[00:18] <Ursinha> who takes care of uvirtbot?
[00:18] <rsalveti> does anyone knows who maintains uvirtbot?
[00:18] <rsalveti> :-)
[00:18] <Ursinha> rá
[00:18] <rsalveti> just wanted to check the code
[00:18] <Ursinha> it's supybot rsalveti
[00:20] <rsalveti> Ursinha: that I know, just wanted to know where is the code of the specific plugin that probes for bugs
[00:21] <rsalveti> seems soren is the contact guy for this bot
[00:22] <rww> what does uvirtbot do that ubottu doesn't do?
[00:22] <Ursinha> rww: shows here when new bugs are filed
[00:22] <rww> ah, shiny
[00:27] <Ursinha> uvirtbot: hi
[00:27] <Ursinha> uvirtbot: owner
[00:28] <Ursinha> meh
[00:41] <zroysch> is it possible to mount and use the data from just one of the partitions in an mdadm raid-1 ?
[00:43] <zroysch> i mean. one of the drives in the raid-1 has failed. one is remaining. the mdadm device is /dev/md0 but I am unable to mount it.
[00:48] <patdk-lap> just tell mdadm to mount it in failed mode
[01:25] <fluvvell> I'm running dhcpd and several client machines on the network are dallying over receiving an IP address. One such incident was at 4am this morning, a cycle of DHCPINFORM and DHCPACK of at least 4 machines.
[01:26] <twb> ITYM "delaying"
[01:26] <twb> Dallying is what you do with a girl in the back of a car
[01:27] <twb> fluvvell: are you nearly out of leases?  If you tcpdump, do you see floods of DHCPREQUESTs for a single response?
[01:28] <twb> I should say: do you see floods of DHCPREQUESTs at all
[01:33] <fluvvell> twb, I've expanded the leases recently, but might need to grow a few more. Not lots of DHCPREQUESTS but I increased by 20 last term and solved some of the issues.
[01:33] <fluvvell> some machines ask twice or 3 times for a lease.
[01:53] <hallyn> SpamapS: don't suppose you're hanging around for a late night?
[02:16] <FIXEDMYNAS> UBUNTU SERVER SAVED MY DAY!
[03:00] <rokr1> hi there !
[03:01] <rokr1> I have just installed kernel 3.0-oneiric
[03:01] <rokr1> The kernel seems to support dom0
[03:01] <rokr1> is there any binary of xen 4.1 in ubuntu
[03:01] <rokr1> natty
[03:02] <rokr1> ?
[03:02] <rokr1> I am running desktop 64bit
[03:02] <rokr1> kernel seems to be generic
[03:02] <twb> Ubuntu doesn't support Xen at all
[03:02] <twb> Everybody has to use kvm or else
[03:03] <rokr1> then why is dom0 support included in ubuntu version of kernel ?
[03:03] <rokr1> if xen is not supported
[03:03] <rokr1> It seems like dependency issue in repository
[03:04] <rokr1> but will it work if I build from XEN-tools source...?
[03:05] <rokr1> is there any reason that ubuntu keeps away from XEN ?
[03:06] <twb> rokr1: because dom0 support is built into 3.0 kernels upstream
[03:06] <twb> I assume they dropped support for everything but kvm because it makes their support contracts cheaper, but it could be just to annoy me
[03:07] <twb> They dropped openvz and vserver from universe in lucid, despite LXC not really being ready to replace them as at .32
[03:07] <KM0201> twb: how would ubuntu server run from an 8gig thumb drive (all i'm doing with it, using 2, 2TB drives as a NAS)
[03:07] <twb> KM0201: in the naïve case, the same as booting from a hard disk.
[03:07] <KM0201> that didn't sound very encouraging..lol
[03:07] <twb> KM0201: if you wanted to get fancy you could do SSD-style optimizations, like not putting swap on there
[03:08] <rokr1> use unetbootin KM0201
[03:08] <rokr1> and change BIOS settings to boot from USB
[03:08] <twb> rokr1: he wants to boot off the USB key, not boot an *installer* off it
[03:08] <KM0201> rokr1: putting it on the USB is not the problem.. i know how to do that...
[03:08] <rokr1> use Windows Version
[03:08] <rokr1> of Unetbootin
[03:08] <KM0201> i'm more curious about performance
[03:08] <twb> And in any case, unetbootin is for retards who are ignorant of both the Ubuntu-provided USB boot images, and isohybrid.
[03:08] <rokr1> performance depends on the I/O
[03:08] <KM0201> considering the USB will almost never be accessed/written to, i can't imagine it being much of an issue.
[03:09] <KM0201> twb: i kinda agree w/ that..
[03:09] <twb> KM0201: FWIW I ship systems that boot from a stock ubuntu install onto USB keys
[03:09] <twb> KM0201: they haven't, you know, exploded or anything
[03:09] <twb> rokr1: http://archive.ubuntu.com/ubuntu/dists/lucid/main/installer-amd64/current/images/netboot/boot.img.gz
[03:10] <twb> None of this Qt GUI wank
[03:10] <KM0201> twb: yeah... one final question... I probably won't roll this out onto my server till end of this week, but.. how difficult would it be to install server to SDA(usb drive), then put SDB and SDC(internal 2tb drives) into a software raid 1?
[03:11] <twb> KM0201: if the array is only used for (say) /srv or /opt, it's trivial
[03:11] <twb> KM0201: if you put /boot or / on the array, you need to do a little dancing, which is why it's easiest to set it up at install time from the debian-installer UI
[03:12] <KM0201> twb: nope, they're just two big storage drives thats it... right now they're mounted at /media/drive 1, and media/drive 2
[03:12] <KM0201> the "OS" partitions, will be on the 8gig usb.
[03:13] <rokr1> KM0201 do you mean to have a persistent then use uuid
[03:13] <rokr1> for drive
[03:13] <rokr1> that will solve the issue
[03:13] <twb> KM0201: incidentally, it would be a good idea to back up the USB's filesystem onto the RAID array
[03:13] <twb> KM0201: in case e.g. someone snaps the USB key
[03:13] <KM0201> twb: i'm not concerned about that, this sits on the top shelf of a closet.
[03:14] <twb> I still advise you to do it
[03:14] <KM0201> rokr1: i'm not concerned about the USB drive install, i can do that, not an issue, i'm curious about Raid1 and the internal drives, would that be possible.
[03:14] <twb> You'll look bloody silly if you need to rebuild the USB key in a hurry from scratch, because you didn't and something goes wrong
[03:15] <KM0201> twb: i'll take your advice under consideration on that, or maybe I'll just do a full copy of the drive onto another thumb drive, and store that thumb drive somewhere..lol
[03:15] <KM0201> twb: so again, my question, is how difficult would that be?..
[03:15] <rokr1> I didnt say about USB at all
[03:15] <KM0201> rokr1: you're talking about a persistant install, my question has nothing to do w/ installing
[03:16] <KM0201> i know how to install ubuntu server to a pen drive.
[03:16] <KM0201> my question, has to do w/ how difficult will it be to make sdb and sdc, part of a raid 1, so they mirror each other
[03:16] <twb> KM0201: I would just have a nightly cronjob that makes a cpio archive of the USB key or similar
[03:16] <twb> I already said the RAID part will be trivial
[03:17] <KM0201> twb: you said it would be trivial if i'm keeping various folders on the hard drive, i'm not.. it's just data (movies, music, etc.)
[03:18] <rokr1> :) never ending
[03:18] <KM0201> ..
[03:18] <KM0201> everybody has either discusssed my usb drive, or a problem that is not going to exist on my system, but not really answered the question, has nothing to do w/ "never ending"
[03:19] <rokr1> okay lets make it simple KM0201
[03:19] <KM0201> it has nothing to do w/ simplicity, you're answering my question, w/ answers that do not apply to my question
[03:19] <rokr1> you need 2 drives sdb and sdc as Linux software raid 1
[03:19] <rokr1> right ?
