[00:02] <CppIsWeird> after installing ubuntu server the computer boots and just hangs. it wont go to the next boot device and it doesnt display anything from booting linux. it just stops.
[00:50] <CppIsWeird> after installing ubuntu server the computer boots and just hangs. it wont go to the next boot device and it doesnt display anything from booting linux. it just stops.
[00:53] <sdfoj8> how i can add new vps to my ubuntu server ? on 10.4 ?
[01:38] <pwnguin> ok this has been bugging the hell out of me for some time
[01:39] <pwnguin> how the hell do i turn off gnome terminal transparency?
[01:40] <CppIsWeird> pwnguin, iirc right click on the terminal, and click on edit options. but im not on a ubuntu desktop atm.
[01:40] <pwnguin> profile is set to black
[01:41] <pwnguin> solid color
[01:45] <CppIsWeird> there is no slider bar under the colors for adjusting transparency
[01:45] <CppIsWeird> ?\
[01:51] <pwnguin> its just rgb
[01:51] <pwnguin> alpha is done seperate
[01:55] <prestonc> anyone here run Ubuntu Server virtually? I'm running it under vmware-fusion and having trouble with the arrow keys.
[01:56] <CppIsWeird> prestonc, i run it in virtual box fine.
[01:57] <CppIsWeird> never tried vmware fusion
[01:57] <prestonc> vm
[01:57] <prestonc> oops
[01:57] <prestonc> hmmm, rather
[01:58] <prestonc> never had to mess with key mappings in virtual box to get the arrow keys to work? Basically my ubuntu server install isn't responding to arrow keys in bash for history. I assume this isn't a setting of server, but rather a vm issue.
[02:05] <CppIsWeird> i could not imagine fixing 600,000 bugs
[02:07] <pwnguin> dont worry
[02:07] <pwnguin> we mark a ton invalid
[02:07] <pwnguin> and wontfix another ton
[02:08] <pwnguin> another ton is fixed upstream, and makes its way to debian and ubuntu by sync, having never addressed the problem
[02:08] <pwnguin> yet another ton is just duplicates
[02:20] <CppIsWeird> any reason why the ubuntu server i just installed have internet access during installation but doesnt now?
[02:29] <unewbie> what is jeos?
[02:29] <unewbie> !jeos
[02:29] <CppIsWeird> just enough operating system
[02:30] <unewbie> CppIsWeird: use it as a guest os?
[02:30] <CppIsWeird> i imagine so.
[02:31] <CppIsWeird> might have less driver support initially due to it expecting a virtual environment
[02:38] <trimeta> I have a couple of questions about disk drive checking.
[02:39] <trimeta> So, I rebooted my system only to find I couldn't ssh into it or anything.
[02:39] <trimeta> I then rebooted the machine (the system was on for less than 15 minutes before this), and got the "disk drive checking" screen, which means that's probably what it had been doing beforehand.
[02:40] <trimeta> The first question is, how likely is it that unattended, it had fully fsck'ed my 4.5 TB RAID5 array and had begun writing changes to disk, changes that I messed up?
[02:41] <CppIsWeird> dont waste your time, no one is around
[02:41] <trimeta> CppIsWeird: Ah. Oh well; how's the main channel? Or is this too esoteric a question for them?
[02:41] <CppIsWeird> dunno, im banned from the main channel for loosing my patience with no one answering my question and spammed the hell out of it
[02:42] <CppIsWeird> childish, but effective at expressing frustration
[02:42] <trimeta> Was it effective at getting an answer?
[02:42] <CppIsWeird> oddly enough, yes.
[02:42] <trimeta> In my experience, asking a second time 5 minutes later is usually helpful, but after that you just piss people off.
[02:42] <CppIsWeird> not that i suggest doing that.
[02:43] <CppIsWeird> i asked intermittently throughout the day to no avail
[02:46] <CppIsWeird> for two days i think
[02:46] <trimeta> I can see how that would get frustrating, yes.
[03:04] <Jeeves_Moss> I'm getting "dovecot: IMAP(scott@moseley.ca): Invalid data in file /mnt/raid/www_root/mail/moseley.ca/scott/.Sent/dovecot-uidlist" as an error.  What file is it requesting, and how do I repair it?.
[03:14] <Error404NotFound> I am setting up a box that will be used as reverse proxy for some of my sites and will host some static content. What is the best option? Apache? Nginx? Lighthttpd?
[03:14] <Jeeves_Moss> Error404NotFound, I prefer apache.
[03:15] <Error404NotFound> Jeeves_Moss, i do as well, but isn't that an overkill for such a simple task?
[03:17] <Jeeves_Moss> 'meh.
[03:17] <Jeeves_Moss> Error404NotFound, I find that I usally "over build" stuff becasue I'm lazy
[03:18] <Error404NotFound> Jeeves_Moss, hmmmm
[03:23] <sdfoj8> how i can add new vps to my ubuntu server ? on 10.4 ?
[03:25] <Jeeves_Moss> sdfoj8, it looks like it's mostly dead in here tonight
[03:25] <sdfoj8> wmware is already installed with ubuntu right ?
[03:25] <sdfoj8> so i can add new vps
[03:25] <sdfoj8> ?
[03:26] <sdfoj8> also, i want to ask that, people are sometimes installing vps and run their works from virtual machines instead of running it from a dedicated pshyically
[03:26] <sdfoj8> what is the reason for that
[03:29] <CppIsWeird> I installed ubuntu-server 10.04 on a server with multiple hard drives. upon installing ubuntu on one of these hard drives the computer would not boot into linux. removing all the hard drives except the drive i want to install ubuntu on and ubuntu will boot up. can anyone help me fix so that i can boot ubuntu and have all my hard drives installed?
[03:31] <Jeeves_Moss> CppIsWeird,  lol, I'm guessing that the BIOS is setting the wrong drive to the first drive.  I would put everything back together the way you're going to keep it, boot from the CD, and do a GRUB repaie
[03:32] <Jeeves_Moss> sdfoj8, I just run VMWare
[03:32] <CppIsWeird> Jeeves, i installed it once with all drives the way they are. that did not work. so i removed all the drives and installed ubuntu again, then it worked. use same course of action?
