[01:10] <leaf-sheep> Help. For some reason, my Ubuntu installation is unable to go beyond 33% of partitions formatting for 2TB HDD. I'm wondering if I did something wrong.  Maybe I'm supposed to set up RAID or something
[01:15] <qman__> leaf-sheep, RAID is only for multiple disk setups
[01:15] <qman__> and only in certain cases
[01:15] <qman__> what filesystem are you using?
[01:15] <PhotoJim> I'm googling, but I'm not aware of any issues with drives that size.
[01:16] <qman__> 2TB is a lot of data, and would take a very long time to format in ext3
[01:16] <qman__> how long did you let it set?
[01:17] <PhotoJim> it took me the better part of an hour to format a one terabyte drive, but that was on an older PIII server.
[01:17] <qman__> I would guess 2TB would take about two hours to format in ext3, based on my experience in setting up my RAID
[01:20] <PhotoJim> ext3 should be fine.  max. vol. size depending on implementation is 2-16 TiB.  and a 2 TB drive would not be 2 TiB, it'd be 2 trillion bytes, significantly smaller.
[01:20] <qman__> yeah, it just takes a while to format the drive
[01:20] <qman__> I have a 3.3TB RAID array
[01:20] <qman__> in ext3
[01:21] <PhotoJim> depends on blocksize, apparently.  1 KiB blocksize caps at about 2 terabytes.
[01:22] <qman__> yeah, I know the limit of a default ext3 setup is ~8TB
[01:22] <PhotoJim> with 4 KiB blocksize, which is the max. on most platforms.
[01:22] <PhotoJim> so that makes sense.
[01:23] <PhotoJim> so either it just needs more patience... or it's a bad drive or controller.
[01:23] <qman__> I need to decide on a new filesystem to use, my array is full and I need to upgrade
[01:23] <qman__> right now the choices are reiser and ext4
[01:24] <PhotoJim> you can't stick with ext3?
[01:24] <qman__> only to 8TB, but if I upgrade I don't think that'll be big enough
[01:24] <PhotoJim> ext4 is still a bit new for me. and reiser's future is in some doubt due to Herr Reiser's incarceration.
[01:24] <leaf-sheep> qman__: EXT4.
[01:24] <PhotoJim> ahh.
[01:25] <qman__> I trust ext3
[01:25] <qman__> I don't trust XFS
[01:25] <qman__> the others are up in the air
[01:25] <giovani> xfs is relatively unstable on a regular basis
[01:26] <giovani> ext4 is the future for ubuntu
[01:26] <giovani> although you may want to be aware of some options that ensure secure journaling
[01:26] <giovani> which aren't default in ubuntu afaik
[01:26] <leaf-sheep> It's not just formatting.  See, when I started the installation process... to the point of formatting and setting up disks to my likings, it'll start right away at 33% and idling -- but I can't be certain of background activities for 2TB is a large HDD.
[01:26] <giovani> leaf-sheep: you can go to the console
[01:26] <giovani> and look at activity
[01:26] <giovani> via iostat, etc
[01:27] <PhotoJim> looks like ZFS is an option too
[01:27] <qman__> leaf-sheep, the progress bar doesn't show the formatting process
[01:27] <giovani> or, if you're physically at the machine, obviously looking at the HD/CD activity will tell you if something's going on
[01:27] <qman__> it's divided up into how many partitions you have
[01:27] <qman__> so if you have a /boot, /, and swap
[01:27] <leaf-sheep> giovani: Console? I came in here because I know people in here maintain servers all times (lot of HDDS).
[01:27] <giovani> ZFS support in linux isn't significant
[01:27] <qman__> so it's logical it would sit there on a large partition
[01:28] <leaf-sheep> I'm setting this up for HTPC.
[01:28] <giovani> leaf-sheep: understood -- but you can switch into the console during installation and check for yourself if it's active
[01:28] <qman__> ...one of my lines didn't show up
[01:28] <giovani> since you were unclear if it was actively writing to the disk
[01:28] <qman__> if you have a /boot, /, and swap
[01:29] <qman__>  /boot is 33%, / is the next 33%, swap is the final 33%
[01:29] <giovani> qman__: depending on the disk order, no?
[01:29] <qman__> even if /boot is a few megs, / is almost 2TB, and swap is a gig
[01:29] <qman__> yes
[01:29] <leaf-sheep> 100MB boot, Encryption --> LVM --> [2GB swap, 10GB root, Remaining for home]
[01:29] <qman__> oh, you're using LVM
[01:29] <qman__> that's why
[01:29] <giovani> and encryption :)
[01:30] <qman__> it'll probably take about two hours to do the LVM
[01:30] <leaf-sheep> Encryption, indeed. ;)
[01:30] <qman__> then a while for the encryption
[01:30] <giovani> nah, it shouldn't take that long
[01:30] <giovani> I just did a new 1.5TB drive on LVM
[01:30] <giovani> took 10-15 min
[01:30] <qman__> wow
[01:30] <qman__> that's quick
[01:30] <qman__> but yeah, the process is generally slow, just let it do its thing
[01:30] <giovani> and ... why speculate
[01:30] <qman__> if it's not done in a few hours, then something's wrong
[01:30] <giovani> just check yourself if there's disk io
[01:30] <giovani> there's no need to wait and guess
[01:31] <leaf-sheep> Oh I'm installing Ubuntu Minimal from USB -- Would that cause issues?
[01:31] <giovani> not during disk partitioning
[01:31] <qman__> no
[01:31] <PhotoJim> shouldn't.
[01:31] <giovani> why are we still speculating?
[01:31] <leaf-sheep> There are no official minimal usb other than UNR.
[01:32] <giovani> leaf-sheep: please switch into the console and run iostat
[01:32] <giovani> this is will answer your question
[01:32] <leaf-sheep> Is it possible to obtain "Release file" for my local apt-mirror?  Getting base packages from official us.archive.ubuntu.com is long. I set up apt-mirror because I'm sick of waiting for packages and get failed installation every time.
[01:33] <qman__> I usually skip networking during the install to speed things up
[01:33] <leaf-sheep> giovani: I'm getting jaunty usb. Somehow I ended up with karmic version.
[01:33] <qman__> then update later
[01:33] <giovani> leaf-sheep: sure ... set up your sources.list -- and it'll work
[01:33] <giovani> leaf-sheep: that's not good -- karmic is not stable at all
[01:34] <leaf-sheep> giovani: I know. I clicked at the bottom and I assumed it was Jaunty. :<
[01:34] <qman__> but yeah, setting up 2TB that way is a slow process, and the progress bar does not indicate how far along each step is, just to which step you're at
[01:35] <leaf-sheep> Yeah, but to start off at 33% right away?
[01:35] <qman__> normal
[01:35] <qman__> your 100MB boot partition takes a split second
[01:35] <giovani> leaf-sheep: have you not been listening to what qman__ has been explaining?
[01:35] <qman__> also
[01:35] <qman__> I suggest you up that to about 256MB
[01:36] <qman__> you'll run out of room for kernel updates
[01:36] <qman__> just my opinion though
[01:36] <giovani> 100MB is standard, and fine, imo
[01:36] <giovani> why would you need more than 3-4 back kernels?
[01:36] <qman__> you don't, but it keeps them automatically
[01:36] <giovani> sure
[01:36] <qman__> it's just an annoyance
[01:36] <giovani> but by that logic, presuming you upgrade to every offered one ... you'll run out in short order anyway
[01:37] <PhotoJim> I agree to make it bigger, you won't miss 156 MB of disk space and it saves hassles if you don't punctually delete old kernels.
[01:37] <PhotoJim> not critical.  but convenient.
[01:37] <giovani> heh
[01:37] <giovani> but then he'll just run out later
[01:37] <giovani> same process, different timing
[01:37] <PhotoJim> significantly later.
[01:37] <qman__> yeah, but you don't have to clean out the old kernels as often
[01:37] <giovani> 1.5 times longer
[01:37] <PhotoJim> 2.5.
[01:37] <giovani> no, 2.5 total
[01:37] <leaf-sheep> giovani: I'm putting it on usb... almost done.
[01:37] <giovani> 1.5 longer
[01:37] <PhotoJim> 256/100 = 2.56 actually.
[01:37] <giovani> anyway
[01:38] <PhotoJim> 3 hours is 3 times as long as 1 hour, not 2 times as long.
[01:38] <leaf-sheep> qman__: I checked my laptop kernels where I kept them. It's in total of 46MB.  3 or 4 kernels there.
[01:38] <giovani> PhotoJim: but it's 2 times longER
[01:38] <giovani> 3 times AS lon
[01:38] <giovani> long*
[01:38] <PhotoJim> "times" = multiplication.
[01:39] <giovani> yes
[01:39] <PhotoJim> by definition.
[01:39] <giovani> longER
[01:39] <giovani> implies in addition to
[01:39] <PhotoJim> you're using it wrong.
[01:39] <qman__> I'ce got only one kernel on my router
[01:39] <PhotoJim> 256% as much time, but 156% additional, if you must.