[03:19] <KM0201> yes
[03:20] <twb> rokr1: he has his answer.  Let him try it; just ignore him until he does.
[03:20] <rokr1> do u have data on existing HDD sdb and sdc ?
[03:20] <KM0201> not right now, no
[03:20] <rokr1> okay use fdisk and format it to raid
[03:20] <KM0201> twb: and no, i don't have my answer, all you've discussed is either backing up my usb drive, or keeping other files on the two large drives, which i am not interested in doing.
[03:20] <rokr1> so MBR is re written
[03:21] <KM0201> ok.
[03:21] <rokr1> then use mdadm to create software raid
[03:21] <KM0201> ok
[03:21] <rokr1> to make it boot as normal raid 1 next time
[03:22] <KM0201> right.
[03:22] <KM0201> the internal drives though, don't boot anyway, (just the external)..
[03:22] <rokr1> no I did not mean that
[03:23] <rokr1> I mean Linux sees the array as a logical drive
[03:23] <rokr1> 2 => 1
[03:23] <KM0201> right... i understand
[03:23] <KM0201> sounds simple enough
[03:23] <KM0201> after what i've been through this past week, i think i'll be able to do it no prob, but i'll do a test run
[03:24] <rokr1> sorry my gnome crashed
[03:25] <rokr1> you there?
[03:27] <rokr1> just have a look at this http://tldp.org/HOWTO/Software-RAID-HOWTO.html
[03:27] <rokr1> bye got to restart
[03:33] <KM0201> rokr1: having probs?
[03:33] <rokr1> yes
[03:33] <KM0201> sorry to hear
[03:34] <rokr1> Just updated to kernel 3.0-oneiric
[03:34] <rokr1> just a stability issue
[03:34] <rokr1> still on the same kernel
[03:34] <rokr1> so did my last message help you ?
[03:35] <KM0201> yeah, i'm not "live" on my server yet.. but.. i've got everything running exactly how i want it (just on a much smaller scale) in vbox, and have been taking serious notes while setting all this up.
[03:35] <KM0201> right now, it seems to be working perfectly.
[03:35] <rokr1> :)
[03:36] <rokr1> Hardware RAID is faster than SoftRAID
[03:36] <rokr1> so y do u want to be with software raid ?
[03:37] <KM0201> well, mainly, cuz i don't have a RAID card..lol, but i was looking at some the other day, i thought they were super expensive, but they aren't at all, i'll probably pick one up before I take this thing "live"
[03:37] <KM0201> how do you configure a "hardware" raid though?
[03:37] <KM0201> i was gonna do some googling on that tomorrow.
[03:37] <rokr1> Yes but I would prefer softraid
[03:37] <KM0201> then why are you asking me why i'd prefer softraid?
[03:38] <KM0201> i'd like a hardware raid, but.. i'm not deadset on it.
[03:39] <rokr1> newer mobos support raid 0 + 1
[03:39] <rokr1> etc
[03:39] <KM0201> yeah, my mobo.. it's kinda old.
[03:40] <rokr1> those are not true hardware raid
[03:40] <rokr1> like intel matrix storage
[03:40] <rokr1> etc
[03:41] <KM0201> hmm, do you know much about rsync?
[03:41] <rokr1> never used it ...! but planning to learn to implement backuppc
[03:42] <KM0201> i'm syncing "folder a" from drive a, to drive b... say I synced it yesterday, then today, i deleted a file in "folder a" on drive a, i want it so that when i sync that folder to drive b, that file also gets deleted
[03:42] <rokr1> not really a case
[03:42] <rokr1> its like RAID 1
[03:42] <KM0201> yeah.
[03:43] <KM0201> i'll just set up a raid... i've got great notes on getting this system exactly back to how i have it now..
[03:43] <rokr1> Full backup + incremental = Delta backup, I prefer it
[03:43] <KM0201> is that in the repos?
[03:43] <rokr1> not really
[03:43] <rokr1> its just a concept
[03:43] <rokr1> of backup
[03:44] <KM0201> oh
[03:44] <rokr1> best thing is to use amanda backup tool or backuppc
[03:44] <rokr1> they have ready made scripts which may suit your needs
[03:44] <rokr1> Auto backup
[03:44] <KM0201> HMM
[03:45] <rokr1> yes based on cron
[03:45] <KM0201> i know nothing about cron
[03:45] <rokr1> its just like scheduler in windows
[03:45] <rokr1> runs app in a regular interval
[03:45] <KM0201> so I could set it to run like, daily at 4am
[03:46] <rokr1> Yes
[03:46] <rokr1> also like every 5 mins do this do that
[03:46] <rokr1> etc
[03:46] <rokr1> wikipedia cron to learn more
[03:47] <rokr1> good thing about backuppc is that it has web GUI
[03:47] <rokr1> and uses samba shares to backup
[03:48] <rokr1> modified version of backuppc is available with zamanda website which includes the ftp support
[03:48] <rokr1> But I dont say rsync is not good
[03:48] <rokr1> its just the way you use
[03:49] <KM0201> right
[03:50] <rokr1> rsync is a base app for backup
[03:50] <rokr1> like xcopy
[03:50] <rokr1> in windows
[03:50] <KM0201> i'm gonna try a software raid.
[03:51] <rokr1> good luck
[03:52] <rokr1> steps are simple 1. format HDD to Linux RAID using Fdisk 2. use mdadm to create raid drive like /dev/md0
[03:52] <KM0201> just gonna do it on a clean install, so it should be easy enough.
[03:52] <KM0201> here we go.. :)
[03:53] <rokr1> also remember chunck size that really improves or degrades the performance of the raid
[03:53] <rokr1> have a eye on it
[03:53] <KM0201> what do you meann "chunk size"
[03:54] <rokr1> its just a small portion of space utilized by raid array to have its parity bits ...! it doesnt matter for RAID1 but it does for RAID 0
[03:55] <rokr1> you can also do RAID5 on 2 disk with mdadm
[03:56] <KM0201> got it installing now...
[03:56] <KM0201> raid1.
[03:56] <rokr1> :)
[03:57] <KM0201> shouldn't take to long, my drives on vbox are small (5gig OS drive, 10gig "storage" drives)
[03:57] <KM0201> only thing I'm not 100% sure on, is how to mount the raid in Samba
[03:57] <rokr1> just use the mounted path
[03:57] <KM0201> actually, i'm not sure how to mount the raid period...lol, so i'll have to figure tha tout.
[03:58] <KM0201> where are they usually mounted by default?
[03:58] <rokr1> its simple
[03:58] <rokr1> you are combining /dev/sdb and /dev/sdc as a singke /dev/md0
[03:58] <KM0201> ok..
[03:58] <rokr1> just mount /dev/md0
[03:58] <KM0201> oh ok.
[03:58] <rokr1> that will do
[03:58] <KM0201> then just add that to Samba.
[03:58] <KM0201> .conf
[03:59] <rokr1> :)
[03:59] <KM0201> and fstab
[03:59] <rokr1> no its raidtab now
[03:59] <KM0201> oh ok.
[03:59] <rokr1> yes also fstab
[03:59] <KM0201> so do i have to add it to fstab, or raidtab?
[04:00] <rokr1> not really look for the instructions in the link which I gave you
[04:00] <rokr1> do not touch raidtab
[04:00] <rokr1> it contains the instructions which you gave in cli
[04:01] <rokr1> if you donot know any config then use cli
[04:01] <rokr1> just mount resulting raid device in fstab
[04:01] <rokr1> simple
[04:02] <rokr1> okay community coming back to my problem
[04:02] <rokr1> with XEN on UBUNTU
[04:02] <rokr1> will try to install XEN from Source
[04:04] <KM0201> what filesystem is on RAID drives?
[04:06] <rokr1> its something like LINUX
[04:07] <rokr1> look for it in instructions
[04:41] <pythonirc101> is there any tool in ubuntu that lets me administer 10 or more fedora machines using one shell? I dont want to ssh to each and execute commands. I was tyring mussh, any other recommendations?