[03:33] <Error404NotFound> If i am creating a reverse proxy, does it actually have to a AMP server if the server behind it is a AMP?
[03:33] <Error404NotFound> I mean, does the reverse proxy box even need php5 mod and all that?
[03:48] <Jeeves_Moss> CppIsWeird, you need to get the drives installed the way they're going to stay, then install the OS?
[03:50] <CppIsWeird> i did that. and nothing worked.
[04:36] <Jeeves_Moss> CppIsWeird, you need to get the drives installed the way they're going to stay, then install the OS?
[04:36] <Jeeves_Moss> I'm getting "dovecot: IMAP(scott@moseley.ca): Invalid data in file /mnt/raid/www_root/mail/moseley.ca/scott/.Sent/dovecot-uidlist" as an error.  What file is it requesting, and how do I repair it?
[04:39] <jmarsden> Jeeves_Moss: That's a file tracking what messages scott@mosely.ca has read in his Sent folder.  You can probably just delete it and it will be rebuilt when needed.
[04:39] <jmarsden> He may end up re-downloading some of his Sent messages, but it should solve the error.
[04:50] <riz0n> Hey guys, I m running an Ubuntu Server w/ Apache. Is there a way to make http://host/~user go to /home/user/www ?? Thanks
[04:50] <twb> It should go to ~user/public_html if the appropriate module is enabled.
[04:51] <riz0n> ok, that will be fine too
[04:51] <riz0n> how would i go about enabling the module for that
[04:51] <riz0n> and is there a way for me to choose which "virtual server" it would place the /~ directories under, or would it apply to all virtual servers running on the machine
[04:51] <twb> I imagine it's a2enmod(8) with some form of argument.
[04:52] <twb> I don't use apache myself.
[04:52] <twb> ITYM "virtual host".
[04:52] <twb> And I don't know.
[04:52] <jmarsden> userdir, I think.
[04:53] <riz0n> I will ask the guys in #httpd and see if they have any further info
[04:53] <jmarsden> Looks like    sudo a2enmod userdir
[04:58] <jturek> riz0n: https://wiki.ubuntu.com/UserDirectoryPHP
[04:58] <jturek> might help as well
[05:38] <riz0n> Thanks for your help guys. Got it all set up and running!
[09:05] <RoyK^> what does it take to make canonical fix a bug? bug 579267 has been open for quite some time now, and the debian bug for this issue is even older. the problem is fixed in redhat and upstream kernels
[09:10] <RoyK^> ubottu: bug #579267
[09:10] <RoyK^> erm, I mean bug 579276
[09:28] <corpsegrindr> Some days streaming tv from my server works just fine and some nights it wants to buffer every 3 sec. I have no other down/uploads going and minimal use of bandwidth. Is there any key targets i can look for to fix this problem?
[09:28] <alvin> Thanks RoyK^. This bug explains a few things.
[09:29] <corpsegrindr> or is there a way to have a video file temp load well on pause (kind of like youtube)
[09:50] <RoyK^> alvin: _
[09:50] <RoyK^> alvin: ?
[12:54] <dsfwea23> can someone tell me how I can change drupal to work in the root directory of apache2 in 10.04?  Default install wants /drupal6/*
[12:55] <trapmax> question about confluences webdav-plugin: the plugin is enabled, but when i try to connect to webdav://<server>/plugins/servlet/confluence/default, i get only get prompted for username/password
[12:55] <trapmax> then get "file or folder does not exist" -error
[12:55] <trapmax> testing with konqueror
[12:59] <Daviey> dsfwea23: I've never used drupal from the repo, but i would imagine the setting is in /etc/apache/conf.d/*drupal*
[13:00] <Daviey> dsfwea23: If you are using it with a virtual host, then you should be able to make the setting in the /etc/apache2/sites-avaliable/$somesite
[13:00] <dsfwea23> Daviey, the config in the repo actually has to be copied in, by default to /etc/apache2/sites-available, then enabled... but what I'm running into is that it doesn't play well when I tell it to use root directory
[13:01] <Daviey> dsfwea23: Hmm.. i don't know..  I'd need to know the error, as i say - i've never used drupal from the archive
[13:01] <Daviey> dsfwea23: The other thing you could do, is an internal rewrite from your docroot
[13:01] <ttx> hallyn: I edited the server-maverick-hypervisor whiteboard: the comments you added were considered as new work items. If you want to make comments, use a separate section... or the wiki spec.
[13:02] <dsfwea23> basically I'd like to run drupa6 for my root site with clean urls, but still have access to things like /squirrelmail etc.  But once I change the directory for drupal to root apache doesn't like the other aliases anymore.
[13:03] <dsfwea23> Daviey, rewrite might work better for me.
[13:03] <Daviey> dsfwea23: Rewrites are pretty good.. i'm sure you'll be able to work out how to get started with that
[13:04] <dsfwea23> Daviey, =)
[13:04] <dsfwea23> gets kind of complicated with drupal's clean urls since that does it's own rewrite
[13:10] <Daviey> dsfwea23: Sounds like this is an ideal oppertunity to document how you resolve the issue on the wiki :)
[13:23] <dsfwea23> well seems it won't matter, there is an issue with drupal6 in 10.04 with php 5.3.
[13:28] <screen-x> The UEC image store is mentioned here: http://www.ubuntu.com/cloud/private/deploy is it possible to browse the available images without having a UEC installation? I want to see what's available.
[13:28] <hallyn> ttx: k
[13:29] <ttx> hallyn: this spike in WI caused a near heart attack to me this morning :P
[13:30] <hallyn> it should've showed up in monday's chart too right?
[13:30] <ttx> ah ?
[13:30] <ttx> it's been there since Monday ?
[13:31] <ttx> then... someone else must have been adding WIs :)
[13:31] <ttx> smoser: around ?
[13:31] <hallyn> ttx: yeah they were there last week, assuming you're talking about what i think you're talking about
[13:31] <smoser> ttx, here.
[13:32] <ttx> smoser: -> pm
[13:32] <ttx> well no.
[13:32] <ttx> smoser: just as a reminder...
