[01:03] <MTecknology> Where are teh ufw logs at?
[01:04] <jdstrand> kern.log
[01:04]  * jdstrand is just passing by
[01:04] <MTecknology> thanks :D
[01:04] <jdstrand> sure
[01:20] <cyphermox> not quite related, but is anybody else having issues with ec2 instances?
[01:20] <cyphermox> issues as in inability to connect to them :)
[01:25] <MTecknology> kindofabuzzkill: hi
[01:25] <kindofabuzz> how do i change auto security updates to manual updates?
[01:25] <kindofabuzz> MTecknology: hello
[01:25] <MTecknology> it's a cron task
[01:26] <kindofabuzz> MTecknology: oh ok
[01:26] <kindofabuzz> MTecknology: sudo crontab -e shows nothing
[01:27] <MTecknology> kindofabuzz: ls /etc/cron.*
[01:28] <kindofabuzz> MTecknology: /etc/cron.daily: apache2  apport  apt  aptitude  bsdmainutils  dpkg  logrotate  man-db  mlocate  popularity-contest  standard
[01:29] <kindofabuzz> maybe aptitude?
[01:30] <MTecknology> kindofabuzz: look in the files and see what they do ;)
[01:36] <kindofabuzz> MTecknology: i've looked through ever one of those scripts and could not find anything about updating, except mlocate had something about updatedb
[01:37] <MTecknology> kindofabuzz: read the apt script ;)
[01:37] <MTecknology> kindofabuzz: don't disable it - but do look at it
[01:37] <kindofabuzz> MTecknology: lol i missed that one. read them all 'cept that one
[01:43] <kindofabuzz> MTecknology: found /etc/apt/apt.conf.d/50unattended-upgrades, can i just comment out the security line?
[01:45] <kindofabuzz> yup. https://help.ubuntu.com/9.10/serverguide/C/automatic-updates.html
[02:49] <maxagaz> hi, what should I add to /etc/apt/mirror.list to have karmic repositories ?
[02:52] <twb> Never heard of mirror.list
[02:52] <twb> Surely you mean sources.list?
[04:02] <jmarsden> twb: mirror.list is used by apt-mirror
[04:02] <twb> Hm.
[04:03] <jmarsden> maxagaz: Lines like    deb http://us.archive.ubuntu.com/ubuntu karmic main restricted universe multiverse
[04:03] <jmarsden> Also similar ones for karmic-updates, karmic-security, karmic-backports and karmic-proposed if you mirror those too...
[04:07] <maxagaz> ok
[04:07] <maxagaz> thanks
[04:07] <jmarsden> You're welcome.
[04:19] <bep> right now i have my 2 1tb drives rsync eachother, and I want to do a raid so things are easier. But for one reason or another i do a lot of fresh installs, so if i decide to reinstall ubuntu server will i have any problems setting the raid back up without losing data?
[04:20] <jmarsden> bep: Assuming you partition them sanely, and don't format/recreate the /home partition where your data is, you should be OK... and you have backups anyway, offline, right?  If not, rsync is safer than RAID for you anyway.
[04:21] <bep> no right now I jsut have the 2 drives in the one machine. i know its not smart. only thing I have i couldnt lose are family pics etc and i do have those backed up else where
[04:25] <jmarsden> It's not smart to "do a lot of fresh installs" either.  Use virtual machines for test installs, not your main/only real physical hardware.
[04:27] <jmarsden> I'd suggest removing one drive, putting in an external case, then use it for regular backups and disconnect it from the machine when not actually backing up...  Since you can "rsync to each other" you must have under 1Gb total family pictures etc. anyway.
[04:28] <jmarsden> You will get more benefit from a good backup regime than from using RAID1, in your current circumstances.
[04:29] <bep> ok thanks, definitely something ill consider
[04:29] <jmarsden> You're welcome.
[04:31] <jmarsden> bep: One example of why ... RAID1 would make you very prone to "accidentally deleting the wrong files" kind of failure... having a separate disconnected backup makes that a lot easier to recover from.
[04:32] <bep> ya, thats exactly why i went with rsync to begin with. because i do web sites and sometimes with a ton of small files with a ton of edits it nice to go back to the rsync files from a few hours before
[04:33] <jmarsden> bep: Yes... also learn to use a version control system for that kind of work :)
[04:34] <bep> heh ya. i dont do web sites professionally or anything (well kinda i do sometimes) but ya
[04:34] <bep> im trying to get organized
[04:34] <bep> i would really like a nas with tiny power consumption to run 24/7 then just have htpc on when needed
[04:35] <bep> but I cant find anything that i like so far
[04:35] <bep> ive looked at the atom boards but idk
[04:36] <jmarsden> bep: Build one you do like? :)  Have you seen the Jetway board: http://www.newegg.com/Product/Product.aspx?Item=N82E16813153096
[04:38] <bep> ya , but i need gigabit lan and would like around 4 sata ports just in case of future expansion
[04:38] <jmarsden> That's not going to be quite so "low power" then...
[04:38] <bep> ya
[04:38] <bep> hence my problem
[04:38] <bep> im hoping the next round of atoms will be better but i have a feeling they wont be
[04:39] <jmarsden> Your problem is at least partly caused by physics... doing stuff faster uses more power... it's not really fair to blame the Atom for that :)
[04:39] <bep> ya
[04:40] <bep> im not atom does seem like a nice platform
[04:40] <bep> especially with ion etc
[04:41] <bep> i dont mind having drives spin down etc
[04:41] <bep> but id like to have it idle at 30 watt or less
[04:42] <bep> with bill here 30 watts would add $3 a month, and for the convenience of the always on its definitely worth it
[04:43] <jmarsden> bep: How much does the HTPC draw?  You might find leaving that always on would work for you, at close to those kind of power levels?