[01:39] <qman__>  /boot is using 24M
[01:40] <qman__> like I said, 100M is enough, it's just my preference to use 256
[01:40] <qman__> won't miss that extra space and updates are less annoying that way
[01:40] <PhotoJim> my /boot hasn't been cleaned out in awhile, 64 MB
[01:40] <PhotoJim> that's only 5 kernels.
[01:41] <PhotoJim> oldest April 1/09.
[01:41] <leaf-sheep> giovani: What did you say about iostat? I'm at install prompt.  Do you mean for me to start a VT in boot process afterward?
[01:42] <giovani> leaf-sheep: I meant to go into the console during the installer, since you're curious what it's doing
[01:42] <giovani> you can do a ps aux
[01:42] <giovani> or if you have iostat you can run that
[01:42] <giovani> and both will make it clear what the installer is doing, if it's doing anything, or if it's hung
[01:42] <leaf-sheep> Console? That's CTRL + ALT + F[1-6] we're talking?
[01:42] <qman__> just alt
[01:42] <qman__> control is only necessary if you're in X
[01:42] <giovani> alt-f2 or f3, i forget which is which
[01:42] <giovani> in the installer
[01:42] <leaf-sheep> Okay. I'll set up everything right to the 33% part.
[01:43] <giovani> this lack of experimentation and hand-holding will get old quickly
[01:44] <leaf-sheep> giovani: Meaning I can set up /etc/apt/sources.list to my localmirror although I tried different mirror which lacks Ubuntu Release File (something) for verification?
[01:45] <leaf-sheep> I guess not.  Now it's getting all packages from server... again.
[01:46] <leaf-sheep> We'll sit tight. Please don't go anywhere. I'm frustrated and would love to get this done with. I have been trying this all night (last night) to this morning.  Fell asleep. Woke up and I'm at it again. :)
[01:54] <HellMind> how tf, i must port forward with UFW , I HATE it :(
[01:55] <qman__> heh
[01:56] <qman__> port forwarding with iptables takes four lines per port forward
[01:56] <qman__> I've got a nice loop in my script that reads from a file
[01:56] <jdstrand> HellMind: ufw does not support port forwarding with the command line interface. Please read the man page for what it supports. You may use iptables-restore style rules in /etc/ufw/before.rules to achieve anything iptables can do. See https://help.ubuntu.com/9.04/serverguide/C/firewall.html for details
[01:57] <HellMind> I hate that
[01:57] <jdstrand> HellMind: patches are welcome
[02:00] <HellMind> my patch will be rm -fr ufw
[02:01] <jdstrand> HellMind: you are free to use any firewall application you wish. if ufw does not suit your needs, try another listed in https://help.ubuntu.com/9.04/serverguide/C/firewall.html
[02:02] <HellMind> Im using 8.04
[02:02] <jdstrand> https://help.ubuntu.com/8.04/serverguide/C/firewall.html
[02:03] <HellMind> stop pasting that
[02:03] <jdstrand> *8.04*
[02:03] <HellMind> its the same doc -_-
[02:04] <oh_noes> is it possible to create a VM with two virtual hard disks in python-vm-builder?
[04:18] <rosa> Hi there, I have a ubuntu domain member with samba and is working fine for a while but lost the connection...the users can't see the samba folder and i need to restart the server to them can see again...
[04:18] <rosa> somebody know what can be the problem?
[04:35] <rosa> hi, somebody know if a have a domain member server which lost the connection with the domain server every 5 hours, what could be the problem?
[04:58] <jmarsden> rosa: I'm not at all an expert on SAMBA domain stuff, but look at the samba log files for clues.  If necessary, turn up samba logging so you get more info in the logs to work from.
[04:59] <rosa> i am not sure if the problem is that but i am getting this error: winbindd: Exceeding 200 client connections, no idle connection found
[05:00] <jmarsden> OK, sounds like you need to configure samba to allow more simultaneous winbindd client connections :)
[05:04] <leaf-sheep> I wonder if Ubuntu-Server.iso would benefit me more for pure XBMC edition -- with occasional samba and cups sharing, even its being a torrentbox?
[05:05] <jmarsden> leaf-sheep: https://help.ubuntu.com/community/ServerFaq#What%27s%20the%20difference%20between%20desktop%20and%20server?
[05:07] <leaf-sheep> jmarsden: I'm looking for something of minimal disc with at least all packages for ubuntu-minimal with it.  Minimal disc itself retrieve packages from the Internet.
[05:08] <leaf-sheep> jmarsden: I'm wondering if Server edition is what I'm looking for... as I'm sick of failed attempts and having to retrieve all packages from Internet more than 10 times. :|
[05:08] <jmarsden> Can't you do a minimal install from the "normal" Desktop install CD?
[05:09] <leaf-sheep> jmarsden: HPTC does not have cdrom. I'm doing this from Ubuntu and by Desktop install CD, it'll install full ubuntu-desktop.
[05:10] <PhotoJim> Alternate CD, I think, not Desktop.
[05:10] <leaf-sheep> PhotoJim: It still installs ubuntu-desktop.
[05:10] <jmarsden> I'll play in a VM...
[05:10] <ScottK> Basic server install is a lot smaller than a desktop install
[05:10] <leaf-sheep> Does Server edition installer supports LVM + Encryption?
[05:11] <leaf-sheep> LVM, yes. I know this for sure.
[05:12] <PhotoJim> leaf-sheep: I'm reasonably sure it has the option of doing a very basic installation.  if you can't get Ubuntu to do that, I know you can do that with Debian.  so it would surprise me that you couldn't with Ubuntu.
[05:14] <rosa> jmarsden, do you know how i can do that?
[05:14] <leaf-sheep> PhotoJim: I'm confidently sure that alternative disk is same as desktop disk minus the GUI, and plus the lvm + encryption.  That's all. I use it many times -- but yeah, they should prompt tasksel by default (and that's in minimal disc).
[05:14] <leaf-sheep> I think I'll remaster one in the future for ubuntu-minimal as default packages instead of ubuntu-desktop in the future for USB.
[05:15] <PhotoJim> leaf-sheep: alright.  can't say I've done it, so I'll defer to your experience.
[05:15]  * leaf-sheep says future twice... Silly me. :<
[05:16]  * leaf-sheep gets server edition because he's certain that ubuntu-minimal is installed by default.
[05:20] <twb> The alternative CD ought to be identical to the desktop CD in the set of packages installed.
[05:21] <twb> The difference is chiefly debian-installer vs. ubiquity + live CD
[05:21] <jmarsden> rosa: No... try googling that error message
[05:21] <twb> However the differences between ubuntu-server and alternative are the set of packages cached on the CD (so they don't have to be downloaded), and the tasksel and locale packages installed by default.  See the preseed/ directory on the server CD.
[05:22] <twb> There's no fundamental reason why you couldn't install either desktop or server using e.g. the mini.iso
[05:22] <jmarsden> rosa: Maybe try the ideas in http://magazine.redhat.com/2008/06/02/tips-and-tricks-i-get-the-error-winbindd-exceeding-200-client-connections-no-idle-connection-found/
[05:26] <jmarsden> Looks like the 200 is a compiled in value, see http://fixunix.com/samba/348340-samba-winbindd-exceeding-200-client-connections-no-idle-connection-found.html for some discussion.
[05:26] <leaf-sheep> twb: Slow network. I'm trying to install a HTPC from laptop (bridged).
[05:26] <twb> leaf-sheep: HTPC?
[05:26] <leaf-sheep> It takes awhile to get a base installation.  (Need a long cable to my room).
[05:27] <leaf-sheep> twb: Yes. Home Theater PC>
[05:28] <rosa> thanks, the weid is i dont have 200 users...that mean connectios? but still i think i dont more than 200 connetions
[05:29] <jmarsden> See the RedHat article for how to get it to list the connections...
[05:34] <rosa> thanks i will check it
[05:48] <rosa> l
[06:40] <cemc> can I somehow refresh the list in /dev/disk/by-label ?
[06:41] <jmarsden> Well, a reboot will most likely do that :)
[06:41] <jmarsden> Probably umount and the mount of the volumes concerned would also do it.
[07:12] <rosa> hi somebody know where i can configurate it ?WINBINDD_MAX_SIMULTANEOUS_CLIENTS ?
[07:24] <rosa> sombody here?
[08:13] <johe|work> good morning
[08:17] <soren> ttx: Any idea what might be causing this? http://launchpadlibrarian.net/30714354/buildlog_ubuntu-karmic-i386.eucalyptus_1.6%7Ebzr452-0ubuntu2_FAILEDTOBUILD.txt.gz
[08:18] <ttx> soren: looking
[08:18] <soren> It's during the java build, where it errs out with a:
[08:18] <soren>       [ERROR] Unexpected internal compiler error
[08:18] <soren> java.lang.StackOverflowError
[08:18] <ttx> soren: strange, it built on lpia
[08:19] <ttx> even stranger, the GWT stuff in compiled arch:all
[08:19] <soren> hm?
[08:19]  * ttx compares build logs, just a sec
[08:23] <ttx> soren: that's strange... This should behave the same whatever the arch, it's some java compile and even the JARs used are arch:all
[08:23] <ttx> soren: so apart from a i386-specific openjdk-6 issue...