[04:42] <SuperLag> How do you expect to administer them, if you're not connected to them? :)
[04:45] <twb> pythonirc101: puppet
[04:46] <twb> SuperLag: he's asking for a SIMO multiplexer for his ssh sessions, but really he wants change management infrastructure.
[04:46] <twb> Unless, perhaps, his nodes are strictly homogeneous, in which case he wants more like a cloud SOE.
[04:48] <pythonirc101> twb: my nodes are homogeneous
[04:48] <twb> lucky bastard
[04:49] <pythonirc101> twb: I would prefer not to write puppet scripts...
[04:49] <twb> I hear ya, buddy
[04:49] <pythonirc101> If I worked that hard, i wud write it in python
[04:49] <pythonirc101> any other suggestions?
[04:50] <twb> There are things that do what you want, but I'm not familiar with any of them
[04:51] <pythonirc101> I tried mussh
[04:51] <pythonirc101> I would prefer something more user friendly
[05:24] <KM0201> twb: got it working... i've got Server running of a 6gig virtual drive, virtual drive b and c(10gigs each) are set up in Raid 1, and I can access it through Samba.
[05:25] <KM0201> twb`:  got it working... i've got Server running of a 6gig virtual drive, virtual drive b and c(10gigs each) are set up in Raid 1, and I can access it through Samba.
[05:25] <KM0201> ive got all my notes, i should be able to roll this out on my server this weekend w/ little fuss.
[06:04] <args[0]> hey, i installed vsftpd on my ubuntu server, edited the .conf file but can't connect, always getting: 530 Login incorrect. The user I'm using is 'root', is it illegal to connect to ftp using root? thanks
[06:04] <twb> args[0]: vsftpd only allows anonymous access by default.
[06:05] <args[0]> twb: i disabled that
[06:05] <twb> If you want to upload files as root, you should use SFTP
[06:05] <args[0]> so i should install sftp?
[06:05] <twb> It's built into SSH
[06:06] <args[0]> i need ftp, i dont want encryption/decryption
[06:06] <args[0]> just pure old ftp
[06:06] <twb> Well, IMO that's bloody stupid, and since I don't know much about vsftpd anymore, I can't help you do that in any case.
[06:07] <args[0]> i am transferring huge files, enc/dec consumes lots of CPU usage and slows down transfers
[06:07] <args[0]> still bloody stupid?>
[06:08] <twb> Yes; doing that on a trusted network, you should use netcat
[06:09] <twb> Although I am not convinced the encryption and decryption overhead are significant; it is very likely to be I/O bound unless your systems are Pentium II vintage, or embedded.
[06:09] <args[0]> i did some research on that, sftp < ftp when it comes to transfers of huge files
[06:10] <args[0]> varies up to 30% in speed
[06:10] <twb> source# nc -l -p 12345 < /dev/sda
[06:10] <twb> sink#   nc -w3 source 12345 > /dev/sda
[06:11] <twb> Can't remember if that's traditional or OpenBSD netcat, but the difference is minimal; an extra -q 0 or so.
[06:12] <args[0]> thanks for your input
[06:13] <greppy> args[0]: doing things over the network as root is generally frowned upon, ftp as another user or use a different transfer method.
[06:15] <twb> That, too.
[06:15] <args[0]> greppy: i've created new users, but still getting 530 Login incorrect
[06:16] <args[0]> not the first time i use vsftpd though, not sure what's happening
[06:17] <greppy> just on a hunch, sudo /etc/init.d/vsftpd stop; sudo /etc/init.d/vsftpd start
[06:17] <greppy> then try to ftp as one of those new users again.
[06:17] <twb> greppy: as opposed to force-reload?
[06:17] <twb> Looking at the logs on the server side would obviously be a good idea, too.
[06:17] <greppy> twb: I'm old fashioned :)
[06:18] <args[0]> this is my /etc/vstfpd.conf file: http://paste.ubuntu.com/657683/
[06:18] <twb> By design, a client is rarely told *why* it was denied access.
[06:18] <greppy> that was my next suggestion, check the logs, see what errors, if any are being given.
[06:18] <greppy>  /var/log/daemon.log and/or /var/log/auth.log
[06:18] <twb> FFS, why does ubuntu pastebin's "download as text" link want me to give a flipping openid
[06:19] <twb> greppy: by default, absolutely everything that passes through syslog will hit either auth.log or messages.
[06:19] <twb> Thus "tail -fn0 /var/log/auth.log /var/log/syslog &"
[06:19] <args[0]> greppy: still, getting 530
[06:19] <twb> Which I write nearly as much as egrep -v '^[[:space:]]*(#|$)' foo.conf
[06:21] <greppy> twb: you haven't setup that egrep as an alias yet? :)
[06:21] <greppy> args[0]: anything in the error logs?
[06:21] <twb> greppy: not on the hundreds of hosts I connect to on an ad-hoc basis, over which I have no authority to reconfigure root's dotfiles
[06:22] <twb> Assuming they're even smart enough to have a persistent root home directory, and not say a flipping Thecus NAS
[06:22] <greppy> ugh
[06:23] <greppy> to keep the root env "clean" I have setup my own .file to source to have my aliases and stuff set
[06:23] <greppy> for instance, EDITOR=nano drives me a little buggy most of the time.
[06:24] <twb> greppy: yes, in principle I'm allowed to do that but ICBF
[06:24] <twb> It's easier to just memorize a few handy commands.
[06:25] <incidence> Hey, I installed OpenVPN. It works flawleslly, but now I can't connect outside, Like www.google.com goes to 10.0.1.7 (openvpn server)
[06:26] <args[0]> greppy: Aug  3 02:19:33 w00dy vsftpd: pam_unix(vsftpd:auth): authentication failure; logname= uid=0 euid=0 tty=ftp ruser=args rhost=173.178.xxx.xxx  user=args
[06:26] <greppy> due to having to work on a variety of different systems, where the versions of tools may or may not be the same, I end up writing scripts/aliases to be as portable as possible.
[06:28] <twb> incidence: so, your DNS resolver is buggered?
[06:29] <args[0]> just created another new user right now, and still getting 530 Login incorrect
[06:29] <twb> greppy: I gave up on that; now I target bash 3.0+ and GNU coreutils 6+, and anyone on somethign stupider can pay me to workaround their box
[06:29] <args[0]> what's a good alternative for vsftpd?
[06:29] <twb> Like that thecus box, which had xargs -0, but not find -print0
[06:32] <greppy> twb: I get to play on debian, ubuntu, cygwin and sun :)
[06:32] <twb> Poor bastard
[06:32] <twb> If I get lumped with the latter two, I hand them off to an intern
[06:32] <twb> FSVO cygwin = windows
[06:33] <greppy> it's lead to me learning a bit more about different utils, or writing perl5 for something instead of using grep/sed/awk
[06:33] <twb> Non-GNU userlands are a joke
[06:47] <args[0]> off to sleep, goodnight
[06:48]  * args[0] ZZzzzzzzzzzz
[07:46] <pcnerd> Hi, has anyone upgraded from 9.04->9.10 and beyond given both are now EOL upgrades?
[07:47] <_ruben> use old-archives.ubuntu.com to fetch updates and upgrades
[07:51] <pcnerd> I've followed both methods listed here (http://mreschke.com/topic/254/Ubuntu+End+of+Life+and+Upgrades)  and both times I receive either no upgrade or "not supported for jaunty to lucid" etc... (++ followed EOL upgrades from ubuntu documentation
[07:53] <Jeeves_> pcnerd: You shouldn't skip karmic
[07:54] <pcnerd> I see, so how cna I tell it not to skip ?
[07:54] <Jeeves_> You just change the values in /etc/apt/sources.list
[07:54] <Jeeves_> type apt-get update
[07:54] <Jeeves_> type apt-get dist-upgrade
[07:55] <pcnerd> 0 upgraded, 0 newly installed 0 to remove adn 0 not upgraded.