[04:43] <bep> 70 watt
[04:43] <twb> Presumably you've already considered IBM SSDs?
[04:43] <bep> yes but 1tb ssds are not in my price range :)
[04:43] <twb> Nod.
[04:45] <jmarsden> Hmmm, interesting tradeoff... running the HTPC costs you $4/month more than running a not-yet-purchased NAS... if the NAS costs you $200 to construct, that's over 8 years to recoup its cost.  I'd just run the HTPC :)
[04:46] <jmarsden> Oh, wait... did I mess up my mental math...?
[04:46] <twb> Signs point to ye
[04:46] <jmarsden> $200/4 = $50, so a year or so recoups the cost.... my bad.
[04:47] <bep> ya ive been considering that
[04:47] <jmarsden> $200/4 = $50, so 50 months recoups the cost.
[04:47] <bep> ive only been planning on spending like 75 though since all i need is motherboard & processor
[04:47] <bep> plus  i have a 25$ credit at newegg so really only out about $50
[04:47] <jmarsden> No case, RAM, PSU?
[04:47] <bep> have it all
[04:48] <jmarsden> OK.
[04:48] <bep> though I would like a new smaller case but its not a must
[04:50] <bep> right now my htpc is about to become both so i am going to see how it works out for me
[06:01] <wizardslovak> hello people
[06:02] <wizardslovak> saturday i will be getting new server box and i need to know how to back up or transfer settings from my current ubuntu box
[06:12] <ScottK> jiboumans: https://blueprints.launchpad.net/ubuntu/+spec/server-lucid-more-mail-integration ready for review again.
[06:17] <qman`> wizardslovak, tar up your /etc directory and copy it over
[06:21] <wizardslovak> what if i want to backup just firewall and apache
[06:32] <SandGorgon> hi guys.. we are migrating from one hosted UBuntu 8.04 box to another. I want to migrate all users and passwords as well. Is this possible ?
[07:07] <qman`> wizardslovak, /etc/apache2
[07:08] <qman`> for firewall, it depends on how you have it set up
[07:09] <qman`> SandGorgon, yes, assuming they're system users. Just copy the users from your /etc/passwd and /etc/shadow files
[07:40] <SandGorgon> qman`, does it mean if I copy the relevant entries for /etc/passwd and /etc/shadow - it should all work right ?
[07:40] <SandGorgon> i would have to create the userhomes using a script
[07:41] <qman`> yes, just copy the users you want to keep out of both files to the destination system
[07:41] <qman`> and recreate the directories
[07:41] <qman`> you probably want to copy the homes over anyway
[08:33]  * starsunflowersu is away: I'm busy
[09:07] <jiboumans> good morning guys
[09:08] <jiboumans> ScottK: i'll take a look today, thanks for pointing me at it
[09:30] <jiboumans> ScottK: thanks for the updates; looks great now so I've approved it
[10:33] <lenios> ScottK, is it necessary to have subtasks: basic (current mail-server) and mail-delivery (dovecot-postfix) ?
[12:18] <seeker_> hello all
[12:19] <seeker_> I am using Eucalyptus 1.6. when I try to run the image I am getting some error,
[12:20] <seeker_> can some one help me in this?
[12:26] <jiboumans> seeker_: what is the 'some error' you are getting?
[12:27] <seeker_> Server: SERVICE: FinishedVerify PROBLEM: null MSG-TYPE: RunInstancesType
[12:27] <seeker_> got the above error.
[12:27] <seeker_> can you help me to resolve this?
[12:27] <jiboumans> seeker_: hmm, that's very sparse. any chance of a log file or so to go with that?
[12:27] <jiboumans> ttx may be more suited, if he's around though
[12:28] <ttx> o/
[12:28] <ttx> seeker_: what command did you run to get that error ?
[12:29] <ttx> seeker_: double check with "euca-describe-availability-zones verbose" that there is enough "space" in your cloud to run your instance type
[12:30] <seeker_> it is failing in ec2-run-instances cmd itself
[12:30] <seeker_> so I cannot  do "euca-describe-availability-zones verbose"
[12:31] <ttx> seeker_: could you paste the ec2-run-instances comand line you used ?
[12:31] <seeker_> am i correct?
[12:31] <seeker_> yes I'll do that
[12:31] <ttx> you can run "ec2-describe-availability-zones verbose
[12:32] <seeker_> ec2-run-instances $EMI -k mykey
[12:32] <seeker_> emi number I got it, when I registered
[12:33] <ttx> seeker_: what does "ec2-describe-availability-zones verbose" return ?
[12:33] <ttx> seeker_: you end up using a default instance type
[12:33] <ttx> seeker_: for which you might not have any "space" in your cloud for
[12:34] <seeker_> no its (ec2-describe-availability-zones verbose) not printing anything.
[12:34] <ttx> seeker_: I'd recommend first getting everything described in the docs working, before starting to deviate
[12:35] <ttx> https://help.ubuntu.com/community/UEC
[12:35] <ttx> theat involves using the euca2ools commands (euca-*)
[12:35] <ttx> if everything works ok, then it's time to try something smarter
[12:35] <ttx> like running with ec2-* commands
[12:35] <seeker_> i am using "Eucalyptus-Jaunty "
[12:36] <ttx> ah! Then your basic premise (" I am using Eucalyptus 1.6") is false
[12:36] <ttx> or is it not ?