[08:24]  * soren is tempted to just retry the build
[08:25] <ttx> soren: that idea crossed my mind as well :)
[08:25] <soren> but first, I'll see if I can reproduce it locally.
[08:25] <ttx> soren: could you reproduce it ...
[08:25] <ttx> you read my mind
[08:25] <ttx> soren: hmm. http://extjs.com/forum/showthread.php?t=73875
[08:29] <ttx> soren: If you can reproduce it locally, I'll PPA this fix in GWT so that you can confirm it fixes the issue: http://code.google.com/p/google-web-toolkit/source/detail?r=5262
[08:29] <soren> ttx: So... Do I need to patch gwt or can I just pass the greater stack size during the eucalyptus build?
[08:29]  * soren tries it locally now.
[08:29] <ttx> The first solution is the fix, the second solution is the dirty workaround
[08:30]  * soren nods
[08:30] <ttx> the fact that the error triggers only on i386 tends to prove you're at the trigger limit anyway
[08:30] <ttx> since there is probably nothing arch-specific tere
[08:30] <ttx> there
[08:35] <ttx> soren: let me rephrase, the second solution is not a dirty workaround. It's perfectly acceptable
[08:35] <ttx> soren: current GWT uses more stack than it should. Increase stack size to make it compile is perfectly ok.
[08:36] <ttx> soren: not sure you have easy access to JVM parameters in the build though... so fixing GWT so that it's leaner might be a better solution.
[08:37] <soren> ttx: Well, there's an ant xml snippet in that extjs thread.. I just don't know where to put it.
[08:37] <soren> ttx: Could you try sticking that gwt patch in your ppa?
[08:37] <ttx> sure, I'm on it
[08:37] <soren> I could not reproduce it locally, by the way.
[08:38]  * soren tries it again, just for kicks.
[08:38] <ttx> heh, sounds like retrying the build might work, then :)
[08:39] <ttx> soren: could you file a quick bug against GWT, so that I reference the fix ?
[08:39] <soren> sure.
[08:41] <ttx> soren: Is there any point in PPAing it, if you can't test the fix locally ?
[08:42] <soren> bug #418022
[08:43] <soren> ttx: Well... I could try uploading it to my ppa to test it.
[08:43] <ttx> ok
[08:43] <soren> ttx: To be honest, at this point, I don't see any particular reason to be so careful.
[08:43] <soren> ttx: Just upload it, I'll retry the build, and we'll see how it goes.
[08:43] <ttx> ok
[08:44] <soren> We've got plenty of time to fix stuff .
[08:44] <soren> It succeeded again locally, it seems.
[08:44] <soren> Yup, just finished.
[08:45] <soren> ttx: Ok, it seems only axis2c and rampart are missing MIRs.
[08:46] <soren> I'll see if I can get someone to file it for me. I'd like to get crackin' on the Elasticfox thing.
[08:46]  * twb wonders when -server became -devel
[08:46] <twb> or -motu or whatever
[08:46] <soren> twb: It always was :)
[08:47] <soren> twb: Up until a couple of years ago, it was /only/ about server development.
[08:47] <twb> Ha
[08:48] <ttx> twb: we just don't use the channel as much as we should.
[08:48] <soren> twb: We've just not been very good at using it for development stuff.
[08:48] <soren> Heh :)
[08:53] <ttx> soren: your fix is in.
[08:55] <kwork> yes use it for development then i can scam off all the knowledege
[08:55] <maswan> Use it for development to fix all my pet issues, even better! ;)
[08:56] <kwork> so any services use upstart with karmic ?
[08:58] <roxy09> hi there, i want to upgrade my version 8.04 to 9.xx..is possible to do easy or i need to install and configurate everything again?
[08:59] <twb> roxy09: upgrading in-place is a well-supported model.
[08:59] <twb> Unlike those RHEL idiots...
[09:00] <roxy09> jaja
[09:00] <roxy09> sorry haha (in english)
[09:00] <roxy09> what is the command to upgrade to the last version?
[09:00] <\sh> roxy09: do-release-upgrade ?
[09:00] <mattt> twb: it's possible on RHEL, just not commonly done :)
[09:01] <roxy09> thanks :)
[09:01] <twb> mattt: it's not SUPPORTED on RHEL
[09:01] <kwork> roxy09, just did it few days ago from 8.04 to 8.10 and then to 9.04 worked like a charm :)
[09:02] <kwork> thou one thing confuzed me, is 9.04 lts ?
[09:02] <\sh> kwork: nope
[09:02] <leaf-sheep> !lts | kwork
[09:02] <kwork> ah okey then no confuzion
[09:04] <roxy09> my problem is i am not sure if the bugs that have 8.04 is solved in the 9.04 or i will have more problems with this version?
[09:04] <kwork> you need to consult launchpad on that
[09:04] <kwork> find your bug id and see where its fixed
[09:04] <kwork> or whats the bug status
[09:04] <roxy09> i have samba bugs
[09:04] <kwork> im not bug database
[09:05] <roxy09> i need to install the vesrion 3.4
[09:05] <kwork> search the db
[09:05] <kwork> !llaunchpad
[09:05] <kwork> !launchpad
[09:05] <kwork> !bug tracker
[09:05] <kwork> !bug
[09:06] <roxy09> dont stress!!! i asked u before u said that...
[09:09] <kwork> no stress here yet :P
[09:26] <roxy09> somebody know how i can replicate and configurate ldap server?
[09:27] <roxy09> i mean condigurate a ldap backup server
[09:45] <twb> roxy09: "configure"
[09:46] <roxy09> yes
[09:51] <roxy09> somebody know how i can configure a ldap backup server?
[09:58] <roxy09> hi i try to change the version to 8.04 to 9.04 but i cant, somebody know the command...i try before with do-release-upgrade
[10:12] <roxy09> somebody know why i can upgrade my version, when i do old me i have the last version but still i have the 8.04
[10:13] <\sh> do-release-upgrade -d
[10:13] <\sh> brb
[10:19] <AlexC_> morning,
[10:20] <AlexC_> SSH has suddenly stopped/crashed on a Lenny server of mine, of which is 200 miles physically away from me. I have zero access to this server now, as even the Dell Remote Access Controller has given up (given XML parse errors). Is there any way you can think of that I can purposely crash this server to get it to restart?
[10:20] <AlexC_> sorry it's non-ubuntu, thought there would be someone with a clever idea, though no worries if it's off-topic I'll leave it
[10:21] <twb> AlexC_: call your colo guy and get him to kick it
[10:22] <AlexC_> see, thing is - the keys to the server are also sitting to my left :P
[10:22] <twb> AlexC_: even the power cable?
[10:22] <roxy09> thanks sh, do you know how long take it?
[10:22] <AlexC_> twb, true, true - that'd work
[10:23] <twb> AlexC_: or maybe serial cable to a box they have that you CAN ssh into and run screen /dev/ttyS0 on
[10:23] <roxy09> also, if i do that what happen with my kubuntu ?
[10:24] <AlexC_> twb, will have to see if they can do that. I'm not even sure if our guy will be in the area today, if not - any sane way of crashing it?
[10:25] <twb> AlexC_: is it running PHP?
[10:26] <AlexC_> twb, yes, though scripts are running as their own user - and the downside to me being security concious, there are no exploits that I could use to bring it down
[10:26] <twb> AlexC_: none you know of, anyway
[10:26] <AlexC_> ;)
[10:28] <maswan> roxy09: do-release-upgrade is roughly the same as the graphical upgrade, and the time depends mostly on the number of packages you have installed
[10:29] <twb> And the number of updates, and the speed of the mirror
[10:32] <roxy09> :) thanks a lot mates!
[10:35] <roxy09> meanwhitle...i am having another problem with amavis is using a lot of CPU, somebody know about that?
[10:39] <twb> Isn't that amavis' job?
[10:43] <roxy09> yes, but 99% od the CPU and stop the job of the mail services
[10:48] <soren> ttx: ..rebuilding eucalyptus on i386.
[11:03] <soren> ttx: \o/
[11:35] <roxy09> hi somebody know what devecop does?
[11:37] <Alblasco1702> Hallo roxy09 did you mean dovecot?
[11:38] <roxy09> yes
[11:38] <roxy09> i am upgrading ubuntu, should keep the conf files or let upgrade the files?
[11:38] <Alblasco1702> dovecot is a IMAP server
[11:40] <Alblasco1702> roxy09 i make a backup from my configuration files en then let the files upgrade but that's up to you
[11:53] <roxy09> hi i am getting error with spammassasin...when i do restart show me doesn find some files, i try to reinstall but still the problem, show me cant find some lib with perl
[12:39] <roxy09> hi there, i am having problme to retsart my spamassasin aplication, i got the error:  Can't locate MLDBM/Sync.pm
[12:40] <kwork> install the lib ?
[12:40] <kwork> or perl module even
[12:44] <roxy09> which one...i try to instal MLDBM but said i have the last one
[12:46] <kwork> http://cpansearch.perl.org/src/CHAMAS/MLDBM-Sync-0.30/Sync.pm
[13:31] <roxy09> hi i got some problem with amavis-new is ussing s lot of CPU, somebody know about this problem?