[07:55] <RoyK> do-release-upgrade
[07:56] <pcnerd> same, jaunty to lucid not supported.
[07:56] <pcnerd> its dl'ing lucid.tar.gz so clearly somewhere lucid is set to the next version :'(
[07:58] <RoyK> pcnerd: in /etc/update-manager/release-upgrades, what's Prompt set to?
[07:59] <pcnerd> have tried normal and default
[07:59] <pcnerd> currently set to default
[08:02] <pcnerd> and have jsut tried normal as well with the same result.
[08:07] <pcnerd> RoyK:  Should have clarified, thats Prompt set to default and normal.. currently at default.
[09:21] <alshain> hi, I have installed ubuntu server last week. Now when I boot my computer, caps lock and scroll-lock are flashing :S I've read that this indicates "kernel panic" and can be caused by faulty hardware. How would I best track down the issue?
[09:22] <shauno> can you hook a screen up to it for more hints?
[09:23] <alshain> shauno: yes, I did. I don't get _any_ output after GRUB
[09:24] <alshain> also, it seems to hang in the GRUB part for a considerable amount of time, >5 seconds after Ubuntu Server has been selected
[09:25] <alshain> also, upon previous boots, I never had any output
[09:26] <shauno> I'd be tempted to edit the command line on grub to make sure the word 'quiet' doesn't appear in it.  otherwise, it's been far too long since I've done bare-metal boots :/
[09:26] <alshain> ok, I can try that
[09:29] <alshain> now I get "alloc magic is broken at <address>", this message is still in GRUB
[09:30] <alshain> it says: "press any key to exit" <-- that doesn't work
[09:33] <alshain> I'm now running memtest for the time being...
[09:33] <shauno> not heard that one before, but atleast we found you some google-fodder.  hopefully you can find someone who's a little more intimate with grub
[09:36] <alshain> I wonder GRUB would work as long as I don't edit any commands...
[12:20] <Narc> Hey everyone. I realized yesterday, when I wanted to access the website, that my 10.04 box (VPS XEN hosting) was unresponsive. No ssh access so I did a restart via the hosting web interface. I then realized that the logs (auth.log, syslog...) stopped suddenly 15 days ago. No intrusion as far as I can tell but I'm not sure. I'm puzzled. syslog shows "rsyslogd: [origin software="rsyslogd" swVersion="4.2.0" x-pid="2143" x-inf
[12:20] <Narc> o="http://www.rsyslog.com"] rsyslogd was HUPed, type 'lightweight' and stops logging a few hours later. Any advice anyone ? Thanks
[12:21] <patdk-wk> did it go out of diskspace?
[12:22] <Narc> No, I'm 3GB used on 8GB allowed.
[12:23] <patdk-wk> yes, but what where you back then?
[12:24] <patdk-wk> I have had systems fill up with like ssh attacks, and the log fill the drive, then logrotation cleans it up later
[12:25] <Narc> Interresting. I'm going to check that. Thanks.
[12:25] <Ynodde> join #ubuntu-nl
[12:26] <Ynodde> oops
[12:26] <patdk-wk> don't wanna
[12:27] <Ynodde> gotta use them / signs first i know, soz for spam
[12:28] <wtham> hiya
[12:30] <Narc> patdk-wk:  Apparently, probe says no disk usage problem.
[12:32] <patdk-wk> hmm
[12:33] <patdk-wk> I dunno if dmesg logs go back forenough, maybe rsyslog crashed?
[12:35] <Narc> Maybe, but I think the whole system crashed, because apache was down and sshd too. I just wanted to be sure. Paranoia :D
[12:36] <patdk-wk> oh, if that much was down
[12:36] <patdk-wk> sounds more like OOM
[12:36] <patdk-wk> when you go out of memory, the kernel starts killing random programs
[12:37] <Narc> Oh... I didn't think about that. 256MB. Possible.
[13:24] <StucKman> I have problems with upstart and two /etc/init scripts. there´s a bug report for one of those, rsyslog, here: https://bugs.launchpad.net/ubuntu/+source/rsyslog/+bug/498531 , but if you look closer, you´ll find several similar errors: https://launchpad.net/+search?field.text=upstart+start+uknown+job . any ideas how to fix it?
[13:44] <Tommy_nmw> hi
[13:45] <Tommy_nmw> how can we copy image of one server to another?
[13:53] <StucKman> Tommy_nmw: you mean, all the files?
[13:55] <Tommy_nmw> yes
[13:56] <Tommy_nmw> StucKman: yes. I have test server at work and which is installed with most packages. I want to create image of that and deploy on my home server where there is no internet. my home server is just not installed with ubuntu
[13:56] <StucKman> rsync can be your friend, or tar
[13:56] <StucKman> ah, hmm
[13:57] <StucKman> well, either you just rsync/tar the files to your server and then do the grub instalation by hand
[13:57] <StucKman> or you install ubuntu and then rsync/tar the user and config files
[13:59] <Tommy_nmw> StucKman: how to rsync ?
[14:04] <StucKman> Tommy_nmw: I think you will need a intermediate media for transporting the files
[14:05] <Tommy_nmw> StucKman: live Ubuntu CD of any version?
[14:08] <jdevel> hey, hope all are doing well.
[14:09] <jdevel> does anyone here have a fair amount of experience with mail servers?
[14:09] <patdk-wk> !ask
[14:09] <jdevel> I'm looking to build one out and wanted to get some opinions
[14:09] <jdevel> I'm of course running this on ubuntu
[14:10] <jdevel> 10.04 server, and I'm thinking about using Postfix to take care of things
[14:10] <patdk-wk> do you have a good reason why you need your own email server?
[14:11] <patdk-wk> cause getting other people to accept email from you, normally is not an easy task
[14:11] <jdevel> because I don't want to pay a provider for it
[14:11] <jdevel> ahh
[14:11] <jdevel> that is a great point
[14:13] <patdk-wk> email is one thing that is unlike anything else a server does, cause it's blocked on so many levels :)
[14:14] <jdevel> yea.. it's true.  I might just forge ahead on a virtual machine and see how things go.. at least as a learning experience
[14:15] <jdevel> use one of my more obscure low traffic domains
[14:15] <jdevel> or no-traffic
[14:15]  * patdk-wk perfers to use postfix + dovecot + amavisd-new
[14:15] <patdk-wk> but then, I'm using versions that aren't in ubuntu yet
[14:16] <jdevel> yea the prototype virtual machine I have setup for servers is U10.04
[14:16] <patdk-wk> mine are 10.04 too, but just using much newer postfix+dovecot+amavis
[14:17] <jdevel> amavis, is that spam/virus management?
[14:17] <jdevel> i see many setups using clamav and spamassassin
[14:17] <Jeeves_> jdevel: It scan's and takes action based on configuration
[14:17] <patdk-wk> it's more of a wrapper, to help intergrade virus and spamfiltering
[14:18] <jdevel> like ufw to iptables
[14:18] <Jeeves_> No
[14:18] <jdevel> k
[14:18] <Jeeves_> It does a lot more
[14:18] <jdevel> I will certainly have to take a look
[14:18] <huats> does anyone have an made some experiment with glusterfs ? well I was wondering if it could be used to "share" files automatically  over a few computers
[14:19] <jdevel> why not just use nfs or something?
[14:20] <huats> jdevel, I'd like to have an offline option and it was my understanding that I could do that with gluster
[14:22] <patdk-wk> heh, glusterfs documentation is sparse :(
[14:29] <jdevel> huats, I'm not sure what you mean by offline..
[14:29] <jdevel> even in a cluster you need the machines to be up and connected..