[12:36] <ttx> Jaunty is 1.5
[12:37] <seeker_> ohh I am sorry then.
[12:37] <seeker_> I followed https://help.ubuntu.com/community/Eucalyptus-Jaunty
[12:37] <seeker_> the above steps
[12:37] <ttx> right
[12:38] <seeker_> I have reached till step6
[12:38] <ttx> seeker_: I fear my experience running the 1.5 "technology preview" is very limited. I'd strongly recommend using the karmic UEC (based on euca 1.6)
[12:43] <seeker_> ttx: you want  me to switch to 1.6 version
[12:43] <seeker_> jiboumans: any help from your side?
[12:45] <jiboumans> seeker_: a lot of improvements in 1.6; switching is much recommended
[12:46] <seeker_> ok I will do that..but is there any solution for this?
[12:46] <seeker_> because I have done lot of work for this
[12:46] <seeker_> jiboumans: ok I will do that..but is there any solution for this? because I have done lot of work for this
[12:48] <VSpike> Is saned on ubuntu server configured by default to bypass inetd/xinetd?
[13:07] <dashnu> Im runing ubuntu-server on Virtual PC 07, The ubuntu server date keeps changing.. Nothing can fix it short of a hard reboot. cant sudo when it happens
[13:07] <dashnu> anyone know why ?
[13:08] <dashnu> i even have ntp installed and running still no luck
[13:08] <dashnu> when i reboot tims is ok again
[13:11] <seeker_> ttx: you want  me to switch to 1.6 version
[13:11] <ttx> seeker_: I encourage you to :)
[13:12] <seeker_> ok...but is there any way to fix this issue?
[13:13] <seeker_> ttx: in this link also http://open.eucalyptus.com/wiki/EucalyptusKnownBugs_v1.4 they mention the issue
[13:13] <seeker_> but i dont know how to change the parameter for VM type
[13:28] <VSpike> Having trouble with saned... it seems to be only binding to ipv6... http://pastebin.com/d4f6d0133
[13:28] <VSpike> It used to work, and other than any updates I'm not aware of having changes anything
[13:28] <VSpike> Also http://pastebin.com/d12914ac2
[13:30] <ttx> seeker_: pass a "-t " to ec2-run-instances ?
[13:30] <ttx> like "-t c1.medium" if your image fits into 4Gb
[13:30] <VSpike> This looks relevant, except that he is using inetd and I'm not. https://bugs.launchpad.net/ubuntu/+source/inetutils/+bug/379621
[13:32] <seeker_> ttx: I getting the same error
[13:32] <seeker_> ttx: my image will fit in 4gb
[13:32] <VSpike> Oh, getting closer http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg352293.html
[13:33] <VSpike> Ah well, I can probably disable ipv6 on the server. Not actually using it
[13:34] <VSpike> It seems that whenever I have IPv6 enabled on a system, it just seems to cause problems :)
[13:34] <jiboumans> VSpike: taking a quick look at the manpage, the saned.conf file is supposed to allow you to set this stuff up
[13:34] <jiboumans> no guarantees that it actually *does* though
[13:34] <ScottK> lenios: I'm not sure, in the end, how we will work out the multiple configurations of mail servers question.  dovecot-postfix exists today and I'm going to do amavis-postfix for Lucid.  Stitching them together is a bit TBD.
[13:37] <ttx> seeker_: my guess would be that there is no room for any instance, like the node controller isn't registered (and you don't get anything in "ec2-describe-availablility-zones verbose"
[13:37] <ttx> but that's a 1.5-uneducated guess
[13:39] <seeker_> ttx: but from the front (using url)  end I am able to see the image and I got the emi number too
[13:40] <ttx> the emi is registered. There is just no free capacity in the cloud to launch it
[13:40] <seeker_> ttx: ok..how to approach this?
[13:40] <VSpike> jiboumans: it seems that if saned binds an address and port in ipv6, it is then prevented from doing so in ipv4
[13:43] <ttx> seeker_: that's the thing, I'm not sure how to debug this in 1.5 :)
[13:44] <seeker_> ttx: have you seen this http://open.eucalyptus.com/wiki/EucalyptusKnownBugs_v1.4 they mention the issue
[13:45] <seeker_> ttx: by going through this, can you provide help?
[13:45] <ttx> That's 1.4 doc and they say the proper solution will appear in v1.5.
[13:46] <ttx> seeker_: I can't, I don't have a 1.5 setup to test with
[13:46] <ttx> hence my recommendation.
[13:59] <zul> ttx: once the package is in main for the canonicala-application-support spec should it be removed from the workitems list?
[14:01] <ttx> zul: no, keep it there with "DONE"
[14:01] <ttx> otherwise it will look like you never worked on anything
[14:02] <zul> ttx: k
[14:08] <VSpike> I'm puzzled.  My scanner is only accessible to root, and I can't see why
[14:08] <VSpike> It's a USB scanner, and it is mention is /lib/udev/rules.d/40-libsane.rules
[14:09] <VSpike> s/mention is/mentioned in/
[14:09] <VSpike> Bus 003 Device 002: ID 04b8:0104 Seiko Epson Corp. Perfection 1200
[14:09] <VSpike> ATTRS{idVendor}=="04b8", ATTRS{idProduct}=="0104", ENV{libsane_matched}="yes"
[14:10] <VSpike> crw-rw-r-- 1 root root 189, 257 2009-12-11 14:01 /dev/bus/usb/003/002
[14:10] <VSpike> What am I missing?
[14:16] <fullstop> Hi all!  Does anyone here have experience with kvm and gpt partitioned block devices?