[13:48] <leaf-sheep> giovani, qman__ PhotoJim: Yoohoo. It's just me and installer's lack of feedback. I tried it again and watch few shows. Came back and it was done. Lulz. ;3
[13:59] <roxy09> :S I cahnge my version to ubuntu 9.04 and now samba doesn't run...
[14:00] <kwork> you should really get people to upgrade your box who knows what they are doing :P
[14:00] <kwork> and on the subject i upgraded from 8.04 to 8.10 -> 9.04 and my samba works just fine
[14:04] <roxy09> thanks !!!
[14:05] <Sky[x]> #samba ? :D
[14:17] <ivoks> could someone look at 416970?
[14:17] <ivoks> but 416970
[14:17] <ivoks> lol... bug 416970
[14:19] <roxy09> im getting this error: final write to client failed: Broken pipe
[14:19] <roxy09> somebody know about that?
[15:12] <garymc> Hi. Anyone know how I replace a faulty hard drive? Im using HP proliant G3 with 6 hotswap scsi drives in each. One gone faulty
[15:12] <Faust-C> garymc, are you using mdadm?
[15:12] <garymc> i wouldnt know?
[15:12] <Faust-C> well unless youre using some kind of raid i don't know what to tell you
[15:13] <garymc> how would i know. what is mdadm
[15:13] <garymc> ?
[15:13] <Faust-C> mdadm is linux's built in software raid
[15:13] <ball> garymc: on a machine like that you should have someone to hand who knows about these things.
[15:13] <ball> Are you responsible for the care and feeding of this beast?
[15:13] <garymc> yep, would should could, cant afford them
[15:14] <garymc> so trial and error for me
[15:14] <Faust-C> garymc, well normally i would agree w/ the "trail and error" method
[15:14] <ball> ...I hope it's nothing mission-critical then.
[15:14] <Faust-C> but its very hard to help someone whom is not familiar w/ the system in question
[15:15] <Faust-C> but if its not mission critical ill try to help
[15:15] <garymc> cool
[15:15] <garymc> it aint mission critical yet
[15:15] <Zerosan> Hello
[15:16] <garymc> so need to know how to do it for when it is
[15:16] <garymc> see my hard disk was displaying a red light on the bay. so normally means some sort of fault
[15:17] <garymc> I took it out put another in its place system didnt work
[15:17] <garymc> so replaced the faulty disk back in the bay restarted the server
[15:17] <ball> The the machine continue to run when you pulled the drive?
[15:17] <garymc> No i shut it down first
[15:17] <Faust-C> garymc, do you know if you are using any kind of raid or LVM
[15:17] <garymc> when i restarted it said press F1 to recover hard disk, i did and now no light showing a fault
[15:18] <Zerosan> Can someone help me with scanner sharing using sane? I'm using 9.04 and the scanner can already be found in the network, but when I try to access it using xsane or scanimage as a client, I always get the message that the Access has been denied.
[15:18] <ball> garymc: when you powered it back up, did the RAID controller tell you it was rebuilding the array?
[15:18] <garymc> Im using RAID 5
[15:18] <garymc> ball: no it never
[15:18] <garymc> Ball: it said recover hard disk data or something
[15:18] <ball> garymc: when you power on, do you see a message to the effect of "hit F8 for RAID configuration"?
[15:18] <Faust-C> Zerosan, is the scanner setup to allow remote connections
[15:18] <ball> (may be some other key than F8)
[15:19] <garymc> yes
[15:19] <Zerosan> Faust-C: yes it is, .... wait
[15:19] <garymc> i do
[15:19] <Faust-C> Zerosan, also check the logs and see what they say
[15:19] <ball> garymc: did you press that key?
[15:19] <garymc> with new hard drive in?
[15:19] <Zerosan> cat /var/log/syslog | grep sane says that access to my client has been granted
[15:19] <Zerosan> "cat /var/log/syslog | grep sane"
[15:20] <garymc> i did with new hard drive in and it said there where no configured logical drive did i want to create one
[15:20] <ball> garymc: yes, power down the system, insert the new hard disk, power the system on and hit that key to enter the RAID Array setup program
[15:20] <garymc> ok i did that and (ABOVE) is what it said
[15:20] <Zerosan> Faust-C: "cat /var/log/syslog | grep sane" tells me that the access has been granted to my client
[15:20] <ball> garymc: you may have lost all the data on your array then, unless someone set it up to use software RAID for some reason.
[15:20] <garymc> I then shut down server and put faulty one back in. I thought maybe that would have killed the whole system
[15:21] <Faust-C> Zerosan, hmm and you still cant scan
[15:21] <garymc> once i put old faulty drive in it worked fine after data recovery
[15:21] <Faust-C> funny part is i was working w/ a client the other day that had a scanner that works flawlessly w/ linux
[15:21] <Zerosan> Faust-C: yes, what bothers me is the fact that there is no group with the name "scanner"
[15:22] <Faust-C> Zerosan, maybe create that group and try it that way
[15:22] <Zerosan> Faust-C: I'll do that now, brb
[15:22] <Faust-C> Zerosan, i havent had much luck w/ linux and network scanners/printers
[15:22] <ball> garymc: Assuming your data is backed up, I'd wipe that box and start from scratch.
[15:22] <Faust-C> kk report back success or failure
[15:22] <garymc> lol
[15:22] <Faust-C> garymc, no seriously
[15:23] <Faust-C> that way you can set it up correctly
[15:23] <garymc> ball: What would i do when im using the sytem for real and a hard drive goes?
[15:23] <garymc> setup the raid?
[15:23] <Faust-C> well raid will be setup before hand
[15:23] <Faust-C> then when hdd dies you will know how to use mdadm to replace the drive
[15:23] <garymc> yes its setup, so why would i start from scratch?
[15:23] <ball> garymc: I would run tests before you put it into production
[15:23] <Faust-C> garymc, because you need to be able to fix it w/o any help
[15:23] <ball> Faust-C: Is mdadm just for software RAID though?
[15:24] <garymc> ball: thats the plan
[15:24] <Zerosan> Faust-C: I just created the group: "scanner" and added the user: "saned" to it. Still, no go | getting the message "access denied"
[15:24] <ball> garymc: I'd try using the RAID array setup program to build an array.  How large are your drives?
[15:24] <Faust-C> ball, yes, depending on the raid card it would be better to use built in raid and not fake raid
[15:24] <Faust-C> Zerosan, what is the make/model of the scanner in question
[15:25] <garymc> ball: I have 146gb and 72gb drives
[15:25] <Zerosan> Faust-C: Doesn't matter so much, scanning works flawlessy when I try scanimage as SU on the server
[15:25] <ball> Faust-C: he's using SCSI and an HP ProLiant I think, so odds are good it's a real RAID controller
[15:25] <garymc> 72Gb seem to go faulty
[15:25] <garymc> but now its showing ok?
[15:25] <ball> garymc: how many of the 146 Gbyte drives do you have?
[15:25] <Zerosan> Faust-C: but not over the network or as a normal user
[15:25] <Faust-C> Zerosan, hmm odd
[15:25] <Zerosan> Faust-C: yup
[15:25] <garymc> only one in this sytem
[15:25] <Faust-C> ball oh yeah didnt think of that
[15:26] <ball> garymc: how many of the 72 Gbyte?
[15:26] <Faust-C> Zerosan, .... man i cant think of what would cause it not to work
[15:26] <garymc> ball: 5
[15:26] <Faust-C> login and back out w/ user you just added to scanner group
[15:26] <ball> garymc: will your RAID controller let you build a RAID 1+0 array?
[15:26] <Zerosan> Faust-C: k, wait
[15:26] <Faust-C> and if al else fails submit help request to forum
[15:27] <garymc> yes, but i heard that wasnt best for my system?
[15:27] <garymc> not that thats true though
[15:27] <Faust-C> garymc, raid 10 is general purpose imo
[15:27] <Zerosan> Faust-C: I restarted saned, now it works, hehe | thanks for the help
[15:27] <Faust-C> lol ok
[15:27] <Faust-C> Zerosan, if you could document your success for others
[15:28] <garymc> so you think I should run all my servers on RAID 1+0
[15:28] <Zerosan> or not xD
[15:28] <Faust-C> i, at least, would greatly appreciate it
[15:28] <Zerosan> Faust-C: didn't work.... *sob*
[15:28] <Zerosan> Faust-C: Only difference now is that I can see the list in xsane, but when I select one scanner and press okay, still getting access denied
[15:28] <ball> garymc: That depends on the machine in question.  If your data fits on a 72 Gbyte disk, you could just mirror three of those.
[15:29] <Faust-C> Zerosan, heh
[15:29] <ball> ...or mirror two and have a third as a hot standby
[15:29] <garymc> yeah, like a backup drive?
[15:29] <giovani> xsane, on a serveR?