[14:30] <huats> jdevel,  I am just saying non sense :)
[14:30] <patdk-wk> he wants a cluster of servers, if anyone dies, you can still access all data
[14:30] <huats> I had made wrong assumptions on the gluster role :)
[14:30] <huats> (I have continued to look at some docs since my question)
[14:30] <jdevel> you could have an nfs and script something to mirror the files on each machine
[14:31] <jdevel> i bet there is something else out there that would handle what you want to do
[14:31] <patdk-wk> unisom :)
[14:43] <huats> patdk-wk unison was our former solution but it gets very very confusing with more than 4 people :)
[14:43] <huats> thanks anyway
[14:45] <patdk-wk> I only used it up to 3 servers at a single time
[14:45] <patdk-wk> and I had it running every 28seconds
[15:01] <zul> NCommander: PING
[15:02] <Ursinha> Daviey!
[15:02] <Daviey> Ursinha: !!
[15:06] <Tommy_nmw> hi
[15:08] <zul> hallyn: ping
[15:11] <hallyn> zul: yo
[15:11] <zul> hallyn: i just checked the omap4 kernel and the lxc is not enabled :(
[15:11] <hallyn> I'm just having no end of tiny little nuisance troubles :)  offlineimap won't work; natty ideapad takes 10 secs to pull up xterm; sigh
[15:12] <hallyn> oh look, mutt just crashed
[15:12] <hallyn> zul: which parts are not enabled?
[15:12] <zul> cgroups looks like it im going through it now
[15:14] <Daviey> hallyn: offlineimap works for me here :/
[15:14] <Daviey> hallyn: using mutt-patched ?
[15:15] <hallyn> wtf is mutt-patched
[15:15] <hallyn> no i guess i'm using mutt 1.5.21-5
[15:15] <hallyn> offlineimap was just copying back my local copies of msgs back to imap!  So I kept getting more and more dups
[15:16] <hallyn> thank god for mut's 'D~='  :)
[15:16] <hallyn> Daviey: oh, mutt-patched *only* adds the sidebar?  no correctness fixes?
[15:17] <hallyn> heh, if i want evolution i'll run evolution :)
[15:18] <Daviey> hallyn: heh, yes - i just wanted to compare fail.
[15:18] <hallyn> Daviey: it was a segfault.  ulimit was default, so no coredump :(
[15:19] <Daviey> ah, i had to raise my ulimit due to thunderbird sucking.
[15:26] <alonswartz> Hi folks. What kernel package is recommend to be used for Xen hosting? In the past (i.e., Hardy) is was linux-image-xen, but that package no longer exists in Lucid+. Is linux-image-virtual the way to go?
[15:27] <smb> alonswartz, for 64bit it would be the server image
[15:28] <alonswartz> smb: linux-image-server, thanks. What about 32bit?
[15:29] <smb> alonswartz, There do not seem to be hypervisor/utils there right now for 32bit. zul you know why?
[15:29] <smb> Otherwise I would use the generic-pae flavour
[15:29] <zul> because you need pae in order to run zen
[15:29] <zul> xen even
[15:30] <smb> zul, Right, does the server install not default to generic-pae?
[15:30]  * smb must admit he has not looked closely yet
[15:30] <Pici> Is linux-image-virtual intended for installs running under a vm?
[15:30] <zul> smb: it should but i havent run 32bit for like ever..
[15:31] <smb> Pici, yes
[15:31] <smb> zul, Hm, maybe something we should think of looking at before release...
[15:31] <alonswartz> I've used linux-image-virtual for our VM optimized images (ie. TurnKey Linux), and when building the vmtools they are actually built against linux-image-generic-pae
[15:31] <zul> smb: yeah...throw it on the list ;)
[15:33] <Pici> smb: thanks, the package description isn't particularly detailed.
[15:33] <alonswartz> in the passed we've optimized our images for Xen hosting providers using linux-image-xen, but now for Lucid I'm looking for the best equivalent - at least that is until we add support for 64bit
[15:33] <alonswartz> which will hopefully be in time for the next LTS
[15:34] <alonswartz> but in the meantime, for 32bit Xen, which kernel is recommended? zul, smb?
[15:34] <smb> alonswartz, Well, tbh Hardy is the only current release usable for hosting xen. And yes, we try to get it right for the next lts
[15:35] <alonswartz> smb: just to be clear, I mean as the guest, not the host
[15:35] <smb> alonswartz, Ah, well as guest either generic-pae or the ec2 images
[15:36] <smb> (that is for lucid)
[15:36] <alonswartz> at the risk of sounding ignorant, what is the difference between linux-image-virtual and linux-image-ec2?
[15:36] <smb> After that the -virtual flavors are meant for guests
[15:37] <smb> alonswartz, In Lucid it is a complete different xen patchset (so not using the codefrom 2.6.32)
[15:40] <alonswartz> smb: so which would you recommend for our use case?
[15:43] <smb> alonswartz, For running as domU I would for Lucid use ec2, for releases after that virtual
[15:44] <alonswartz> smb: excellent, thanks for the help!
[15:46] <alonswartz> smb: quick follow up question. In Hardy there was linux-ubuntu-modules-VER-xen, anything similar for Lucid?
[15:47] <smb> alonswartz, No, there is generally no l-u-m other than hardy
[15:47] <smb> (well anything still supported that is)
[15:48] <alonswartz> smb: ok, thanks
[16:09] <hggdh> folks, I am getting a failure on RAID1 i386
[16:19] <SpamapS> hallyn: what were you up to last night? did you get it sorted? I saw your merge proposal.. will take a look as soon as I get through the morning deluge of email
[16:29] <baffle> Anyone know of any TFTP server that reply on port 69 instead of a random sourceport? I.e. to punch thru firewalls/NAT. I've found "Open TFTP Server" on SF, but it doesn't work.
[16:42] <Ursinha> soren: hi, can you please set the importance/status of bug 809646? as you're the assignee I believe you know how to triage that correctly :)
[16:57] <hggdh> Daviey: do you need to worry about bug 820469?
[16:57]  * SpamapS will be fixing that one
[16:58] <hggdh> SpamapS: you knew about this one?
[16:58] <Daviey> hggdh: i rarely worry
[16:58] <SpamapS> hggdh: saw it
[16:58] <hggdh> heh
[16:58] <SpamapS> hggdh: I'm pretty sure though, that this has been the case since lucid
[16:58] <SpamapS> hggdh: the test case specifically says you may have to re-assemble the arrays
[16:59] <Daviey> Hmm
[16:59] <Daviey> is this a dupe of the one jamespage raised?
[16:59] <SpamapS> hggdh: well, step 16 does
[16:59] <Daviey> which i thought we left in WON'T FIX?
[16:59] <SpamapS> Wait..
[16:59] <SpamapS> There should be no need to add any missing devices back to the RAIDs manually. Otherwise, there is a bug! A manual addition would be:
[16:59] <SpamapS> Did we change the test?
[17:00] <SpamapS> Or have I always been reading it wrong?
[17:00] <hggdh> no, we did not change it
[17:00] <Daviey> Lets put this into a real situation.
[17:00] <hggdh> it _should_ be automagic
[17:00] <Daviey> Server running..
[17:00] <Daviey> disk becomes degraded why?
[17:01] <hggdh> disk fails. You stop the system,, remove it, but you do not have a backup disk to insert. You reboot degraded
[17:01] <Daviey> Then you reinsert the *broken* disk?
[17:01] <Daviey> and expect it to be added automagically?
[17:02]  * SpamapS has seen that scenario play out to sinister consequences
[17:02] <hggdh> later on you find there was a moth on the printed circuit of the failing disk, and reinserts it
[17:02] <hggdh> (after taking off the moth)
[17:03] <Daviey> hggdh: sure thing, but i think it should require re-adding manually... there is clearly an issue that requires a sysadmin to intervene
[17:03] <SpamapS> An admin I worked with saw a failed drive light.. pulled it, pushed it back in.. and said "sometimes they do that, just pull/push it, it goes back to green"
[17:03] <hggdh> you reboot. One md device is recognised. Two are not
[17:03] <hggdh> SpamapS: seen this happen, did it myself
[17:03] <SpamapS> That led to a most awesome multi-drive failure costing the company 1 week of transactional data.