[14:18] <fullstop> http://www.pastebin.org/63522
[14:21] <fullstop> I created a /dev/sda3, 400 GB, and gave it to a KVM guest to use.  When I installed Ubuntu server 9.10 on the guest, I partitioned /dev/sda3 in the installer.. using DOS partitioning with fdisk.
[14:22] <fullstop> After this, the host linux will no longer boot from grub, and I have to boot into rescue mode and delete /dev/sda3 before it will boot.
[14:24] <fullstop> quiet in here.  :)
[14:28] <VSpike> sure is
[14:32] <fullstop> is that normal for here?
[14:33] <ScottK> It varies a lot
[14:36] <hackeron> I get messages like this in my syslog every hour or so for the past year: "smartd[2644]: Device: /dev/sda, 8 Currently unreadable (pending) sectors" -- what does pending mean and what do I do about this? (specifically to stop the messages flooding my logcheck reports)
[14:38] <_ruben> replace the disk?
[14:41] <hackeron> _ruben: well, it's been working with these bad sectors for a year so far, another similar disk has been working for 3 with 5 unreadable sectors -- why bother replacing?
[14:42] <VSpike> hackeron: perhaps you could run badblocks on it?
[14:43] <_ruben> or just add an ignore to your logcheck config, if you dont care about the message
[14:44] <VSpike> hackeron: "e2fsck -ck/dev/sda" or something like that
[14:44] <VSpike> hackeron: "e2fsck -ck /dev/sda" even
[14:44] <hackeron> VSpike: I'm using raid though
[14:45] <VSpike> ah
[14:45] <hackeron> VSpike: so there is no filesystem on /dev/sda
[14:45] <hackeron> VSpike: I just have /dev/md0 which maps to /dev/sda1 and /dev/sdb1
[14:46] <hackeron> _ruben: I don't want to ignore it, I want the disk to remap the blocks and reconstruct the array
[14:46] <_ruben> could try to break the mirror, wipe the disk (possibly with a disk management tool from the vendor), recreate the raid
[14:46] <[diablo]> afternoon all.. guys, can abgx360 change the region code of an ISO or not please?
[14:46] <[diablo]> whoops wrong window
[15:11] <stacmandu> i just plugged a modem into my ubuntu-server 9.04.  how do i tell if it's working?
[15:12] <hackeron> _ruben: seems all I need to do is echo repair > /sys/block/mdX/md/sync_action -- and the drive appears to remap the bad sector
[15:14] <_ruben> hackeron: nice to know :)
[15:26] <zul> has anyone thought of packaging spice now for kvm since its open now
[15:27] <Status0> hello ppl
[15:27] <zul> http://www.spice-space.org
[15:28] <Status0> i have installed to my qa the ubuntu 9.10 server
[15:28] <Status0> and i have some question
[15:28] <Status0> i personaly try develop ubuntu community in israel
[15:29] <Status0> can anybody help me about ubuntu server ?
[15:29] <zul> Status0: someone might be able to answer
[15:30] <Status0> thanks, i wonder why there is no gui to server ?
[15:30] <arj> i wonder why one would need a gui to server
[15:31] <Status0> arj, becuse ppl are stupid [most of them] but again this is some required by consumers
[15:32] <Status0> the want to make one click and it will work
[15:32] <Status0> some-where i can understand them
[15:32] <Status0> about this point of view
[15:33] <arj> they can use webmin
[15:33] <arj> or plesk
[15:33] <Status0> yeh but i want "sell" ubuntu server and i have nothing why ubuntu server better ?
[15:36] <Status0> ok i will re-ask the question : why ubuntu server is better ?
[15:36] <arj> better than what?
[15:36] <arj> if you want a gui, ubuntu-server might not be the best choice
[15:37] <Status0> arj, generaly why is better ? (not about the gui) i try to promote more ubuntu. but i cant find resone..
[15:38] <lau> hi, i am trying to set an tls signed mta in order to send signed emails thru my web php/mail() function
[15:39] <lau> i am running hardy, apache2 and apache2/php5/php.ini shows nothing particular except SMTP = localhost and smtp_port = 25
[15:39] <lau> i assume exim is my mta but ps aux|grep exim returns nothing
[15:40] <lau> how does the php/mail() function send mail ?
[15:40] <Status0> yes
[15:40] <Status0> lau, but you need to configure the php.in
[15:41] <Status0> to use your mail
[15:41] <lau> Status0: currently the .php application is sending mails
[15:42] <lau> I want to add tls functionnality in order to sign the emails sent thru this application
[15:42] <lau> thus I am assuming a mta is already up and running on my server, is that a good assumption ?
[15:43] <ScottK> lau: Why do you assume exim is your MTA?
[15:46] <Jeeves_> ScottK: It's the only sane mailserver ;)
[15:46] <Status0> try opensocket to tls
[15:47] <ScottK> Since Ubuntu Server ships no MTA by default and prefers Postfix in general, assuming Exim is an odd choice.
[15:47] <Jeeves_> ScottK: I know :)
[15:47] <ScottK> One might prefer Exim (I don't), but I don't see how one should assume it.
[15:47] <Jeeves_> But so is the preference for Postfix :)
[15:47] <lau> ScottK: dpkg -l exim4 returns something, I got files in /etc/exim4/...
[15:48] <ScottK> lau: OK.  That's a good reason then.
[15:48] <lau> also I ls -lsa /usr/sbin/sendmail
[15:48] <lau> and this returned /usr/sbin/sendmail -> exim4
[15:48] <lau> thus now I am assuming that the php application is sending emails thru calls to sendmail
[15:49] <lau> that is a simlink to exim4
[15:49] <lau> so exim4 is used in order to send emails
[15:49] <lau> am i right ?