[15:29] <Zerosan> giovani: no, sane on a server | xsane on the client
[15:29] <giovani> ah
[15:30] <Zerosan> giovani: to be precise saned
[15:30] <giovani> scan-to-server saves everyone's lives :)
[15:30] <ball> garymc: not a backup drive, but a drive that the server can use in the event that one of the other drives fails
[15:30] <garymc> whats the diff between RAID 5 and RAID 1+0
[15:30] <Zerosan> right now it only hurts :P
[15:30] <ball> (backups are something else)
[15:30] <Zerosan> giovani: do you know how to make it work properly?
[15:30] <garymc> ball: ok im not sure how I would do that
[15:30] <_ruben> garymc: a lot of performance
[15:30] <ball> RAID 5 requires three drives.  RAID 1+0 requires 4, but is slightly faster.
[15:30] <giovani> garymc: massive difference -- google has lots of answers -- it's not simple to explain quickly
[15:31] <garymc> but sounds like thats what I should do
[15:31] <ball> ...and arguably a little more resiliant.
[15:31] <ball> garymc: do you have a manual for your RAID controller?
[15:31] <garymc> nope
[15:31] <giovani> RAID10 is less efficient with space
[15:31] <garymc> I bought the servers second hand
[15:31] <_ruben> raid5 is nice for bulkstorage (fileserver) .. raid10 is good for performance (virtual machines/database/etc)
[15:31] <ball> garymc: how large is your data set?
[15:31] <garymc> not so large yet
[15:31] <giovani> RAID5 'wastes' 1/Nth of the space you have -- where N is the number of disks
[15:32] <giovani> RAID10 'wastes' half of the space you have
[15:32] <ball> giovani: it's not wasted ;-)
[15:32] <giovani> ball: hence the ''
[15:32] <ball> It's used to save your ass ;-)
[15:32] <_ruben> raid5 'wastes' your performance as well ;)
[15:32] <giovani> I thought that was very clear
[15:32] <_ruben> (when writing, not when reading)
[15:32] <garymc> infact i think its the other way around with disk space
[15:32] <Zerosan> why not use the word "reduces" instead of waste?
[15:32] <garymc> lol
[15:32] <ball> How about "uses"
[15:32] <giovani> garymc: no ... it's not the other way around
[15:33] <garymc> RAID 5 i have less space than with RAID 1+0
[15:33] <giovani> false
[15:33] <Zerosan> or "takes up"
[15:33] <ball> garymc: if you're on 72 Gbyte drives, space is probably a non-issue for you
[15:33] <ball> ...or you would be upgrading to larger disks.
[15:33] <Zerosan> now, does anyone have expierence with saned on ubuntu-server 9.04?
[15:33] <giovani> if you have 3 1TB drives in RAID5, 1TB will be used for parity, and 2TB will be usable
[15:33] <ball> (which might be an idea anyway)
[15:33] <garymc> yes, well I got someone picking up 6 146gb scsi right now
[15:34] <giovani> if you use 4 1TB drives in RAID10, 2TB will be used for mirroring, 2TB will be usable
[15:34] <garymc> which i will use for my asterisk server
[15:34] <garymc> thats another story
[15:34] <garymc> :)
[15:34] <ball> garymc: Also be aware that every disk you put in the server adds to the heat inside the case.
[15:34] <garymc> ahh right ok
[15:34] <ball> ...so use the minimum number that gives you the reliability and performance that you need.
[15:34] <garymc> i got them in an air conditionded room
[15:34] <giovani> honestly
[15:34] <giovani> there are plenty of studies
[15:35] <ball> garymc: that doesn't matter ;-)
[15:35] <giovani> that show that slightly hotter temps don't affect drives longevity
[15:35] <garymc> no?
[15:35] <giovani> so if the server was properly cooled with 2 drives, adding 4 more drives won't be a problem
[15:35] <giovani> this obsession with super-cooling servers is wasteful
[15:35] <garymc> not my electricity so not too bothered
[15:35] <garymc> anyway I need to change or check a faulty disk drive
[15:36] <giovani> well is the cooling auto-adjusted?
[15:36] <garymc> whats the best way to go about it?
[15:36] <ball> giovani: heat is the enemy of reliability though, not just for disk drives but for RAM and processors too.
[15:36] <giovani> ball: that's a concept that's been taken beyond reality
[15:36] <ball> giovani: not the way I do it.
[15:36] <giovani> yes, in extreme temperatures, longevity/reliability are issues
[15:36] <garymc> giovani: if you mean my Air con then yes, it keeps the room and 16 degrees
[15:36] <garymc> *at
[15:36] <giovani> but the suggestion that adding a few drives will in any significant way affect their longevity, is false
[15:37] <ball> 16C seems excessively cold
[15:37] <ball> Anyway
[15:37] <ball> garymc: build a RAID array using the utility on the RAID controller card.
[15:37] <giovani> 16C is FAR below standard
[15:37] <garymc> well.... not a prob right now
[15:37] <_ruben> garymc: find out the type of raid controller and then go find a linux managment tool for it
[15:37] <giovani> I would not recommend wasting energy that way
[15:37] <ball> garymc: then test that array.
[15:38] <giovani> 22-23C is standard in most DCs
[15:38] <garymc> is a raid array like create a new logical drive? etc?
[15:38] <giovani> and even that is often not required
[15:38] <ball> 23C sounds about right
[15:38] <giovani> 72F is standard here
[15:38] <ball> garymc: some RAID controllers refer to it that way
[15:38] <garymc> yeah i can set it to 23C
[15:38]  * ball breaks out the xcalc to do F->C
[15:38] <giovani> the issue has more to do with airflow management
[15:38] <giovani> and less to do with temperature
[15:38] <garymc> right ok
[15:38] <giovani> google runs their datacenter about 10-15 degrees hotter than "standard"
[15:39] <giovani> because they know how to manage airflow
[15:39] <garymc> so build an array using RAID 1+0
[15:39] <ball> giovani: 72F ~= 22C
[15:39] <ball> ...so that sounds good.
[15:39] <ball> garymc: try RAID 1 with three disk drives, if it will let you.
[15:39] <garymc> I wonder how big googles data centre is
[15:39] <garymc> ok
[15:39] <ball> garymc: they have several
[15:39] <garymc> what about sparE?
[15:40] <giovani> heh, 'several'
[15:40] <giovani> they have dozens
[15:40] <ball> garymc: RAID 1 with three disk drives writes the same data to three drives.  Up to two drives can fail before you lose all your data
[15:40] <garymc> yep they are worth alot of dosh now
[15:40] <ball> giovani: dozens == several
[15:40] <garymc> and if one drive fails?
[15:41] <giovani> then you replace it
[15:41] <ball> garymc: if one drive fails the array controller should notify you of that
[15:41] <garymc> how?
[15:41] <giovani> by taking out the old one
[15:41] <giovani> and putting in a new one
[15:41]  * giovani smacks head
[15:41] <ball> ...you replace it and tell the RAID controller to rebuild the array
[15:41] <garymc> wball: when the system bootS?
[15:41] <ball> garymc: ideally you should run software that talks to the RAID controller and asks it about the health of the array.
[15:41] <garymc> so a drive fails, while system is up. I see red light on drive
[15:42] <garymc> ok so what software can i download in ubuntu server?
[15:42] <garymc> and run in the gui?
[15:42] <giovani> there is no gui
[15:42] <giovani> if your raid controller supports hotswap
[15:43] <giovani> you replace the drive, while the system is on
[15:43] <garymc> hmmm?
[15:43] <giovani> unless you *want* to take the system offline
[15:43] <ball> garymc: You *may* be able to ask the RAID controller to rebuild the array without bringing the server down.
[15:43] <garymc> ok just to let you know Im running LTSP
[15:43] <garymc> hmm ok
[15:44] <ball> ...if not, then you'll have to reboot, hit that key to get into the software on the RAID card and then rebuild the array that way
[15:44] <giovani> if your raid controller supports hotswap (all decent ones do)
[15:44] <ball> (which shouldn't take long on a 72 Gbyte disk)
[15:44] <giovani> you don't take the system offline
[15:44] <ball> giovani: right, I'm allowing for a possible lack of Linux tools for his RAID controller)
[15:45] <garymc> rebuild mean create new logical drive?
[15:45] <ball> garymc: no.
[15:45] <garymc> cos I will try that now
[15:45] <garymc> ahh ok
[15:45] <giovani> ball: well then I'd ask/investigate the controller, rather than assuming
[15:45] <ball> giovani: we've asked already ;-)
[15:45]  * ball asks again
[15:46] <Steve[mbp]> morning everyone!
[15:46] <ball> garymc: when you power up your system, do you see a line of text that says "HP SmartArray" something-or-other?
[15:46] <garymc> yes
[15:46] <ball> garymc: please type here the entirity of that line.
[15:46] <garymc> ok ill be back in a sec with that
[15:47] <ball> giovani: in fairness, I'm also not used to Linux, so I don't know how functional the available tools for a SmartArray controller are.
[15:48] <ball> At a minimum I'd like them to tell me about the health of my array.
[15:48] <ball> It would be *really* nice if I could also ask it to rebuild the array in the event of a drive failure.
[15:48] <garymc> right here we go
[15:49] <garymc> slot 0: HP SMART ARRAY 5i Controller ........... initializing
[15:49] <ball> garymc: hopefully someone here knows about Linux tools for those.