[13:32] <ttx> smoser: alpha2 freeze is next Thursday
[13:32] <smoser> yeah. thanks.
[13:32] <ttx> smoser: so I suggest you concentrate on what would land in main packages... cloud-init.
[13:32] <smoser> i'm making progress on the cloud init stuff.
[13:33] <ttx> smoser: the kernel upgarde stuff can actually be handled during the freeze.
[13:33] <ttx> since it won't result in formal landing in packages yet.
[13:34] <ttx> SpamapS: same for you... everything that lands into main packages on your side should be completed by next Thursday.
[13:34] <ttx> SpamapS: (that should include cloud-utils and web20-workloads)
[13:36] <ttx> kirkland, Daviey: I propose that we have a quick Mumble session when you're both around, to prepare the euca call.
[13:39] <ifoo> hi, i'm trying to install ubuntu server 10.04 on my intel ss4200 NAS box. for the system drive i use an IDE DOM. problem is the ubuntu installer doesn't load the right IDE module (ide_generic + ata_piix).
[13:42] <ifoo> and it seems they are not even contained on the installation cd. the SATA controller is an intel ICH7 chip. It works with the debian installer. are these modules removed from the default ubuntu installer?
[13:52] <sommer> morning
[14:11] <amstan> hey guys, i'm looking to regular backups with a script
[14:12] <amstan> i know how to use cron and copy files, the interesting part is getting the script to keep only the latest 5 backups for example
[14:14] <ElTonerino> amstan, I use rsnapshot.  More than hapy with it.
[14:16] <ElTonerino> amstan, http://rsnapshot.org/
[14:16] <amstan> i like that idea of hardlinks
[14:17] <ElTonerino> I use it to back up a server and a couple of laptops.  It's saved me a few times.  Yesterday being the most recent.
[14:18] <amstan> does it delete old backups automatically?
[14:19] <ElTonerino> yes
[14:31] <Kartagis> hi
[14:31] <Kartagis> how do I generate some passwd.db out of passwd file?
[14:38] <soren> passwd.db for what?
[15:00] <konryd> hello everyone. How do I connect to a freshly-baked vm instance?
[15:01] <konryd> I prepared one with vmbuilder. It has openssh-server installed, and I know the username and the password
[15:03] <bogeyd6> konryd, which virtualization package did you use?
[15:03] <konryd> bogeyd6: kvm
[15:04] <bogeyd6> sorry but i only use vmware
[15:18] <tio_> hello
[15:18] <tio_> can somebody help me<
[15:18] <tio_> ?
[15:18] <guntbert> !ask | tio_
[15:19] <tio_> ok, so Hello, I need help with RTL8111DL 1gbps wired and broadcom4311 wireless network interface. I have mainboard d945gsejt. I have in PCI one 100Mbps wired network interface. When I was installing system know both wired intrfaces. I select 100Mbps as primary (will be WAN port). 1Gbps will be bridged with wireless to LAN. Problem is, that system not recognize these two cards. after install
[15:19] <tio_> root@atom:~# ifconfig
[15:19] <tio_> https://gist.github.com/4986fb78b804ff2911ea
[15:20] <Kartagis> soren, I figured out thanks
[15:20] <Kartagis> the command was postmap
[15:20] <tio_> This show only one - primary. So I downloaded latest realtek driver and install it. But no change. So I fix kernel modules, because was loading r8169 instead r8168.ko. It is ok now. But no change
[15:20] <_ruben> first mistake: forget about the existence of ifconfig (and route)
[15:20] <tio_> Broadcom have same issue. I installed drivers -- package b43-fwcutter - extracted fw, rebooted. But no change. So I looked here
[15:21] <tio_> root@atom:~# lshw -c network
[15:21] <tio_> I see, that these two cards are recognized but disabled
[15:21] <tio_> https://gist.github.com/cb05afd045e3dc8eb1a3
[15:21] <tio_> I attached this output (look last three devices)
[15:21] <tio_> root@atom:~# lspci -v
[15:22] <tio_> https://gist.github.com/9a0e3277853e34e960f7
[15:22] <tio_> Would somebody help me? I have installed Ubuntu server 10.04 LTS 32bit. Thanks
[15:25] <hallyn> hm, centos 5.5 won't boot (once installed) in lucid kvm - whines about apic.  sounds like a 3 year old kernel!
[15:33] <hallyn> ah, it's a xen kernel thing
[15:41] <tio_> haha, I am lama, I just add interfaces to /etc/network/interfaces and its ok now
[15:43] <zul> hallyn: smoser had some documentation about that
[15:43] <hallyn> oooh, cool, thanks
[15:44] <smoser> oh yeah, documentation is my specialty
[15:44] <smoser> :)
[15:44] <smoser> but, no, i dont know much about that.
[15:44] <hallyn> gah
[15:45] <kirkland> smoser: speaking of documentation ....
[15:45] <smoser> i did get centos 5.3 booting in kvm, and running xen inside it
[15:45] <smoser> though
 smoser: could you take a look at https://help.ubuntu.com/community/UEC/Provisioning/DD-WRT
[15:45] <smoser> that was probably what zul was thinking of.
[15:45] <hallyn> smoser: i'm just surprised that even booting with --no-acpi didn't work
[15:46] <zul> smoser: yeah it was
[15:47] <hallyn> (http://us.generation-nt.com/bug-528077-kvm-fails-setup-timer-interrupt-via-io-apic-help-168961491.html suggests there's not much to be done)
[15:48] <smoser> i think that looks reasonable, kirkland
[15:48] <kirkland> smoser: cool, anything to add?
[15:49] <kirkland> smoser: i'm working on documenting all of the uec auto provisioning steps
[15:50] <smoser> no
[15:50] <smoser> kirkland, you might want to point at that page from https://help.ubuntu.com/community/UEC/Tips
[15:50] <smoser> or replace that one
[15:56] <Genk1> hello
[15:58] <hackeron_> hey, I have ubuntu set up as a nat router to share the internet connection. I also want to add NAT-PMP/UPnP for automation NAT traversal so I don't have to manually manage the firewall for games/bittorrent - how would I set that up?