[15:49] <Jeeves_> probably
[15:49] <lau> (enven if apache2/php5/php.ini is not configured)
[15:51] <lau> if I setup TLS/SSL for exim4 I should be ok ?
[16:38] <zul> kees: ping
[18:23] <TopKatz> what procedure do you guys use that have hardware raids when a update to the headers comes out?  I know Im not doing it right.
[18:24] <arj> what's the relation between hardware raids and "the headers"? :)
[18:24] <TopKatz> after installing new headers, the driver is gone
[18:25] <TopKatz> so the raid comes up as seperate disks, causing parity faults (I think)
[18:25] <TopKatz> and all around bad times
[18:25] <arj> seperate disks should not cause parity faults
[18:25] <arj> are you sure it's hardware raid and not fake raid?
[18:25] <TopKatz> not really
[18:25] <TopKatz> Im using a rocketraid 1740
[18:25] <TopKatz> I have the driver installed
[18:25] <TopKatz> and the managment software
[18:26] <TopKatz> so I think Its a real raid
[18:26] <arj> reinstall the driver for the correct kernel before you reboot
[18:26] <TopKatz> hmm
[18:26] <arj> if you need a driver I doubt it's real
[18:26] <TopKatz> I jsut dont think its suported out of the box
[18:26] <TopKatz> so your sugesting a make install before rebooting after header update
[18:27] <TopKatz> will it use the new header in that case?
[18:27] <TopKatz> I never tell it which to use, it just does it right now
[18:27] <arj> I don't know, if you're lucky you can specify what kernel to build for
[18:27] <TopKatz> hmmm
[18:27] <TopKatz> its an interesting idea
[18:28] <TopKatz> right now I jsut go to the driver dir and do sudo make install and it does the whole thing
[18:29] <arj> is the root or boot fs on those disks?
[18:29] <arj> if not, I'd halt, disconnect them, boot the new kernel, build the driver, connect them, boot
[18:29] <TopKatz> no, it has my mysql and www stuff
[18:29] <TopKatz> yeah, I have been thinking about that
[18:29] <TopKatz> or, how to do all that
[18:30] <arj> and I'd also check how well a raid card this card is, and consider software raid
[18:30] <TopKatz> its not a high end one, but this is also just a test dev enviorment
[18:31] <arj> if it's not high end, consider using it as a sata controller and do mdraid
[18:32] <arj> or run some benchmarks on the driver thingy and mdraid
[18:32] <TopKatz> when you update the header does it put it in /lb/modules... before the reboot?  so would they be there after update but before reboot?  The driver make builds against that
[18:33] <TopKatz> so in theory your inital idea might work
[18:33] <arj> I got to go now, good luck :)
[18:34] <TopKatz> thanks for your time
[18:34] <arj> no problem
[18:45] <jiboumans> time to pretend it's weekend; later guys
[19:02] <zroysch> I have an ubuntu 9.04 server that is having a multitude of problems that seem to be all related to network activities. rtorrent will randomly crash, mpd server also randomly crashes. My ssh and any other tcp connection will randomly go down for a few minutes at a time while i can still ping it, then tcp will come back up again. I have found no logs to tell me what the hell is going on. could anyone tell me where to find 
[19:52] <fullstop> are there any known issues w/gpt KVM setups and grub2?
[19:54] <fullstop> I have a KVM host which uses GPT due to large partition sizes.  I gave a KVM guest a raw partition to use.   The guest used a MS-DOS partition table on that raw partition, such that fdisk -l /dev/sda3 would show sda3p1 and sda3p2.
[19:55] <fullstop> However, this messed up grub2, which was noticed after a reboot.  It just sat there at "Loading GRUB"
[19:55] <fullstop> booting into rescue mode and deleting /dev/sda3 and re-installing grub recovered.  I don't know if reinstalling grub was necessary.
[19:56] <fullstop> I am more familiar with xen, but KVM seems to be the way the winds are moving right now.
[20:04] <fullstop> wow, netsplit sucks
[20:05] <Belgarath> hi
[20:06] <Belgarath> does anybody know how to clear kerberos instalatioon/configs
[20:06] <Belgarath> so I can start all over
[20:06] <Belgarath> without reinstalling ubuntu ?
[20:07] <Belgarath> dpkg --purge is nto good enough apparently
[20:22] <Listerthrawn> I want to set up KVM on a server with LVM on Raid and use the LV's as the disks for KVM.  Anyone know a good guide on aligning the 3?
[20:28] <Clusty_> still trying to jail a user and allow him to use just scp and not ssh nor to snoop around and reall all the stuff
[20:28] <Clusty_> i tried this: http://geekzine.org/2007/09/28/easy-sftp-and-chroot-sftp-with-scponly/
[20:28] <Clusty_> but it chokes
[20:29] <Clusty_> connection closes for no good reason
[20:38] <mjs2> hey - just installed karmic koala but i'm stuck behind a firewall.  what's the proper way to make apt use an http proxy?
[20:39] <mjs2> simple google resulted in a few methods - setting "http_proxy" env var (didn't work), setting /etc/apt/apt.conf (doesn't exist, apparently changed in 9.1)..
[20:46] <mjs2> well, never mind.  apparently i have to apt-get update
[20:50] <ruben23> hi
[20:50] <BlueT_> ruben23: morning
[20:51] <ruben23> BlueT_:what time in your place now..?
[20:52] <BlueT_> ruben23: 4:51am here in Taiwan :P
[21:01] <[diablo]> evening #ubuntu-server
[21:01] <Clusty_> ahoy
[21:01] <[diablo]> guys, I have a box I wish to run ubuntu server 8.04 on ...