[17:04] <hggdh> heh, not with me ;-)
[17:04] <hggdh> Daviey: why *one* md would be recognised, and two not? Same disk, same partition table
[17:04] <Daviey> bug #791454
[17:04] <SpamapS> I believe this is actually a problem in the test case. I would fully expect that a degraded array would require me to re-add any failed disks.
[17:05]  * SpamapS said he'd fix it because he thoguht it was the other bug
[17:05] <hggdh> SpamapS: we might need to look at it, it never happened with me on amd64
[17:06] <hggdh> but it is the same scenario, indeed. So I will close it dup, and get back to the test
[17:07] <Daviey> it's not like it's hard to re-add to mdad, and booting degraded works.
[17:07] <Daviey> hggdh: Can you update the test case aswell please?
[17:07] <jamespage> Daviey, SpamapS, hggdh: I did have a dig around in the md-ressemble code when I encounted the issue
[17:08] <hggdh> and?
[17:08] <jamespage> I think if was something todo with ext4 filesystem checkpointing not matching - which is why the swap auto-resyncs
[17:08] <jamespage> and the filesystems don't
[17:08] <jamespage> it looked like the behaviour as coded - but that does not mean its a bug
[17:09] <hggdh> oh, this might be it
[17:09] <hggdh> I agree
[17:09] <Daviey> nobody disagrees that the current behaviour is a problem?
[17:09] <hggdh> Daviey: I will update the test with a statement that it will probably fail, and the mds will have to be manually added back
[17:09] <hggdh> I do not disagree
[17:09] <Daviey> hggdh: rocking!
[17:21] <hggdh> Daviey: test instructions & result update
[17:21] <hggdh> d
[17:24] <Daviey> hggdh: thanks!
[17:26] <SpamapS> hallyn: reviewing your libvirt change.. nicely done.. its intended to only stop on *shutdown* or *reboot* right, like, you don't want to shutdown domains if somebody says 'stop libvirt-bin' right?
[17:38] <hallyn> SpamapS: exactly
[17:38] <hallyn> SpamapS: I was a bit torn on runlevel 1, but shutting them down there seems right
[17:38] <hallyn> (so i'm doing it)
[17:42] <SpamapS> hallyn: agreed
[17:43] <SpamapS> hallyn: this is also indepenent enough of a change that I think its 100% SRU'able
[17:53] <hallyn> SpamapS: people will be glad to hear that :)
[17:53] <hallyn> SpamapS: though, if the script goes wrong, it could prevent systems from shutting down
[17:53] <hallyn> (as it did for me when I didn't have the 'break' at end of loop :)
[17:57] <SpamapS> hallyn: no an error won't cancel shutdown
[17:58] <SpamapS> hallyn: oh so it just went forever? ;)
[17:59] <hallyn> :)
[17:59] <hallyn> yup
[18:00] <hallyn> do you think it should sit in oneiric for a week or so before we try to sru?
[18:05] <AceKing> I tried to setup Ubuntu Server 11.04 on one of my PC's. I am trying to set it up where I can share files with family in another state. I installed it, but when it was time to restart it boots to a black screen and just sits there. This is the first time attempting to setup a server, so I'm not familiar with getting it to work
[18:06] <AceKing> Can someone help me to get it up and running?
[18:11] <SpamapS> hallyn: yes
[18:12] <SpamapS> hallyn: libvirt sees enough usage in the platform team (and we all reboot enough) .. we should root out any bugs rather quickly
[18:12] <SpamapS> AceKing: it should have a login screen for you
[18:13] <SpamapS> AceKing: do you get the menu where you can select booting into a recovery console?
[18:13] <AceKing> SpamapS, It just comes up to a flashing prompt.. Nothing written at all
[18:14] <AceKing> SpamapS, I tried reinstalling, because I thought it didn't install correctly, but it's doing the same thing
[18:15] <SpamapS> AceKing: hold down left shift after the BIOS .. no menu?
[18:15] <AceKing> SpamapS, I'll check
[18:17] <AceKing> SpamapS, No, right after the BIOS I held down the left shift and it still went right to the flashing prompt
[18:17] <SpamapS> AceKing: during install, did you install GRUB to the MBR ?
[18:18] <SpamapS> AceKing: also what kind of computer is it?
[18:18] <AceKing> SpamapS, Yes, GRUB is installed
[18:18] <AceKing> SpamapS, It is a Dell
[18:19] <AceKing> SpamapS, Dell XPS 400 to be more specific
[18:19] <SpamapS> hallyn: do you have upload rights to libvirt ?
[18:19] <SpamapS> AceKing: thats very strange, you should be getting a grub menu.
[18:20] <SpamapS> AceKing: just regular SATA disks?
[18:20] <AceKing> SpamapS, Yes
[18:21] <AceKing> SpamapS, I installed it from a USB drive that I setup with UNETbootin. Does that make a difference?
[18:23] <SpamapS> AceKing: shouldn't
[18:23] <SpamapS> AceKing: but its possible that it installed grub to the USB drive instead of to your install disk.. which would be a bug.
[18:24] <pmatulis> SpamapS, AceKing: installing from USB key can be problematic.  GRUB may end up... meh
[18:24] <SpamapS> how to fix that though? drop to a console and manually grub-install ?
[18:24] <AceKing> SpamapS, I am going to burn it to a CD and see if that works
[18:25] <AceKing> pmatulis, How do I do that?
[18:25] <pmatulis> AceKing: can try booting with the key in, see if it comes up
[18:25] <pmatulis> :)
[18:25] <SpamapS> AceKing: if that does work, thats probably worth reporting a bug against the installer
[18:25] <AceKing> pmatulis, I'll give that a shot
[18:25] <AceKing> SpamapS, Ok
[18:26] <SpamapS> https://bugs.launchpad.net/ubuntu/+source/debian-installer/+filebug
[18:28] <AceKing> SpamapS, pmatulis, I tried with the key in, and it went back to the install screen. When I installed it, I just let it run from the Default setting. Should I have tried a different setting?
[18:28] <pmatulis> https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/752694  <--- this seems to pertain more to HP and RAID devices but someone does mention a "dell"
[18:31] <AceKing> Is it a pain in the ass to move over from the key?
[18:31] <AceKing> Or is it worth just reinstalling with the CD?
[18:32] <pmatulis> AceKing: depends if you want to learn something.  if you just want to get it to work i would use a cd
[18:33] <AceKing> pmatulis, I would love to learn how to do that :)
[18:33] <pmatulis> AceKing: go for it
[18:34] <SpamapS> I think the work around is to switch to a console before the first reboot, and grub-install onto the root device
[18:34] <hallyn> SpamapS: I do (have libvirt upload rights)
[18:35] <SpamapS> hallyn: looks good.. I haven't tested it, but I think its likely to suceed. :)
[18:35] <SpamapS> succeed even
[18:36] <AceKing> How do I know if GRUB installed on the key? where would I look?
[18:36] <AceKing> Sorry if my question sounds dumb, this is new to me
[18:36] <SpamapS> AceKing: it would have booted to your new system if it did
[18:36] <SpamapS> I think
[18:37] <SpamapS> AceKing: still its worth reporting as a bug if your CD install works
[18:37] <AceKing> SpamapS, ok, thanks. I think I'm going to try installing from the CD. I will report it.
[18:37] <SpamapS> AceKing: sweet
[18:38] <AceKing> SpamapS, I'll problably be back in to figure out how to setup a file server if I get too confused.
[18:38] <hallyn> SpamapS: thanks for looking.  will push it.
[18:40] <AceKing> Thank you both helping me!
[18:44] <NCommander> zul: pong
[18:44] <zul> NCommander: i forgot what i was going to ask
[18:45] <NCommander> zul: thats the problem with contentless pings
[19:07] <hggdh> SpamapS: can you please mark your raid test completed?
[19:13] <sneakyimp> I'm trying to set up an Amazon EC2 compute instance to send postfix mail using Amazon SES mail service.  I do not want any mail (local or otherwise) to be delivered to mailboxes on this machine and want all mail instead to be sent to mydomain.com.  I've got this mostly working.  mail to "root" is properly redirected to root@mydomain.com.  However, root@localhost gets sent to root@localhost.mydomai
[19:13] <sneakyimp> n.com and fails with the message "local delivery is disabled"
[19:14] <sneakyimp> Any postfix pros around?