[15:49] <garymc> ok. Does the line give any indicatioins?
[15:50] <ball> garymc: Yes, it tells us what your RAID controller is.
[15:50] <ball> (HP Smart Array 5i)
[15:50] <ball> ^- you may want to write that down.
[15:50] <maswan> ball: You want hpacucli from the hp websites to manage arrays from the command line in the OS
[15:51] <garymc> ball: is it any good?
[15:51] <ruben23> hi
[15:51] <leaf-sheep> Hi alligators. :)  Is it possible to mirror from a website using rsync --no-parent (eg, getting all images on the said directory)?
[15:52] <ball> maswan: good to know, thanks.
[15:52] <Faust-C> leaf-sheep, look into using wget for that
[15:52] <Faust-C> wget has a web spider feature iirc
[15:52] <ball> garymc: I don't know, but I imagine it's adequate.
[15:53] <leaf-sheep> Faust-C: I already have some local files and when I run wget, I see that it overwrite the files instead of checking to see if it's same.
[15:53] <Faust-C> leaf-sheep, iirc there are options to _not_ overwrite
[15:53] <ruben23> i have installed samba on my ubuntu desktop trying to access a windows workgroup..but get error when trying ot access windows pc on gui--stating unable to mount network location.
[15:55] <pmatulis> ruben23: gvfs should handle that, shouldn't need samba
[15:56] <ruben23> pmatulis: whats that...?
[15:56] <pmatulis> ruben23: Places (in the Panel)
[15:57] <ruben23> pmatulis: can you give guide on it how do i do it..
[15:59] <pmatulis> ruben23: sure, Places > Connect to Server
[16:00] <ruben23> then..? thats all--->ill be able to connect to my windows client PC..?
[16:00]  * pmatulis is taking the risk of getting swatted by the local server nazis
[16:08] <leaf-sheep> ruben23: Try "network://" in the Nautilus address bar.  That may be all you need.
[16:10] <gst> what is Nautilus?
[16:10] <ball> Gnome's file browser I think
[16:17] <roxy10> Hi there ..i can't to connect from my windows client to samba server, i got the error teh route form the machine doesn't found
[16:18] <roxy10> i update my version of ubuntu and i got this error, before was working ok, somebody know what could happen?
[16:19] <garymc> ball: so what should I do. Do you know?
[16:19] <roxy10> i really need help is 1 am and i need to resolve this problem ...please help!
[16:19] <ball> garymc: How far did you get?
[16:20] <garymc> i never
[16:20] <ball> garymc: do you have any document files (or other data that you want to keep) on this server now?
[16:20] <garymc> not really no
[16:20] <garymc> but it took me a while to setit all up
[16:21] <ball> garymc: set up in terms of software?
[16:21] <garymc> yes... easily done again
[16:21] <garymc> are you suggesting re install using RAID 1+0?
[16:21] <garymc> not RAID %?
[16:21] <ball> garymc: okay.  Are you using a different computer to access IRC?
[16:21] <garymc> *5
[16:21] <garymc> yes
[16:22] <ball> Good.  How many 72 Gbyte disks are in the machine now?
[16:22] <garymc> 5
[16:22] <garymc> +1 146
[16:23] <garymc> i could put 2 more 146 in instead
[16:23] <ball> garymc: save them for when your data outgrows the 72 Gbyte drives
[16:24] <garymc> right
[16:25] <garymc> so.......................
[16:25] <ball> gary: I'm thinking.
[16:26] <garymc> cool
[16:26] <ball> Does your RAID controller give you the option of using four disks to create a RAID 1+0 array?
[16:27] <garymc> i go and check now
[16:27] <ball> Thinking about it, if your data set is smaller than 72 Gbytes, I'd just shove two drives in there and mirror them.
[16:27] <garymc> you dont want me to use 6?
[16:27] <ball> (RAID level 1)
[16:27] <ball> why use 6 if your data will fit on two?
[16:27] <garymc> well im hoping my databse will become vast eventually
[16:27] <garymc> my data?
[16:27] <garymc> like the OS?
[16:27] <ball> garymc: Use two now and then upgrade as it expands.
[16:28] <garymc> well since i got loads of hard drives does it not make sense to just bung them all in?
[16:28] <ball> Remove three 72 Gbyte disk drives and use the software on the RAID card to create one 72 Gbyte logical drive from a pair of 72 Gbyte physical drives (RAID-1)
[16:29] <ball> garymc: no, that does not make sense.
[16:29] <garymc> lol ok
[16:29] <garymc> so i just bought loads of drives for no reason?
[16:29] <ball> garymc: spares == a good reason.
[16:30] <garymc> ok
[16:30] <ball> You may want to create a second RAID-1 pair to backup onto (in addition to the tape backups you'll hopefully be making)
[16:30] <garymc> you mean i need tape driveS?
[16:30] <ball> ...either way, by using two drives now, you're leaving some drive bays empty for when you outgrow that 72 Gbyte array
[16:30] <ball> garymc: a tape drive is a sensible investment.
[16:30] <jmarsden> ball: Or set up RAID1 with a hot spare , if you really have plenty of drives available?
[16:31] <garymc> i do have plenty of drives
[16:31] <ball> jmarsden: I did think about that.
[16:31] <garymc> ok
[16:31] <ball> jmarsden: I like the thought that he could sustain two drives failing at once.
[16:31] <garymc> so if I have say 4 drives and two hot spares?
[16:31] <ball> garymc: will your RAID controller let you create a RAID1 array with three disk drives?
[16:32] <ball> brb
[16:32] <ball> (phone)
[16:32] <garymc> hold on ill check
[16:32] <garymc> ok
[16:32] <roxy10> I can browse form windows to samba server, somebody know how i can detect which is the problem?
[16:34] <IvanCosta> Hi, guys.
[16:35] <IvanCosta> Does someone know the difference of ECC memory buffer and unbuffered?
[16:35] <jmarsden> garymc: I'd be looking to create a RAID1 array (two drives) plus a hot spare (third drive).  If you want another one for backup... duplicate that.  Now you have used 6 drives, which seems to be what you want to do anyway :)
[16:35] <jmarsden> IvanCosta: Different technology.  Use the one your motherboard supports.  One has buffers on the modules, one does not...
[16:37] <garymc> Ball: It will only let me create RAID 5 or RAID 0 with 3 drives in
[16:37] <giovani> garymc: no, RAID5 requires 3+ drives
[16:38] <giovani> RAID0 requires 2+ drives
[16:38] <giovani> RAID1/RAID10 require 2+ drives in even increments; 2,4,6 etc
[16:38] <garymc> what does RAID 1+0 require?
[16:38] <jmarsden> garymc: 4 drives
[16:38] <IvanCosta> Jmarsden: If I used one by another, will break the machine?
[16:38] <giovani> err, RAID1 is 2 drives only rather, RAID10 requires 4,6,8,etc
[16:38] <garymc> ok so Bill is on the phone at the minute
[16:39] <jmarsden> IvanCosta: Probably not, it just won't work... I think!  But don't deliberately put the wrong kind of RAm in a motherboard, that's just not wise...
[16:39] <garymc> he was asking if my RAID array will let me create RAID 1 with 3 drives. The answer is no
[16:39] <giovani> garymc: no, RAID1 only works with 2 drives
[16:40] <jmarsden> garymc: OK.  You may be able to create ie RAID1 with 2 and then add a hot spare drive to it?
[16:40] <giovani> or just use RAID5
[16:40] <ball> giovani: usually.
[16:40] <garymc> so should I remove a drive and start again?
[16:40] <garymc> I only get 3 RAID options
[16:40] <garymc> RAID 5 - RAID 1+0 - RAID 0
[16:41] <jmarsden> Deselect one of the drives and pick RAID1.
[16:41] <IvanCosta> Jmarsden: Yes, you right. But a only have buffered memory and it's very difficult to find unbuffered...
[16:41] <ball> garymc: okay, so it looks as though your RAID controller wants to use three or four drives
[16:41] <giovani> selecting RAID1+0 may have RAID1 as a suboption
[16:41] <jmarsden> ball: He has 3 selected... :)
[16:41] <giovani> look into it
[16:41] <ball> garymc: doe jmarsden's suggestion work if you select just two drives?
[16:42] <ball> s/doe/does/
[16:43] <garymc> yes i can select RAID 1+0 and use one as a spare
[16:43] <ball> garymc: that's strange.
[16:43] <garymc> so TWO as RAID 1+0 and 1 as a spare
[16:43] <garymc> it is?
[16:44] <ball> RAID 1+0 requires four drives (minimum).  Hoepfully they've just mislabelled RAID-1 by putting it under the same heading.
[16:44] <ball> If it lets you pair two drives as RAID-1, with one spare... that's a sensible choice.
[16:45] <garymc> well i cant find no choice for RAID 1, unless your abbreviating RAID 1+0
[16:45] <ball> btw, that phone call was my boss saying he'd plugged something into the hub.  Unfortunately he plugged it into a port that he should never touch.