[21:01] <[diablo]> I can only slap a wireless (atheros) pci card in it... will server still have support for the ath card?
[21:03] <Clusty_> [diablo]: this is a kernel issue
[21:03] <Clusty_> 8.04 is running 2.6.24
[21:04] <Clusty_> if support is not there you can add it
[21:04] <[diablo]> sure, true
[21:04] <[diablo]> actually I should really set up an AP -> ethernet
[21:04] <[diablo]> less messy
[21:04] <Clusty_> any reason to put 8.04 and not 9.x?
[21:05] <[diablo]> yeah, the project I wish to develop I prefer LTS
[21:05] <Clusty_> i see
[21:05] <[diablo]> either that or would have to go RHEL or CentOS
[21:05] <Clusty_> i am also unfortunately on 8.04
[21:05] <[diablo]> but for server, its a good move
[21:05] <Clusty_> cursing already for 2 days cause of this
[21:06] <Clusty_> i want to set up a username that cnot see stuff outside his home folder via sftp
[21:06] <Clusty_> next from 8.04 works out of the box
[21:08] <[diablo]> that can be done
[21:09] <[diablo]> Clusty_, there is an option in sshd to do it
[21:09] <Clusty_> yeah, chroot with scponly
[21:09] <[diablo]> nope
[21:09] <Clusty_> [diablo]: for next version from 8.04
[21:09] <Clusty_> :D
[21:09] <[diablo]> well, sort of
[21:09] <[diablo]> you sure?
[21:09] <Clusty_> 8.04 has old openSSH
[21:09] <[diablo]> one second
[21:09] <Clusty_> yeap
[21:10] <[diablo]> Clusty_, http://www.openbsd.org/cgi-bin/man.cgi?query=sshd_config
[21:10] <[diablo]> Clusty_, check the ChrootDirectory option
[21:11] <Clusty_> the message is: cannot do chroot inside a blahhh block
[21:11] <Clusty_> the one that catches one user
[21:12] <[diablo]> Clusty_, do a man sshd_config
[21:12] <[diablo]> does it show the ChrootDirectory in there?
[21:13] <Clusty_> nope
[21:13] <[diablo]> Clusty_, what version of sshd is it?
[21:14] <[diablo]> Clusty_, actually if its not supported in the version shipping with LTS, personally, for I would just build from source the openssh server..
[21:14] <Clusty_> OpenSSH_4.7p1 Debian-8ubuntu1.2, OpenSSL 0.9.8g 19 Oct 2007
[21:14] <[diablo]> Clusty_, you could even have it running on a higher port and not mess up your ubuntu sshd
[21:14] <[diablo]> Clusty_, you should be able to have it up and running in a couple of minutes
[21:14] <[diablo]> Clusty_, openssh is easy to compile
[21:15] <Clusty_> [diablo]: i would have to write my own rc.d script, which feels scary :D
[21:16] <[diablo]> Clusty_, just copy the existing one
[21:16] <Clusty_> [diablo]: plus machine is behind a firewall and i do not have easy access to firewall to punch holes through
[21:17] <[diablo]> Clusty_, then rebuild the source package with the latest version
[21:17] <Clusty_> good point
[21:17] <[diablo]> Clusty_, also double check your sshd_config file to be sure the ChrootDirectory is not listed
[21:18] <Clusty_> i searched through man
[21:18] <[diablo]> yeah, but look in the actual config file too
[21:19] <Clusty_> ohh ok
[21:19] <Clusty_> nope
[21:19] <[diablo]> ah ok...
[21:19] <[diablo]> apparently it seems its from 4.8 onward
[21:20] <[diablo]> you could try grabbing the .deb from 8.10 ... and installing it
[21:20] <[diablo]> sorry Clusty_ I have to walk my dog... bbl
[21:21] <Clusty_> :D
[21:36] <Listerthrawn> I notice you can create a 2 device raid10 array in mdadm.  Are there any benefits/drawbacks with using raid10 over raid1 with 2 disks?
[21:38] <qman`> Listerthrawn, that would effectively be exactly the same
[21:38] <qman`> different in name only
[21:38] <jeiworth_> Listerthrawn: i wouldn't do it, so the idea is to have 2 equally sized partitions on each disk, mirror the 2 disks and  use 1 partition each for striping?
[21:38] <Listerthrawn> What I want to do is this: -
[21:39] <Listerthrawn> I have 4 disks 2x500Gb and 2x750Gb (I think, it's been a while since I've seen them)
[21:39] <Listerthrawn> I want to install a small raid1 on the 2x750's and install ubuntu server.
[21:39] <ruben23> hi, any application whihc i can make a clone image of my local server form an ubuntu server then can restore then any time to the state i imgae thm.
[21:39] <jeiworth_> ruben23: dd
[21:39] <jeiworth_> Listerthrawn: ok...
[21:40] <neonfreon> Listerthrawn i like to setup 2 disk RAID 1 groups and then if i want to extend beyond that use LVM
[21:40] <neonfreon> add another 2 disks for another raid 1 group and then just add the new physical volume to my volume group
[21:40] <Listerthrawn> I then want to use the rest of the space on the 2x750's as another raid1, then the 2x500's as another raid1 then LVM them
[21:41] <neonfreon> yes
[21:41] <Listerthrawn> neonfreon, We are so on the same wavelength here.  Great so far :)
[21:41] <neonfreon> i think that's a good setup
[21:41] <neonfreon> except i didn't raid1 my OS
[21:41] <neonfreon> because it's easy to replace
[21:41] <neonfreon> (compared to data)
[21:41] <qman`> Listerthrawn, the correct thing to use for that setup is raid 1, not raid 10
[21:41] <Listerthrawn> True, but I want to run KVM on top of this and use the LVM'd RAID arrays as backing storage for VM's.