[15:59] <kirkland> smoser: yeah, when it's all ready to go
[15:59] <kirkland> smoser: https://help.ubuntu.com/community/UEC/Provisioning will be the entry point
[15:59] <smoser> yeah, i dont really care.
[15:59] <kirkland> smoser: so far, i've documented dd-wrt and setting up a mirror
[15:59] <smoser> just pointing at the old tips page, at least put that all together
[15:59] <kirkland> smoser: i still need to do the pxe, tftp, wakeonlan, preseed, and web front end docs
[15:59] <kirkland> thanks
[16:01] <Genk1> I just want to find a way to disable nat in my station
[16:01] <Genk1> how can I proceed
[16:01] <Genk1> ?
[16:05] <_chris__> heja. i have mounted a windows share to my server. now i want to copy a folder from linux to windows share. ive done t hat with "cp /folder/ -R /share/" problem is: i have some folders/files there with german chars in it lik 'äöü' and when they are copied to the windows share the names are pretty weird cause he makes stranges things out of the special chars
[16:05] <_chris__> is there a way to solve this ?
[16:06] <nimrod10> is there a way to force passwd to unlock users without a password ?
[16:06] <bogeyd6> nimrod10, can you be more general
[16:07] <bogeyd6> _chris__, is the windows share a en_US bawx?
[16:07] <nimrod10> what do you mean bogeyd6 ?
[16:07] <_chris__> no its german
[16:07] <_chris__> ad bogey
[16:07] <bogeyd6> nimrod10, you mean when they login to a box, a samba share, or the http, what have you?
[16:08] <bogeyd6> _chris__, did you set your term to UTF8?
[16:08] <_chris__> bogeyd6, u mean while mounting the drive ?
[16:09] <nimrod10> :) creating a user as a linux account with a blank password is possible but the account is locked, and doing passwd -u username won't unlock the user. Is there a way to force passwd ?
[16:09] <nimrod10> that is on a ubuntu-server 9.10
[16:09] <bogeyd6> _chris__, your system locale should be UTF8 to include those
[16:09] <bogeyd6> nimrod10, so you want a user to have a blank password and then upon login they are forced to change it?
[16:10] <_chris__> bogeyd6, okay, im pretty new to linux, how can i set this ?
[16:10] <_chris__> bogeyd6, its just a virtual machine for testing purpose
[16:11] <nimrod10> bogeyd6, the user logs in via ssh keys auth, after that I want them to be able to set the passwd they want. If I create their account with an empty passwd then when they set the pass it won't ask for the old pass
[16:11] <bogeyd6> _chris__, server or desktop?
[16:11] <bogeyd6> nimrod10, i guess you answered your own question. set it to a generic password first
[16:11] <nimrod10> but if  they have a blank pass ubuntu locks the account. and passwd -u won't unlock it
[16:12] <bogeyd6> this command will solve it too # usermod -p “” foo # chage -d 0 foo
[16:12] <_chris__> bogeyd6, the machine with the share is a desktop
[16:12] <amstan> ElTonerino: thanks for that suggestion, rsnapshot works great
[16:12] <nimrod10> bogeyd6, if I set it to a generic pass then everyone knows the pass and they can become the other user that hasn't changed his pass yet
[16:12] <ElTonerino> amstan, Glad to hear it.
[16:14] <bogeyd6> _chris__, https://help.ubuntu.com/community/LocaleConf
[16:15] <bogeyd6> i dont know if you can make a blank password and login through ssh with it
[16:15] <bogeyd6> i know you can use an ssh key instead of a password
[16:16] <bogeyd6> but as far as just a str8 up blank login, dont think so
[16:16] <hackeron_> anyone? I have ubuntu set up as a nat router to share the internet connection. I also want to add NAT-PMP/UPnP for automation NAT traversal so I don't have to manually manage the firewall for games/bittorrent - how would I set that up?
[16:17] <bogeyd6> hackeron_, https://help.ubuntu.com/community/Servers#UPNP%20Mediaserver
[16:18] <corpsegrindr> Some days streaming tv from my server works just fine and some nights it wants to buffer every 3 sec. I have no other down/uploads going and minimal use of bandwidth. Is there any key targets i can look for to fix this problem?
[16:18] <corpsegrindr> or is there a way to have a video file temp load well on pause (kind of like youtube)
[16:18] <_chris__> bogeyd6, set it to de_DE.UTF-8 now, but still same :/
[16:19] <hackeron_> bogeyd6: I don't want to set up a media server, I just want to enable NAT traversal
[16:19] <bogeyd6> _chris__, you didnt log out and log back in
[16:20] <_chris__> hm
[16:20] <_chris__> true :)
[16:20] <_chris__> brb
[17:11] <SpamapS> ttx: should make some progress on cloud-utils today. :)
[17:16] <jdstrand> hallyn, kirkland, smoser: fyi-- I uploaded libvirt 0.8.1 yesterday. the only oddity is 0.8.1 ftbfs on powerpc because of a test case failure. I'm not sure that is a problem for you, but wanted to explicitly mention it. I don't plan to look at that any time soon
[17:16] <ttx> SpamapS: good :)
[17:16]  * ttx disappears
[17:17] <jdstrand> hallyn, kirkland, smoser: I can say that redhat disabled some tests in their builds because of powerpc test suite failures, but not this one. In other words, our powerpc build didn't fail the one they blacklisted and their powerpc didn't fail on the one that ours did
[17:18] <smoser> nice
[17:18] <jdstrand> the test suite has always been a little fidgety...
[17:19] <jdstrand> it shouldn't be hard to disable that test on powerpc only, but like I said, I don't have a lot of time for that atm
[17:22] <jdstrand> actually, it wouldn't be earth shattering if just the nwfilterxml2xmltest was disabled... but rather than just disabling it, I think it should be investigated, which is precisely the part I don't have time for
[17:23] <jdstrand> hallyn, kirkland, smoser: ^
[17:24] <hallyn> jdstrand: so if i update a maverick box i'll get the new package?
