /var/log/dist-upgrade/main.log up somewhere please?
[22:36] <mindframe> damnit pastebin wont let me paste.. saying i trip the spam filter
[22:36] <hads> paste.pocoo.org
[22:37] <lukehasnoname> http://mibbit.com/pb/9mtFY2
[22:38] <mindframe-> http://paste.pocoo.org/show/89757/
[23:11] <LordDicranius> is there an easy way to move cron jobs from one system to another?
[23:13] <mindframe> copy / paste?