[19:18] <RoAkSoAx> adam_g: did do you write a puppet ensemble formula?
[19:19] <hggdh> SpamapS: duh. Forget, read it wrong
[19:27] <sneakyimp> I'm sincerely hoping someone can help me with this problem: http://paste.ubuntu.com/658159/
[19:33] <AceKing> SpamapS, are you still here?
[19:34] <SpamapS> AceKing: about to step out for lunch
[19:34] <AceKing> SpamapS, ok, I'll just ask my question in the room.. Thanks
[19:37] <AceKing> I installed Ubuntu Server from the CD. It started up to a command prompt asking for my username, then the password. It then drops to a command prompt. I tried to type startx like I would normally do on Ubuntu, but that obviously did not work. How do I boot to a desktop?
[19:38] <smoser> where is the jenkins server ?
[19:39] <smoser> zul, hggdh ?
[19:39] <smoser> bueller ?
[19:40] <hggdh> smoser: on the internal QA lab (access to it is still restricted, public-facing instance is being built
[19:40] <hggdh> smoser: PM
[19:54] <lifeless> hallyn: hi
[19:54] <lifeless> hallyn: do you know any reason why lxc containers cannot nest? [other than shallow bugs like 'noone has tried it']
[20:02] <AceKing> I am a total beginner with Ubuntu Server. All I am trying to do is setup a simple file server to share large files (video) to family in different states. Is there a tutorial that can tell me how to accomplish this? I installed the server software, and I am LOST to say the least.
[20:04] <JanC> AceKing: share in what way?  (for what purpose?)
[20:05] <alshain> shauno: memtest was running all day, no error :S I really hoped the faulty, as suggested in some threads as possible cause, well, I'll continue my search tomorrow
[20:06] <AceKing> JanC, I converted a lot of home videos, and my family is scattered all over the states. Instead of putting them on DVD's and mailing them out, I want to be able to set it up where I can send them a link and they can download them
[20:09] <JanC> AceKing: home videos, so I guess you want some privacy too...
[20:09] <AceKing> JanC, That's not too important.. A lot of holiday get togethers. It would be boring for anyone else to watch. LOL
[20:11] <patdk-wk> aceking hope you have a crapload of bandwidth
[20:11] <patdk-wk> normal videos are like 1-3mbit upload speeds, for a single user
[20:12] <JanC> patdk-wk: for streaming, you mean?
[20:12] <JanC> not sure that's what AceKing wants?
[20:12] <AceKing> patdk-wk, they all wont be downloading at the same time
[20:12] <patdk-wk> janc, that is what it sounds like to me, make like a youtube site for himself
[20:12] <Bernhard> just set up  a webserver .. so they can donwload the files from your webserver.
[20:12] <patdk-wk> doesn't gallery2 support videos?
[20:13] <Bernhard> apt-get install apache2
[20:13] <AceKing> JanC, I want them to be able to download the vids and burn them, or just play them off their HDD
[20:13] <JanC> if privacy isn't all that important, just serving them over http should be okay
[20:13] <Bernhard> apt-get install apache2
[20:14] <AceKing> JanC, I'm not even sure where to begin. At the risk of sounding stupid, I thought after install it would boot to a desktop
[20:15] <patdk-wk> server edition? to have a desktop?
[20:15] <JanC> AceKing: if you want a desktop, install the desktop version
[20:15] <JanC> you can use it as a server too
[20:15] <patdk-wk> desktop version isn't going really help you setup and configure this though
[20:15] <AceKing> JanC, You mean Ubuntu can be used as a server? That is what I am on right now
[20:16] <JanC> patdk-wk: there is some GUI tool to configure apache IIRC
[20:16]  * patdk-wk runs scared
[20:17] <AceKing> patdk-wk, like I said, I'm new to this and don't have a clue
[20:17] <Pici> AceKing: They use the same package repositories. 99.9% of the things you'd do on a 'server' you can also do on the desktop, although there normally aren't graphical configuration tools.
[20:18] <AceKing> Pici, OK, is there somewhere I can find a tutorial to learn how to accomplish what I am trying to do?
[20:18] <Pici> !serverguide
[20:18] <AceKing> Pici, thanks!
[20:18] <JanC> patdk-wk: hm, I thought there was a gadmin-* tool for it, but apparently not  ☺
[20:19]  * patdk-wk would be supper lazy
[20:19] <patdk-wk> install gallery2 for apache/php
[20:19] <patdk-wk> and just setup a password protected folder in gallery2
[20:19] <patdk-wk> then upload videos to it
[20:19] <patdk-wk> since it's all webased, would be easy to use
[20:19] <patdk-wk> even for other people to load up stuff
[20:20] <AceKing> patdk-wk, I know how to install gallery2, but how do I upload the vids?
[20:20] <patdk-wk> you have to configure gallery2
[20:20] <patdk-wk> and make sure you have ffmpeg installed
[20:20] <patdk-wk> you upload a video just like you would upload a picture
[20:21] <patdk-wk> when gallery2 says add picture just take that as, add new object
[20:21] <JanC> so, sort of a private youtube?  ☺
[20:21] <patdk-wk> yep
[20:21] <AceKing> Ahhhh, OK
[20:21] <patdk-wk> http://codex.gallery2.org/Gallery2:Features#Supported_MIME_Types
[20:21]  * JanC never used gallery2
[20:21] <patdk-wk> janc, I have all my kids pics and stuff in it
[20:22] <patdk-wk> also makes it easy to backup/store :)
[20:22] <patdk-wk> also nice that it does all the conversions for me, since I always take raw pics, it converts to jpeg, and auto-rotates them for me :)
[20:23] <AceKing> JanC, patdk-wk, Pici, Thank you for helping!
[21:19] <photon> hi. I'm planning on buying a 3TB drive. will it work with ubuntu, ext4?
[21:19] <Ursinha> zul: hi, are you there by chance?
[21:19] <RoyK> it will
[21:21] <photon> RoyK: will I have to use GPT?
[21:23] <RoyK> photon: to boot from it, I guess yes
[21:23] <RoyK> photon: but for storage, that won't matter
[21:23] <photon> ok
[21:24] <RoyK> just use an old drive or two (in a mirror) for the boot
[21:24] <RoyK> then use the rest for storage
[21:24] <zul> Ursinha: barely
[21:25] <Ursinha> zul: just wondering if you could give some input in bug https://bugs.launchpad.net/ubuntu/+source/backuppc/+bug/782890
[21:26] <Ursinha> it's marked to expire soon and I wasn't sure how to bring your attention to that bug without directly pinging you
[21:26] <Ursinha> (that might be the reason people assign others for them to comment, I guess I get it now Daviey)
[21:35] <hallyn> lifeless: I dunno.  hav eyou tried it?  :)
[21:35] <hallyn> lifeless:  a few things I can think of:
[21:35] <hallyn>  1. our cgroups handling may not be sufficient.  In fact I'm pretty sure it's not
[21:35] <hallyn>  2. the lxc monitor may want access to a device that we don't normally grant access to (which would be easily fixable)
[21:36] <hallyn> It's a known TODO to make lxc more flexible with respect to where it tries to put its cgroups.  So if that's stopping nested containers, we can bump teh priority of that.
[21:37] <lifeless> so conceptually it should be fine ?
[21:37] <lifeless> i.e. just some expected bugs rather than a inherent design issue
[21:38] <hallyn> lifeless: yeah...
[21:38] <hallyn> lifeless: cgroups are nestable, and namespaces are all either nestable or fully isolated...  so there *shoudl* be no problems
[21:38] <lifeless> (this is relevant to ensemble)
[21:50] <Ursinha> zul: thanks
[21:54] <zul> Ursinha-afk: n
[22:00] <hallyn> jdstrand: are you around?  (I *think* he's out this week, just checking...)