[17:24] <jdstrand> hallyn: oh yes :)
[17:24] <hallyn> rockin'  thanks
[17:24] <jdstrand> https://launchpad.net/ubuntu/+source/libvirt
[17:24] <jdstrand> hallyn: sure! :)
[17:25] <jdstrand> hallyn: I cleaned out some old packaging cruft too. the delta is still significant, but at least it is smaller than lucid
[17:26] <jdstrand> hallyn: and upstream didn't break apparmor this time either :) only thing I added for the apparmor driver was a patch that they forgot to commit to 0.8.1 that will be in 0.8.2
[17:27] <jdstrand> (I did fix some profiling bugs, but that is ubuntu-specific)
[17:29] <jdstrand> hallyn: if you are interested in the powerpc ftbfs (and I hope you are ;), this would be a good opportunity to become acquainted with porting boxes (ie davis.canonical.com for powerpc)
[17:32] <hallyn> jdstrand: IIUC you're saying the new package does not build from source on powerpc, and I should take a look at why...
[17:33] <hallyn> i'll take a look at that - if nothing else it'll test my account's ability to get into the machines :)   (after testing lxc merge)
[17:34] <jdstrand> hallyn: that is exactly what I am suggesting :) though, it builds, it fails a test in the test suite (which is enabled in the build). See http://launchpadlibrarian.net/50454753/buildlog_ubuntu-maverick-powerpc.libvirt_0.8.1-2ubuntu1_FAILEDTOBUILD.txt.gz
[17:38] <kirkland> hallyn: jdstrand: powerpc is *very* low priority for us, however
[17:38] <kirkland> should be fixed, but should not consume your time
[17:39] <smoser> SpamapS, i just shoved a 'Usage.txt' file to lp:~smoser/junk/uec-run-instances/ that describes kind of what we're hoping to hvae.
[17:39] <LowValueTarget> I have a deb package directory structure I built with dh_make..... is there an easy way to increment the package version without having to manually put entries in the changelog etc?
[17:40]  * jdstrand nods
[17:41] <jdstrand> part of why I can't jump on it atm
[17:43] <jdstrand> hallyn: I neglected to mention that only amd64, i386 and armel are officially supported architectures. all others are community driven. I brought this up cause I don't know how important powerpc is in reality to the server team, which kirkland answered
[17:43] <jdstrand> hallyn: earlier releases have a different list, but last few releases are those 3 archs
[17:47] <jdstrand> hallyn: my rule of thumb when working on devel is to not break the community archs needlessly, but I'll fix when I have time for it
[17:51] <jdstrand> mdeslaur: oh, I forgot to mention to you that I uploaded a new libvirt yesterday (0.8.1) in case you want to refresh virt-manager and friends
[17:52] <mdeslaur> jdstrand: ok, thanks
[17:57] <SpamapS> smoser: thats super helpful thanks. :)
[18:00] <SpamapS> smoser: should I branch your uec-run-instances and work from that? It looks to have quite a few things different from the one in lp:ubuntu/cloud-utils
[18:01] <smoser> SpamapS, you can start from that branch that i have (the junk one).
[18:01] <SpamapS> smoser: sweet thanks
[18:01] <smoser> i hope to basically be bacwards comaptible with the cloud-utils/uec-run-instances
[18:01] <smoser> but a whole lot more powerful
[18:10] <unlex> anyone using a dual p4 xeon as servers?
[18:13] <JackTO> hi all, i'm configing my ip for the first time ibn linux, in the /etc/networks/interfaces file, what is the field "network" for, its under netmask, i nebver had this in wondows
[18:16] <steffan> JackTO: "man interfaces" explains that
[18:18] <JackTO> just says requied for 2.0 kernels, but what does it do, mean?
[18:19] <JackTO> if my address ip is 212.203.83.4, does that mean that addess always ends in a 0, like 212.203.83.0
[18:22] <kaushal> hi
[18:22] <kaushal> Anyone using collectd on ubuntu ?
[18:29] <unlex> i have a dual P4 Xeon and was wondering if i should derack it considering it's got only two physical CPUs with one core each which is rathr quaint in my opinion but would it fit to buoild a cluster/server?
[18:30] <unlex> or replace it with an entry level X3430 single socket quad core box ?
[18:39] <elnur> How to force bind9 to update its cache?
[18:47] <_ruben> elnur: rndc flush (it doesnt really update it, it flushes it)
[18:47] <jilly> hi. im setting up dns
[18:47] <jilly> how do i find my network address?
[18:47] <jilly> like 192.168...
[18:49] <SpamapS> jilly: whose network adress?
[18:49] <SuperLag> SpamapS: I want yours, so I can set up some scripts on your server. :)
[18:50] <SuperLag> SpamapS: and don't tell me 127.0.0.1, either :)
[18:51] <LowValueTarget> why would apache freeze on serving up css files
[18:53] <kirkland> SpamapS: what's your LP id?
[18:54] <SpamapS> kirkland: clint-fewbar
[18:54] <SuperLag> You're fubar, alright :)
[18:54] <SuperLag> How goes it, dude?
[18:55] <SuperLag> I'm just in one of those moods.
[18:55] <SuperLag> SpamapS: are you working hard, or hardly working?
[18:55] <SpamapS> SuperLag: a bit of both :)
[19:00] <owen1> i created new user. when i run something with sudo i get:
[19:01] <owen1> sudo: must be setuid root
[19:01] <owen1> any clues?
[19:04] <jonny2> Hi, I am trying to setup a server with 6 IP addresses on 1 card, but only one IP address works, /etc/network/interfaces http://pastebin.com/bX7nEsFW, ifconfig output http://pastebin.com/sU31SMys, any pointers would be helpful, I have based /etc/network/interfaces on the server guide
[19:07] <Rigorm0rtis> Hello, I am having problems with Ubuntu crashing when I copy a large file to it over the network. Whenever I transfer a file > 60GB the system crashes. I have had this happen over ssh and samba. If I write a large file locally, the system does not crash. I have logs. This is happening on Ubuntu 10.04 server x64 on an EXT4 partition that is 2tb. I have run memtest on the machine and it held...
[19:07] <Rigorm0rtis> ...up for three passes. Any ideas?
[19:09] <elnur> _ruben, it didn't help. my server still has old domain's data. maybe i should flush something else, not bind?