[21:41] <qman`> raid 10 would be if you want to set aside 500GB on all four disks for one array
[21:42] <Listerthrawn> so while the OS is easy to replace, the downtime isn't easy to stomach.
[21:42] <ruben23> jeiworth_: i read on ubuntu pro book- about DRBL and clonezilla..to do imaging of server.
[21:42] <neonfreon> sure Listerthrawn makes since for time critical setups
[21:42] <neonfreon> sense
[21:43] <jeiworth_> you could also create a 4x500gb raid5 or 6 and use the 2x250gb that are over for another raid1 for system, programs, etc...
[21:43] <neonfreon> boo software raid5
[21:43] <qman`> ruben23, that's more suited to making multiple copies, like setting up a bunch of identical client computers
[21:43] <Listerthrawn> What I'm really having trouble with is finding a good guide on how to align all the different chunk/stripe sizes from disk/raid/lvm/filesys
[21:43] <qman`> ruben23, you can use dd to make a single raw backup, and compress with gzip or bzip2
[21:43] <neonfreon> software raid5/6 is a no go for me
[21:43] <jeiworth_> ruben23: no, but i also read about drbl, on the fly synchronization between 2 machines over network
[21:44] <Listerthrawn> i've used software raid5 before when I wanted to store lots of static data with some resilience but with the relatively high random writes I expect from the VM workload it wouldn't be suitable I don't think.
[21:44] <jeiworth_> ruben23: is this a one time thing or do you want to make regular backups that way?
[21:45] <ruben23>  jeiworth_: yes, regular backup and aside that i have an imgae of my server so i can restore it on a good working condition then copy my backup files to make it run if any donw time..
[21:45] <jeiworth_> Listerthrawn: well, if you prefer you culd make a raid10 the same way, 4x500gb and a raid1 for the remaining 2x250gb
[21:45] <qman`> I use software raid 6 on my file server, but VMs are a much heavier load
[21:46] <ruben23> or what steps you can suggest best would fit on the purpose of failover.
[21:46] <qman`> ruben23, for failover you need two servers up and running
[21:46] <qman`> dedicated to that purpose
[21:47] <Listerthrawn> I could I suppose but eventually I'll run out of disk space and I intend to replace the 2x500's with 2x1TB by adding them in as another raid1 and pvmoveing the LVM off the 2x500's
[21:47] <Listerthrawn> Anyone got any advice on how to align all the stripe stuff?
[21:56] <Listerthrawn> anyone still here?
[21:56] <Listerthrawn> i've not used IRC for nearly 10 years.  Thought they'd have fixed netsplits by now!
[21:58] <neonfreon> what size writes and reads will you be doing?
[21:58] <neonfreon> if you're doing raid 1 there is no stripe size to configure
[21:58] <neonfreon> at the raid level
[22:00] <IrCYop> Could someone be so kind to paste me their /etc/init.d/networking file for 9.10?
[22:00] <qman`> IrCYop, http://pastebin.com/f2ff08c3c
[22:00] <IrCYop> qman`: <3
[22:00] <crohakon> Awww, a cute couple.
[22:00] <qman`> IrCYop, this system has been upgraded since 7.10, so that might not be the stock 9.10 file, but it's a working 9.10 system
[22:00] <qman`> and I haven't messed with it
[22:00] <crohakon> Should I bother upgrade my server to 9.10?
[22:00] <crohakon> I just upgraded my laptop, and it seemed to go well...
[22:00] <ruben23>  jeiworth_: what is best option on an external drive or another server..?
[22:00] <qman`> crohakon, if it's 9.04, yes, if it's 8.04, no
[22:00] <qman`> reason is, unless you really need something, 8.04 will be able to upgrade straight to 10.04
[22:00] <ruben23>  jeiworth_:how do dd backup an entire image of my linux server
[22:00] <qman`> so you can avoid a lot of hassle if you can wait until april
[22:00] <jeiworth_> ruben23: hehe well, that really depends, do you want to be able to move your backup to safe places or is the other server in a safe place, i.e. far away from the 1. server, another building, another city, another country or even continent? ;)
[22:00] <qman`> ruben23, dd if=/dev/sda1 of=/path/to/backup-file
[22:00] <jeiworth_> yes, with dd you can backup the entire block device as is, e.g. dd /dev/sda
[22:00] <qman`> where /dev/sda1 is your server's filesystem
[22:00] <jeiworth_> haha
[22:00] <qman`> or you can compress on the fly, dd if=/dev/sda | gzip > /path/to/backup-file
[22:00] <ruben23>  jeiworth_:how about the restoration..?
[22:00] <jeiworth> ruben23: if you make e.g. sudo dd if=/dev/sda of=//server/share/backup this will actually backup the entire disk including mbr and everything, you should pipe it through tar and bzip2 though, just to decrease size
[22:00] <jeiworth> ruben23: exactly the same, boot with a live-cd and then: sudo dd if=//server/share/backup of=/dev/sda
[22:00] <qman`> yeah, dd will create a raw file, including free space, so you definitely should compress it with something, even if you can't use the CPU on good compression
[22:00] <jeiworth> qman`: say, doesn't dd itself also have a parameter to ignore unused blocks?
[22:01]  * jeiworth checks man:dd....