[22:08] <hallyn> SpamapS: in that libvirt-bin.upstart...  I'm using RUNLEVEL (and it seems to work), but am I supposed to be declaring 'env RUNLEVEL' at top?  Or is that only to give it a default if it's not defined by the 'stop on' line?
[22:09] <hallyn> (working on a version for lucid right now)
[22:09] <lifeless> hallyn: apt is fixed :)
[22:10] <hallyn> lifeless: woohoo
[22:11] <hallyn> I wonder if I should then remove that new package requirement...
[22:11] <lifeless> less cruft is less cruft
[22:11] <hallyn> true.  i'll remove it when i add your lxc-start-aufs script
[22:11] <hallyn> any ideas what i should call that?
[22:12] <hallyn> lxc-start-once?  lxc-start-ephemeral?
[22:12] <SpamapS> hallyn: I don't believe script sections are run with set -u .. so you should be ok
[22:14] <hallyn> lifeless: (i don't want to do -afs bc I'd like to reuse it with -overlayfs)
[22:14] <lifeless> hallyn: ephemeral sounds good
[22:15] <hallyn> a bit long for people whose tab key doesn't work, maybe
[22:15] <lifeless> now, I need to debug why postgresql isn't starting with ...-aufs - may be more shutdown-race damage
[22:15] <lifeless> hallyn: its mostly going to be inside other scripts that it gets used.
[22:15] <lifeless> hallyn: what might be really neat is a no-baseline version of it, for hard-core-ephemeral, run-until-reboot, no-trace-left kindof stuff.
[22:16] <lifeless> but thats probably you-ain't-gonna-need-it territory
[22:16] <SpamapS> hallyn: oh, and to be clear, you're covered using RUNLEVEL on the stop on condition because rc exports it.
[22:17] <hallyn> SpamapS: thx
[22:18] <lifeless> hallyn: so you've mentioned poweroff support
[22:18] <lifeless> hallyn: how deep does that go ?
[22:18] <lifeless> brb
[22:18] <hallyn> kernel
[22:19] <hallyn> lemme find a url
[22:19] <hallyn> lifeless: here is the middle of a long discussion about kernel support: http://lkml.indiana.edu/hypermail/linux/kernel/1101.1/02730.html
[22:21] <hallyn> lifeless: http://lkml.indiana.edu/hypermail/linux/kernel/1101.1/01973.html   is the problem description
[22:24] <SpamapS> hallyn: hey, looking at /etc/init/cgconfig.conf ... why doesn't it have a stop on ?
[22:26] <adac> Hi guys. I installed denyhosts and added a hostname in hosts.allow but it still get blocked from time to time. any ideas whats wrong? shouldn't it be whitelisted when its in hosts.allow?
[22:27] <hallyn> SpamapS: I guess bc you can safely shut down with cgroups mounted?
[22:28] <SpamapS> hallyn: mk, will leave it alone
[22:28] <hallyn> SpamapS: perhaps we shoudl ask jbernard if he thinks we should make it stop at some point
[22:29] <SpamapS> hallyn: I think the regular late shutdown umount will probably do it, but it won't run 'cgclear' so not sure if thats enough
[22:30] <hallyn> SpamapS: yeah, umount is all it would do anyway i think
[22:30] <hallyn> oh no, i guess cgclear actually moves all tasks to root cgroup first
[22:33] <hallyn> SpamapS: so to be sure, you want me to wait a week on the SRU, or should i just go ahead and request it now?
[22:33] <hallyn> (I'm fine either way)
[22:34] <SpamapS> hallyn: I figured you could wait a week just to see if there's anything fundamentally wrong with it. If you want to request it now, and let that happen in parallel w/ the SRU.. thats fine too
[22:35] <hallyn> ok
[22:35] <hallyn> i guess i'll push the tree to lp:~serge-hallyn and let it sit
[22:36] <Demosthenes> so what's the preferred package for monitoring syslog nowadays?
[22:36] <SpamapS> hallyn: I do think we need to fix it for maverick and natty too.. the upstart job didn't really change in those releases much/at all, so should be a straight accross the board update
[22:36] <SpamapS> Demosthenes: define "monitoring" :)
[22:36] <hallyn> SpamapS: yeah, i'll do them too
[22:37] <Demosthenes> SpamapS: pattern matchign & email
[22:37] <SpamapS> Demosthenes: I used to use 'swatch' a lot.
[22:38] <SpamapS> Demosthenes: but thats more real-time...
[22:38] <SpamapS> Demosthenes: logwatch is good for a daily summary
[22:40] <Demosthenes> i tend to prefer logmuncher, but its not in apt
[22:40] <Demosthenes> logcheck appears to be its closest kin
[22:41] <SpamapS> package it up, we'll sponsor it. :)
[22:41] <Demosthenes> just don't look it up in the urban dictionary ;]
[22:41] <SpamapS> lol.. now I *have* to
[22:41] <SpamapS> err.. on 2nd thought, no thanks
[22:41] <Demosthenes> *laugh*
[22:43] <n2deep> Hi, does anyone know of any System 76 competitors?
[22:45] <jbernard> hallyn: that's the upstart script, no?
[22:47] <jbernard> hallyn: if policy calls for having a stop on, I have no objections
[22:48] <hallyn> jbernard: but is there any point to it?
[22:48] <hallyn> hm, well maybe there is -
[22:48] <hallyn> if you enter runlevel 2, then runlevel 1, then 2 again,
[22:48] <hallyn> cgconfig may fail to restart without first being stopped?
[22:49] <jbernard> if you switch runlevels, you may want that behaviour
[22:49] <hallyn> which?
[22:49] <hallyn> meaning you think there should be a stop-on?
[22:49] <jbernard> proper stopping
[22:49] <hallyn> ok
[22:49] <hallyn> SpamapS: ^
[22:50] <hallyn> so that suggests a simple stop on runlevel [016]
[22:51] <hallyn> jbernard: thx
[22:51] <SpamapS> hallyn: yes
[22:52] <SpamapS> hallyn: tho if you switch to 1, since there's no stop on, it just won't stop.
[22:52] <lifeless> hallyn: oh that reminds me, did you see the need to touch utmp ?
[22:52] <SpamapS> its totally acceptable to have no stop on when you leave no process running.. but if you cause umounts to fail, that could be a problem.
[22:53] <Demosthenes> woot. i have a debian fileserver, 5 years old, with raid5... i just bought a new fileserver running ubuntu LTS, raid6, and was setting it up... when i had my first drive failure on the old one ;]
[22:54] <Demosthenes> talk about timing
[22:54] <Demosthenes> oh, and i must report, that running ubuntu booting from a pair of 16 GB USB keys with raid0, encryption, and LVM works great.
[22:55] <hallyn> lifeless: yes i did, i'll toss that in.
[22:56] <hallyn> SpamapS: not sure what you're saying
[22:56] <lifeless> hallyn: I'm not sure why its not auto created, but it didn't seem to be
[22:56] <hallyn> lifeless: i'm hoping to throw out a new package tonight
[22:57] <lifeless> \o/
[23:04] <hallyn> SpamapS: what would you say would be the best channel on which to ask UDD questions (i.e. about lp:ubuntu/*-updates)
[23:04] <hallyn> there's no ubuntu-udd...  i thought there was...
[23:12] <SpamapS> hallyn: I usually ask in #ubuntu-devel
[23:12] <SpamapS> hallyn: but I feel knowledge on the deeper parts of it is hard to come by
[23:13] <hallyn> SpamapS: ok.  i suppose i jsut need to wait for just the right one of two people to come by and see the q :)
[23:26] <hallyn> lifeless: actually i think i'm going to move the varrun+utmp tweaking into /etc/init/lxcguest.conf.  It doesnt' seem to belong in lxcmount.conf
[23:55] <adam_g> RoAkSoAx: no, i didn't
[23:56] <adam_g> RoAkSoAx: (write a puppet formula)