[19:11] <_ruben> elnur: it depends on your dns infrastructure
[19:19] <kaushal> Can someone please guide me about my query regarding collectd on ubuntu server ?
[19:27] <_ruben> !poll
[19:58] <hggdh> Rigorm0rtis: what do you mean by "crashing"? You get a kernel OOPS?
[20:00] <Rigorm0rtis> hggdh: The system completely locks up. I can't get a picture from attached monitor, and any network connections drop.
[20:15] <hggdh> Rigorm0rtis: yeah, sounds bad. Is this a server?
[20:17] <hggdh> brb
[20:24] <corpsegrindr> Some days streaming tv from my server works just fine and some nights it wants to buffer every 3 sec. I have no other down/uploads going and minimal use of bandwidth. Is there any key targets i can look for to fix this problem?
[20:27] <Plecebo> I'm having trouble getting postfix to send mail on my server. I tried testing with telnet but i've never tested postfix before so not sure what success/failure looks like. Can someone help me?
[20:40] <pmatulis> Plecebo: telnet is just a means.  the result is that mail is sent.  check /var/log/mail.log or recipient inbox.  however, i'm sure googling for 'testing smtp telnet' will help
[20:42] <jeremyn> Plecebo: i like to use nc to the smtp port and then type "ehlo <domain>"
[20:43] <jeremyn> Plecebo: or submission port, like run, on the server "nc localhost 587" and then type "ehlo <domain>" and hope you get back some responses that sound like mail
[21:17] <Rigorm0rtis> hggdh: Yes, a file server primarily.
[21:18] <cwillu_at_work> Rigorm0rtis, are you familiar with netconsole?
[21:19] <_ruben> cwillu_at_work: netconsole as in performing the install procedure over ssh?
[21:19] <cwillu_at_work> Rigorm0rtis, also, do the magic-sysrq keys work?  (alt-sysrq-[r, e, i, s, u, b] to cleanly reboot)
[21:20] <cwillu_at_work> _ruben, no
[21:20] <cwillu_at_work> _ruben, netconsole is a low level output for the kernel log, which usually allows you to capture oops and such when the system goes down
[21:24] <Rigorm0rtis> cwillu_at_work: I am about to leave work right now. I will return in about 30m.
[21:27] <Ichat> can someone help me with an upgrade issue....
[21:27] <guntbert> !ask | Ichat
[21:27] <Ichat> atm im running ubuntu 8.04 - with ebox    on a   3TB raid5  array (ext 3)
[21:28] <Ichat> now i want to install  the new lts,  with  it converted (if posible and safe)  to  ext 4 ) can i do that
[21:30] <Ichat> as its being  3tb data,   id dont like the idea of loosing it,   even though i have backups  (on mostly old  160gb spares..     one can emagine that  restoring data wont be an option to consider light
[21:30] <cwillu_at_work> Ichat, for the most part, ext4 is just ext3 with some extra features turned on, in the same way that ext3 is just ext2 with a journal
[21:31] <cwillu_at_work> however only newly written files will gain benefit from things like extents
[21:31] <cwillu_at_work> investigate "tune2fs"
[21:32] <Ichat> meaning it should be done after i installed the new server and server software
[21:32] <guntbert> cwillu_at_work: and a few bugs added as anyone can tell who tries to use a virtual box guest with a virtual SATA adapter when the host FS is ext4
[21:34] <Ichat> so i should just 'install'  the new server, and ebox  (im not going to try upgrading it) ....  than get my / old     /home   partition mounted again,   and than worry about convertiong it to  ext 4  ???
[21:35] <Ichat> strange (i think) that the server cdrom didn't  offer me an upgrade ?
[21:35] <SpamapS> grr.. why doesn't vim have the modeline option on by default!? argh.
[21:35]  * SpamapS wants his 5 minutes back
[21:38] <_ruben> SpamapS: "security feature" i think
[21:42] <SpamapS> smoser: so I think I'm going to use twisted.conch for the ssh not paramiko.. that way we can check all the instances asynchronously :)
[21:43] <smoser> i have no opposition to that.
[21:43] <smoser> i had gone the twisted route, but then decided that it wasn't terribly useful
[21:44] <smoser> as
[21:44] <SpamapS> smoser: The bash script calls the euca/ec2 tools for describe.. is there an API equivilent?
[21:44] <smoser> a.) we dont want to finish until all things are finished
[21:44] <smoser> b.) i forget what b was.
[21:45] <SpamapS> smoser: yeah b was probably that you can't start ssh'ing until you have the console scrape anyway.. right?
[21:45] <smoser> there are definitely api equivalents to describe images anda get-console-output
[21:45] <smoser> that might have been it :)
[21:45] <smoser> one thing the twisted would really help witih, that we dont want to be serialized on is multiple instances running a long running command
[21:45] <SpamapS> smoser: if I can use twisted for that step then if nothing else we can finish a little earlier. Is this so people can start 100 instances efficiently, or 5?
[21:46] <smoser> so the reason for execing PREFIX-run-instances, PREFIX-describe-images, PREFIX-get-console-output
[21:46] <smoser> instead of using an API is that you can the assume that the user has set those commands up
[21:47] <SpamapS> smoser: ah, the API's don't share the same setup as the tools?
[21:47] <smoser> well, euca and ec2- use different ways.
[21:47] <SpamapS> makes sense
[21:47] <smoser> i didn't want to invent another.
[21:47] <smoser> but, yeah, that is prettier.
[21:48] <smoser> SpamapS, i dont know if its 100 or 5.
[21:48] <smoser> i'd say "and run command" is likely a 1 thing
[21:48] <smoser> but can't hurt
[21:48] <smoser> i'm not at all opposed to having things twisted wher epossible.
[21:49] <SpamapS> smoser: I mean, part of me just wants to do exec("ssh blah &") for that. ;)
[21:50] <SpamapS> smoser: I'll keep it in mind... paramiko *is* simpler... and I like simple
[21:50] <ghost_> hi all I'm having a problem running rails I get a could not find RubyGem active support error http://pastebin.com/NgRh2fLz
[21:50] <ghost_> any help is appreciated
[21:50] <SpamapS> smoser: but I can see waiting for 100 ssh's taking a long time if not done in parallel somehow.