[16:45] <garymc> cos I cant find no RAID 1
[16:45] <garymc> ball: stupid bosses
[16:45] <garymc> :)
[16:45] <giovani> garymc: it's not going to offer you RAID1 with an odd number of drives selected
[16:45] <giovani> it shouldn't
[16:45] <ball> garymc: I think HP have just (confusingly) lumped RAID 1 and RAID 1+0 into one menu option.
[16:46] <giovani> this is so not an ubuntu issue though
[16:46] <kinnaz> you can create raid1 with one hot spare
[16:46] <kinnaz> so 3 drives :P
[16:46] <giovani> indeed
[16:46] <giovani> we've mentioned this a number of times
[16:46] <garymc> ok lets assume i have
[16:46] <ball> giovani and kinnaz are right
[16:46] <giovani> garymc: we've been hand-holding you this entire time
[16:46] <giovani> I think it should stop
[16:46] <giovani> call HP, use google, poke around the menus yourself
[16:46] <giovani> we don't know your server, this isn't ubuntu-related at all
[16:47] <ball> ...or /msg me.  We are tying up the channel.
[16:47] <ball> Thanks everyone for being so patient with us.
[16:47] <garymc> thanks
[16:47] <garymc> sorry
[16:47] <kinnaz> no worries
[16:53] <giovani> anyone here running ext4 in production?
[16:54] <sgsax> I hear it's stable
[16:54] <kinnaz> too sceard yet
[16:54] <kinnaz> ill let you guys try it out :P
[16:54] <giovani> heh, I've been running it in testin for a few weeks
[16:54] <kinnaz> for sake of the testing
[16:54] <giovani> but was talking in #ubuntu+1 about how data=ordered might have severe consequences for power loss
[16:54] <kinnaz> try to poweroff the box
[16:54] <giovani> testing*
[16:54] <Sam-I-Am> mathiaz: you around?
[16:55] <kinnaz> giovani, so have you tested it with powerloss ?
[16:55] <giovani> kinnaz: not yet
[16:55] <giovani> I'll need to set up a new box for that
[16:55] <giovani> it seems data-ordered is the issue to be concerned with, and can be changed
[16:56] <kinnaz> i havent really dig into it, but is there noticebale performance increase ?
[16:56] <mathiaz> Sam-I-Am: hey
[16:56] <Sam-I-Am> mathiaz: fixed the nssov bug
[16:56] <mathiaz> Sam-I-Am: \o/
[16:56] <giovani> kinnaz: there may be some -- there's the potential for online-defrag, which is a huge deal to me
[16:56] <mathiaz> Sam-I-Am: what was the issue?
[16:56] <giovani> the online-defrag code is still unofficial at this point though
[16:56] <Sam-I-Am> mathiaz: turns out it was calling an undefined symbol
[16:57] <kinnaz> i think i havent ever defraded ext partision, if fsck doesnot do it
[16:57] <mathiaz> Sam-I-Am: which library is missing?
[16:57] <kinnaz> defraged
[16:57] <Sam-I-Am> mathiaz: gave output from ld_debug to hyc... he fixed it in HEAD... quanah is going to commit the fix to 2.4.18.
[16:57] <mathiaz> Sam-I-Am: and how did you debug it?
[16:57] <Sam-I-Am> mathiaz: it was just a bad function call... something that was apparently renamed
[16:58] <Sam-I-Am> which oddly translated to 'file not found' as far as slapd's module loader was concerned
[16:58] <mathiaz> Sam-I-Am: ok - so it was upstream, rather than the packaging
[16:58] <Sam-I-Am> yes
[16:58] <mathiaz> Sam-I-Am: great - thanks for taking this up to upstream, debugging it and reporting back
[16:58] <Sam-I-Am> so the question is... do we want to wait for the fix in 2.4.18 or want me to submit patches for 2.4.17?
[16:59] <mathiaz> Sam-I-Am: good question - we'll probably wait for 2.4.18
[16:59] <mathiaz> Sam-I-Am: 2.4.18 will have support for disconnected mode in the pcache overlay
[16:59] <Sam-I-Am> ok
[16:59] <mathiaz> Sam-I-Am: which is something that was discussed at the last Ubuntu Developer Sumiit
[16:59] <Sam-I-Am> what should we do with the ubuntu bug report i filed?
[16:59] <mathiaz> Sam-I-Am: leave it open - I'll fix it with the upload of 2.4.18
[16:59] <Sam-I-Am> sure
[17:00] <mathiaz> Sam-I-Am: it's a bug in the current package and should be kept on track
[17:00] <Sam-I-Am> guess we're good then... however, i will get some patches going to fix the nssov build/cleanup
[17:00] <mathiaz> Sam-I-Am: if you could add a note to the bug stating that it was fixed upstream, with a link to the commit/ITS it would be helpful
[17:00] <mathiaz> Sam-I-Am: did you talk with upstream about this too?
[17:00] <Sam-I-Am> sure... not sure hyc got an its # yet, i'll ask
[17:01] <Sam-I-Am> mathiaz: well, the nssov makefile could use a  'clean
[17:01] <Sam-I-Am> er...
[17:01] <mathiaz> Sam-I-Am: ok - even if there isn't an ITS, a link to the upstream cvs commit could be helpful
[17:01] <mathiaz> Sam-I-Am: or at least a note that it has been fixed in 2.4.18
[17:01] <Sam-I-Am> 'clean' target... but also since the deb build doesnt copy the contrib tree into the build die, it doesnt remove the temporary build files
[17:01] <mathiaz> Sam-I-Am: just to keep track of where things are and where to look for if needed
[17:02] <Sam-I-Am> so i was thinking about adding a clean target to nssov... then in debian/rules, rm the temp files.. or call make clean (if thats possible)
[17:02] <Sam-I-Am> sure
[17:03] <mathiaz> Sam-I-Am: isn't a clean target in the Makefile + rm in debian/rules redundant?
[17:03] <mathiaz> Sam-I-Am: I though the two options were:
[17:03] <mathiaz> Sam-I-Am: 1. add clean to nssov Makefile and call make clean in debian/rules
[17:03] <Sam-I-Am> yeah it would be... i just didnt know if it was good practice to call another makefile directly from rules
[17:03] <mathiaz> Sam-I-Am: 2. Add rm in debian/rules
[17:03] <Sam-I-Am> depends on if we want to touch the nssov makefile or just debian/rules
[17:04] <mathiaz> Sam-I-Am: option 1. would be suited for inclusion in upstream source
[17:04] <mathiaz> Sam-I-Am: while option 2. would be debian specific
[17:04] <Sam-I-Am> yup
[17:04] <Sam-I-Am> your choice... i've done both
[17:04] <mathiaz> Sam-I-Am: let's option 1 and ask upstream about their opinion - if they take the patch, then go for 1.
[17:04] <Sam-I-Am> will do
[17:04] <mathiaz> Sam-I-Am: let's *try* option 1 and ask upstream about their opinion - if they take the patch, then go for 1.
[17:11] <Sam-I-Am> guess i'll file the its... dont see one yet
[17:11] <Sam-I-Am> hyc will just bless it
[17:17] <roxy10> hi i got his error with winbind ,Exceeding 200 client connections, no idle connection found ...somebody know about it?
[17:23] <zklaus> Hi there, I have kerberos problems and could use some help.
[17:24] <zklaus> I am trying to auth against my universities kdc and the kinit succeeds.
[17:24] <zklaus> But I fail to find the configuration that allows me to login.
[17:25] <zklaus> The auth.log says attempting authentication...
[17:25] <zklaus> success
[17:25] <zklaus> but then: FAILED LOGIN.
[17:25] <zklaus> Any ideas?
[17:50] <roxy10> somebody know what mean broken pipe?
[17:51] <KillMeNow> usually it means you're trying to Pipe something to another application
[17:51] <KillMeNow> you can do it where email comes in and "pipes" it to say RT
[17:51] <KillMeNow> other wise google the error message
[17:51] <KillMeNow> could mean something completely opposite than what i'm telling you
[17:55] <slestak> cjwatson: hiya man.  good weekend?
[17:55] <zklaus> roxy10: What's the full command, that led to that?
[17:55] <slestak> cjwatson: got some interesting (to me at least) putty testing input
[17:56] <cjwatson> slestak: ok ...
[17:56] <slestak> cjwatson: i am using your karmic package with 9.04.
[17:57] <slestak> cjwatson: i get visible artifacts when scrolling through text files with less, vi, and even when using my ERP system (ssh'd to aix, System Builder GUI toolkit)
[17:57] <slestak> the artifacts only occur when I use compiz,
[17:58] <slestak> I disables all effects, and they have gone away
[17:58] <slestak> i am using an nvidia card with the binary nvidia driver
[17:59] <slestak> does not occur with gnome terminal
[17:59] <slestak> vim is the strangest, because whole lines can disappear or reappear as you use j-k to go up and down in a file
[18:01] <slestak> can we refresh the package soon so I can get the Font Translation Selection patch included?
[18:02] <slestak> i should be able to do some comparisons to other jaunty machines.  I have another machine here that has an intel chipset, wonder if it shows the same issue.
[18:02] <slestak> its a netbook, so I do not use it nearly as much
[18:08] <slestak> cjwatson: i had to reset sth, sorry.  did i miss a post?