[22:01] <qman`> not that I know of, it would have saved me a lot of effort in the past
[22:01] <jeiworth> hehe so i better not find it then, not to embarass you? ;o)
[22:02] <qman`> learn something new every day :)
[22:03] <Listerthrawn> neonfreon, I'm not too bothered about the size, the default may well be OK,  I just want to make sure that 1 fs block is 1 lvm block is one Raid block.  If they don't align 1 fs read can mean reading 2 raid blocks from disk and that's gonna kill performance
[22:04] <jeiworth> hmm qman`nope, don't find anything in the man doc
[22:04] <ruben23> i will find how to for dd...
[22:05] <ruben23> are there any, just a guide for reading
[22:05] <jeiworth> hehe i think for dd you won't have trouble finding
[22:07] <ruben23>  jeiworth: if you have a good link there, that would be deeply appreciated..:-D
[22:08] <neonfreon> Listerthrawn: i dont think there is anything to worry about if you're doing raid 1 and linear LVM
[22:08] <neonfreon> there is no striping occuring there
[22:08] <crohakon> How do I upgrade to 9.10 from the command line?
[22:09] <neonfreon> so you just want to look at your file system's block size to make sure it's most appropriate for the size of writes you'll be getting typically
[22:11] <jeiworth> sry ,eetnig
[22:11] <jeiworth> meeting
[22:12] <Listerthrawn> neonfreon, I'm not sure.
[22:12] <crohakon> How do I upgrade to 9.10 from 9.04 from the command line?
[22:13] <ScottK> crohakon: You mean 9.04 to 9.10?
[22:14] <crohakon> ScottK, thats what I said...
[22:14] <crohakon> to 9.10 from 9.04
[22:14] <ScottK> Ah.
[22:14] <crohakon> hehe
[22:14] <ScottK> Right.
[22:14] <ruben23> hi
[22:14] <ruben23> guys
[22:14] <ScottK> crohakon: sudo do-release-upgrade should do it.
[22:14] <crohakon> okay
[22:14] <crohakon> thanks
[22:15] <ScottK> No problem.
[22:17] <jumbers> Does update-motd run every time a user logs in? I've added a script to the update-motd.d folder and now every time I log in, it takes a while before the motd displays
[22:20] <jumbers> I see there's an update-motd cron job, but it seems like it also runs every time somebody logs in
[22:43] <neonfreon> Listerthrawn: what are you not sure about
[22:44] <Listerthrawn> http://wiki.tldp.org/LVM-on-RAID
[22:44] <Listerthrawn> neonfreon, the link i just said
[22:44] <Listerthrawn> neonfreon, My head's just toast at the moment lol.
[22:45] <Listerthrawn> neonfreon, I've been trying to work this out whilst off work sick with Swine Flu.  Trying to get your head around Raid/LVM and aligning the lot with a fever of 104 has been a challenge I'll tell you.
[22:51] <Listerthrawn> neonfreon, I think I may have just worked it out.  If I set my partition tables to start say 1024k into the disk rather than skipping 1 sector everything should align.
[22:51] <neonfreon> ah cool
[22:52] <neonfreon> start of your partition tables or start of your partitions?
[22:52] <Listerthrawn> partitions
[22:52] <neonfreon> yeah that's the conclusion i came to as well
[22:52] <neonfreon> if there are any alignment issues that will be it
[22:52] <neonfreon> issues caused by label or whatever at start of disk
[22:55] <Listerthrawn> i'll have to do this on a test rig.  do a bit of bonnie'ing
[22:55] <neonfreon> ya
[22:55] <Listerthrawn> see if it actually does make a difference.
[22:55] <Listerthrawn> you on here often?
[22:55] <neonfreon> have been lately
[22:56] <Listerthrawn> I'll let you know what I find.  Thanks for the help, much appreciated.
[22:56] <Aison> evening, i'm searching a way that my server can send SMS to my cellphone
[22:56] <neonfreon> i've tuned stripe alignment with raid 5 and XFS before
[22:56] <neonfreon> i used ioread to do my perf tests
[22:56] <Listerthrawn> i'll look at that tool too
[22:56] <neonfreon> i only saw differences on read performance, but it was really big
[22:56] <neonfreon> err, iometer
[22:56] <neonfreon> sorry
[22:56] <Aison> I heard, that it's possible to connect nokia devices with the server
[22:56] <Aison> but i cant find anything that explains that
[22:57] <neonfreon> lets you set size of write/read, number of streams, and change what % are sequential
[22:57] <Listerthrawn> only trouble is, the machine I want to use for this is currently used hosting Xen VM's and one of which is my firewall/router.  Arranging downtime with the wife is harder than arranging it at work.
[23:11] <bep_> does ubuntu server support sleep etc or do i need to install something else?
[23:12] <RoyK> bep_: sleep?
[23:13] <jumbers> RoyK: Suspend
[23:13] <RoyK> afaik suspend to disk should be there
[23:13] <jumbers> Suspend to disk would be hibernate, suspend is the one where it suspends to RAM
[23:13] <RoyK> but not sure if the userspace parts are installed by default
[23:14] <RoyK> k
[23:14] <RoyK> I don't really want that on a server, though
[23:14] <jumbers> Neither would I, especially because I don't have physical access
[23:15]  * RoyK only has ubuntu server VMs by now
[23:15] <jumbers> My box runs Ubuntu server on a dedicated machine. Pretty cheap monthly costs too
[23:19] <RoyK> I have an opensolaris pizzabox with three ubuntu VMs
[23:19] <RoyK> zfs for storage management and ubuntu for the apps part
[23:58] <IrCYop> Could anyone give me a hand with openvpn