[21:51] <smoser> mostly i was going th eparamiko route for verifying keys
[21:51] <smoser> i didn't know about twisted.conch
[21:51] <smoser> and also, paramiko can read .ssh/known_hosts and such
[21:52] <smoser> but... id ot know what others can do
[21:54] <smoser> SpamapS, one thing i'd really like to do well is get the command execution remotely done well
[21:55] <smoser> there are a couple thing si dont know really what to do about
[21:55] <SpamapS> smoser: when spawning instances, how is the stuff in .ssh/known_hosts useful ?
[21:55] <smoser> well if you're connecting to them
[21:56] <cwillu_at_work> SpamapS, making sure you're logging into the machine you think you're logging into is good
[21:56] <smoser> the idea was that --known-hosts would update your known-hosts such that subsquent 'ssh host' would have the entries
[21:56] <SpamapS> cwillu_at_work: spawning EC2 instances means not necessarily having a known key.
[21:56] <smoser> you scrape console for keys, verify that versus ssh-keyscan, and then write the entry to the users known hosts.
[21:56] <SpamapS> smoser: ah, so its more about saving to it
[21:57] <smoser> right, but in saving to it, you also have to remove
[21:57] <smoser> and remove with hashed values
[21:57] <smoser> which, i had no idea how i was going to do. paramiku does that sort of thing . i've not actually done it, but read about it.
[21:57] <smoser> the things i want to well but dont know exactly how to do are in remote execution:
[21:57] <smoser> - ssh somehost foo bar "wark two"
[21:58] <smoser>   that always gets screwed up. the commands are never delimited correctly as they go through multiple levels of shell interpretation
[21:59] <smoser>  i'd like to make it work correctly. i'm not sure if the deficiency is in the ssh client, or the ssh server.  ie, i dont know if going through a library rather than system(ssh) is going to help that.
[21:59] <SpamapS> smoser: right you have at least 3 shells to screw it up (your login shell, the shell that the script uses to run ssh, and the shell on the other side)
[21:59] <smoser> - for a long running command on the remote host, your network connection could get lost
[22:00] <smoser> i'd like to somehow recover or "be correct"
[22:00] <cwillu_at_work> you're using python?  Don't use os.system or any of the routines that do shell calls if you can avoid it, having the extra shell just causes problems
[22:01] <SpamapS> cwillu_at_work: we're more talking about how it has been done in the past. I intend to send the command directly down the pipe w/ either paramiko or conch. :)
[22:01] <cwillu_at_work> okay
[22:01] <cwillu_at_work> none the less, os.system calls are a bad sign in a codebase
[22:02] <cwillu_at_work> subprocess.popen is the correct approach, where you pass a list of strings rather than allowing the shell to perform the splitting for you
[22:02] <cwillu_at_work> ignoring the conch approach :p
[22:02] <SpamapS> cwillu_at_work: noted.. since I may be calling some commands as well. :)
[22:03] <cwillu_at_work> hmm
[22:04] <smoser> SpamapS, i dont know if "right down the pipe" is truely "right down the pipe"
[22:04] <smoser> in the end, the thing that i want is 'sudo' like behavior rather than 'su' like behavior
[22:04] <cwillu_at_work> explain?
[22:05] <smoser> sudo -u ${USER} sh -c 'for f in "one 1" "2 two" "three 3"; do echo $f; done'
[22:05] <smoser> one 1
[22:05] <smoser> 2 two
[22:05] <smoser> three 3
[22:06] <SpamapS> smoser: playing w/ paramiko, I can feed it a raw key and it will save to a known_hosts file. :)
[22:06] <cwillu_at_work> ugh, sudo -u ${USER} -i 'for f in "one 1" "2 two" "three 3"; do echo $f; done would work
[22:07] <SpamapS> smoser: so while it seems strange, I can use paramiko for the local key management while still using conch for the transport.
[22:07] <SpamapS> which it doesn't seem to do
[22:07] <SpamapS> er
[22:07] <smoser> cwillu_at_work, the above command and output is correct
[22:07] <SpamapS> conch doesn't seem to do local key management
[22:08] <smoser> basically, su takes '-c' "command string"
[22:08] <smoser> sudo takes -- command arg arg2 arg3
[22:08] <smoser> the latter is greatly superior in my experience
[22:12] <SpamapS> smoser: I believe su simply switches user and runs a shell whereas sudo execs the binary you told it to exec
[22:15] <holmser> is there any way to forcibly unbind a port without rebooting my server?
[22:15] <SpamapS> holmser: kill the process that has it bound
[22:15] <holmser> already did
[22:15] <SpamapS> kill -9 ?
[22:16] <holmser> killall
[22:16] <SpamapS> try the same killall but 'killall -9 whateveryoudidbefore'
[22:17] <holmser> got it... thanks
[22:34] <Rigorm0rtis> Hello, I am having a problem with Ubuntu 10.04 server. When I transfer a file >60 GB over the network to the server hard locks after about 60GB is transferred. This happens over ssh, and samba. The destination directory is an ext4 2tb hardware raid10 array. When I create a large file locally on the machine it does not crash. Memtest reports no errors after 3 passes. Does anyone have any...
[22:34] <Rigorm0rtis> ...ideas? I have a few logs.
[22:35] <Ng> kirkland: hrm, I'm trying to test that qemu-kvm upload and getting weird failures - I don't appear to even be able to start a VM
[22:35] <Ng> kirkland: I dug in some logs to find the command libvirt is executing and got: http://paste.ubuntu.com/451300/
[22:38] <qman__> Rigorm0rtis, sounds like a bug in the ext4 driver to me, is your kernel up to date?
[22:39] <qman__> if it were something else, it probably wouldn't hard lock, but rather fail in some other more graceful manner
[22:39] <qman__> that, or hardware failure
[22:39] <Rigorm0rtis> qman__: I'm using the most recent kernel on 10.04.
[22:40] <Black_Prince> bad sector looks like hw problem
[22:40] <Black_Prince> descriptor
[22:40] <Rigorm0rtis> But then why would it work when I create a file locally?
[22:41] <Ng> kirkland: hrm, belay that, qemu-common wasn't installed properly, so /usr/share/qemu was empty
[22:42] <Ng> yep that was it, VMs start now :)
[22:42] <mathiaz> Ng: is there a missing dependency in qemu-kvm
[22:42] <mathiaz> Ng: ?
[22:42] <Rigorm0rtis> Black_Prince: If I just copy from /dev/urandom to a file it gets past 100GB no problem. But if I try to send a file over the network via ssh, or samba it consistently dies around the same point. The outfile shows as being ~60GB.
[22:43] <Ng> mathiaz: no, I'm not really sure what happened - I was downgrading from packages I'd backported from debian to the ones in -proposed to test a bugfix there, and somehow qemu-common was installed, but empty
[22:43] <qman__> Rigorm0rtis, could be the NIC too, try copying several smaller files, like 10GB, then start the big file
[22:44] <qman__> if the failure point changes to ~50GB, then the network card is a real possibility
[22:45] <Rigorm0rtis> qman__: I regularly copy copy windows images and partimage captures over the network. Today alone I probably transferred >100GB. It very much seems to be the size of one transfer.
[22:52] <kirkland> Ng: why wasn't qemu-common installed properly?  archive, package, or user error?
[22:55] <cwillu_at_work> Rigorm0rtis, does anything show up in kern.log from the actual crash?
[22:56] <cwillu_at_work> Rigorm0rtis, do the sysrq key-combo's do anything
[22:56] <cwillu_at_work> and you should really set up netconsole and reproduce the crash to make sure you're getting any panic or oops messages
[22:57] <Rigorm0rtis> cwillu_at_work: I can check kern.log tomorrow when I am at work. I am not familiar with the sysrq key combos (but I do know that the capslock and numlock keys stop responding). I don't know how to set up a netconsole.
[22:57] <cwillu_at_work> Rigorm0rtis, then I suggest you google them :)
[22:59] <cwillu_at_work> Rigorm0rtis, the sysrq combos may allow a clean shutdown even with a partially hung system, and netconsole will allow you to capture kernel messages on a different machine using a very low mechanism that generally still operates during all but the worst crashes
[22:59] <Rigorm0rtis> I have noticed that right before a crash, I start getting these errors in syslog: http://paste.ubuntu.com/450086/
[22:59] <Rigorm0rtis> they also pop up on the terminal.
[23:00] <cwillu_at_work> okay
[23:00] <Rigorm0rtis> the first two lines, not the entire thing. Right after that seqence in the log file, there is usually another, or messages from the system booting after I reset it :)
[23:01] <Ng> kirkland: I'm honestly not sure, but since I was downgrading let's go with user error :)
[23:01] <kirkland> Ng: heh
[23:01] <cwillu_at_work> Rigorm0rtis, yes, the first two lines are at a different debug level than the rest, this is configurable
[23:01] <kirkland> Ng: how long to tell if it's still memleaking?
[23:02] <Ng> kirkland: I'm just setting up the tests now, but to be fair I should leave it running overnight
[23:02] <Ng> it ought to be obvious when I check in the morning
[23:02] <kirkland> Ng: cool
[23:02] <kirkland> Ng: take your time, just respond in the bug
[23:03] <cwillu_at_work> Rigorm0rtis, might try mounting with nodelalloc
[23:03] <Rigorm0rtis> cwillu_at_work: So is this netconsole thing equivalent to remote syslog?
[23:03] <cwillu_at_work> Rigorm0rtis, yep
[23:03] <cwillu_at_work> but at a low level that survives most crashes
[23:03] <Ng> kirkland: I want to get it done so I can get back to setting up my new server as a VM host ;)
[23:03] <Rigorm0rtis> cwillu_at_work: I guess I do know how to set that up then!
[23:04] <cwillu_at_work> Rigorm0rtis, it's a kernel module you load with parameters
[23:04] <Rigorm0rtis> nodelalloc is an fstab mount option correct?
[23:04] <cwillu_at_work> http://www.mjmwired.net/kernel/Documentation/networking/netconsole.txt
[23:05] <cwillu_at_work> believe so, although if it's your /, you might need to pass it in a different way, not sure
[23:06] <cwillu_at_work> i.e., via rootflags= on the kernel command line or some such
[23:07] <Rigorm0rtis> The drive that the writes are occurring to is separate from the root. This is a storage server, I've got one drive each for / and /home, and then a 2TB raid for the general storage.
[23:10] <Rigorm0rtis> cwillu_at_work: So, tomorrow I should remount the drive with nodelalloc, set up the netconsole, crash it again (if possible) and then grab the logs?
[23:11] <Rigorm0rtis> I'm at home right now and the machine isn't configured for remote access. :(
[23:11] <cwillu_at_work> Rigorm0rtis, you need another local machine anyway to capture the netconsole output
[23:11] <cwillu_at_work> but something like that, yes
[23:12] <Rigorm0rtis> I'll use my personal linux box to capture it. I just have to enable the receipt of syslog messages from remote hosts correct?
[23:12] <cwillu_at_work> no, you need to read the documentation :p
[23:13] <Rigorm0rtis> I am looking at this doc you sent me, and it says The remote host can run either 'netcat -u -l -p <port>', 'nc -l -u <port>' or syslogd.
[23:13] <cwillu_at_work> looks like it can talk to syslogd though, yes
[23:14] <Rigorm0rtis> syslogd seems like a cleaner way to do it to me.
[23:20] <Rigorm0rtis> I will do this and report back tomorrow. Thank you for you help. FYI I've got a thread on the ubuntu forums about it I'm keeping updated as well. http://ubuntuforums.org/showthread.php?t=1510233
[23:21] <Rigorm0rtis> Thanks again. I know my way around Linux I would say, but I don't have much troubleshooting experience (never really have problems ;) ).
[23:22] <cwillu_at_work> ah, ubuntuforums, where good information goes to die :p
[23:23] <Rigorm0rtis> Yeah I hate it when I google a problem and the top result is an ubuntuforums post with zero replies. :(
[23:28] <rapha> Hi all!
[23:28] <rapha> Regarding the dovecot-postfix meta-package ... what would the proper way to set that up with MySQL be?