[18:10] <cjwatson> slestak: the artifacts really sound like Somebody Else's Problem, perhaps compiz's; I don't use compiz so I have no idea what might be going on there
[18:11] <cjwatson> slestak: I've got a refresh lying around on disk, planning to upload this week
[18:11] <cjwatson> thanks for reminding me
[18:13] <slestak> I've since turned it off on my workstation.  i had it on bc I love gnome-do.  should I post it somewhere in launchpad (Maybe Answers?) in case someone else runs into the issue, or if someone can corroborate this on another Nvidia machine
[18:14] <twint_> Boa tarde
[18:15] <twint_> alguem pode me ajudar?
[18:15] <KillMeNow> say what?
[18:15] <guntbert> !br | twint_
[18:17] <twint_> 	
[18:17] <twint_> like to know how to pop the apache and php and mysql
[18:17] <kinnaz> !lamp
[18:18] <twint_> what would lamp?
[18:19] <twint_> say what?
[18:19] <twint_> 	 	
[18:19] <twint_> I will investigate this link a Polco thank
[18:21] <roxy10> hi sorry, i lost the connection...my winbindd is show me this error. request location of privileged pipe
[18:25] <twint_> #ubuntu-br
[18:37] <Sam-I-Am> mathiaz: with all the recent problems reported on 17 by richton, quanah says 18 might not be ready by the feature freeze for karmic...
[18:41] <mathiaz> Sam-I-Am: right - I've emailed howard about that
[18:41] <mathiaz> Sam-I-Am: the question is when 2.4.18 could be released
[18:41] <mathiaz> Sam-I-Am: we could ask for a Feature Freeze Exception if 2.4.18 is released within the next two weeks
[18:50] <Sam-I-Am> mathiaz: ok... just figured id let you know
[18:50] <Sam-I-Am> mathiaz: got an ITS # for the nssov bug... updating ubuntu bug.
[18:54] <mathiaz> Sam-I-Am: thanks for your work on this.
[18:54] <Sam-I-Am> no problem
[18:54] <Sam-I-Am> i hang out in #openldap and #openldap-devel :)
[18:54] <mathiaz> Sam-I-Am: yeah - I should do the same - I'm on the mailing list
[19:51] <StefanWray> seeking info on cloning a hard drive with ubuntu server
[19:53] <andresmujica1> partimage, dd, dd_rescue, rsync
[19:55] <SJr> With intel speed step, my CPU seems to only ever be running at 2.0 GHz instead of 2.83 GHz, how do I clock it up.
[20:08] <PhotoJim> SJr: put it under load?  it might just idle at the lower speed to conserve electricity, and step up when load demand requires it.
[20:16] <SJr> I don't think so PhotoJim it's still staying there
[20:17] <SJr> oh there we go
[20:46] <Doonz> hey guys was wondering if someone could help me set up my nic's to be static and not dhcp. i only have command line interface so i need some guidance
[20:47] <VirtualDisaster> Doonz, if i point you to the proper doc will you use it?
[20:48] <Doonz> very much so
[20:48] <Doonz> rather read then be babied
[20:48] <VirtualDisaster> Doonz, which version? 8.04 LTS or juanty
[20:49] <Doonz> 9.04
[20:49] <VirtualDisaster> https://help.ubuntu.com/9.04/serverguide/C/index.html
[20:49] <Doonz> thanx
[20:49] <VirtualDisaster> it has a good article on setting static interfaces
[20:49] <VirtualDisaster> yw, and good luck
[20:51] <StefanWray> anyone use clonezilla
[20:52] <giovani> I've used it a few times, yeah
[20:54] <StefanWray> giovani: i'm intending to use clonezilla live off a usb flash drive. any issues i should watch for?
[20:55] <giovani> not that I know of
[20:55] <giovani> that's how I've used it
[20:56] <giovani> but I wouldn't necessarily remember everyhting
[20:56] <giovani> everything*
[20:57] <StefanWray> giovani: can you recommend best tutorial or instructions?
[20:57] <giovani> nope ...
[20:57] <giovani> why would you need a tutorial?
[20:58] <giovani> it's just formatting the usb stick, setting it bootable and copying files
[20:58]  * FastZ is away: Away from keyboard
[20:58] <StefanWray> ok, so it's pretty straightforward then
[20:58] <giovani> should be
[20:59] <StefanWray> thanks
[21:02] <trothigar> Hi, I'm having trouble installing jaunty x64 on virtual box on jaunty desktop (x64). It keeps freezing at 50% when scanning the mirrors.
[21:03] <giovani> trothigar: do you have networking properly set up?
[21:04] <trothigar> giovani, a NATed Network adaptor should do the trick shouldn't it?
[21:05] <giovani> trothigar: should ... if you have another NAT to the internet, it can get nasty, but often works
[21:06] <trothigar> giovani, hm i'm behind a router, so i do have double NATing as it were.
[21:07] <giovani> yeah, try bridgin or something
[21:07] <trothigar> giovani, Surely the jaunty install should time out eventually?
[21:07] <giovani> could be unrelated
[21:07] <giovani> it should
[21:07] <giovani> I'd switch into the console
[21:07] <giovani> and check what it's doing while "frozen"
[21:08] <trothigar> is "Host interface" what they now call bridging?
[21:09] <giovani> maybe
[21:09] <giovani> try it out
[21:09] <giovani> I don't use virtualbox
[21:21] <Sam-I-Am> sommer: you around?
[21:31] <sommer> Sam-I-Am: yeppers
[21:31] <Sam-I-Am> sommer: so i hear you're the documentation gyu
[21:31] <Sam-I-Am> guy
[21:32] <sommer> sure, I do what I can :)
[21:33] <Sam-I-Am> well, i tend to set up a lot of openldap-samba-kerberos-dhcp-bind systems (infrastructure stuff) and wondered if i can help with either a) the server guide and/or b) wiki pages for various things
[21:33] <sommer> Sam-I-Am: sure, all help is greatly appreciated
[21:33] <Sam-I-Am> so how would i suggest updats?
[21:34] <sommer> Sam-I-Am: one thing that would be great is if you could review the openldap, kerberos, and ldap and kerberos sections of the server guide
[21:34] <Sam-I-Am> no problem
[21:35] <Sam-I-Am> one thing i noticed is no mention of sasl with openldap, yet the ldaputils stuff tries sasl by default... might be nice to have a basic setup
[21:35] <sommer> Sam-I-Am: if you find any issues grammar, syntax, etc you can report them in LP
[21:36] <Sam-I-Am> so just report things as bugs?
[21:36] <Sam-I-Am> and suggested changes...
[21:36] <sommer> Sam-I-Am: ya, I haven't done much with sasl, but if you have experience with that a new section would be great
[21:36] <Sam-I-Am> sure
[21:37] <sommer> there's a quick guide to get the docbook xml source here: https://wiki.ubuntu.com/ServerTeam/KnowledgeBase#Documentor%20resources
[21:37] <Sam-I-Am> i'm writing ground-up docs here at work, so they might apply well to the public
[21:37] <sommer> Sam-I-Am: cool, any help would be awesome
[21:37] <Sam-I-Am> figure why keep stuff internal :)
[21:38] <Sam-I-Am> especially if i'm going through the pain of writing this so just about anyone can plug-n-chug
[21:39] <mathiaz> Sam-I-Am: I've been working on some script to integrate kerberos+openldap
[21:39] <sommer> heh, ya sometimes writing things up can be painful... but every time I refer back for a command I can't remember it all becomes worth it :-)
[21:39] <mathiaz> Sam-I-Am: https://code.launchpad.net/~mathiaz/openldap-dit/dynamic-backend
[21:40] <sommer> mathiaz: sweetness :)
[21:40] <mathiaz> Sam-I-Am: have you integrated dhcp/dns to use openldap as the backend?
[21:40] <Sam-I-Am> i've integrated dhcp
[21:40] <mathiaz> sommer: it probably won't make it in time for karmic though
[21:40] <Sam-I-Am> however, theres some caviats with that
[21:40] <mathiaz> Sam-I-Am: are you using isc dhcp?
[21:40] <Sam-I-Am> yes
[21:40] <mathiaz> Sam-I-Am: with the ldap patch?
[21:40] <Sam-I-Am> yes
[21:40] <Sam-I-Am> i'm also trying to package dhcp4
[21:40] <Sam-I-Am> with the ldap patch
[21:40] <mathiaz> Sam-I-Am: IIRC this patch is not in debian
[21:41] <mathiaz> Sam-I-Am: has the ldap patch been submitted to upstream?
[21:41] <Sam-I-Am> yeah, for a while now
[21:41] <Sam-I-Am> they dont even put it in contrib
[21:42] <mathiaz> Sam-I-Am: IIRC there was some push back from the debian maintainer to not include the ldap patch
[21:42] <Sam-I-Am> i see a dhcp-server-ldap in debian
[21:42] <mathiaz> Sam-I-Am: where?
[21:42] <Sam-I-Am> http://packages.debian.org/lenny/dhcp3-server-ldap
[21:43] <mathiaz> Sam-I-Am: oh right - I remember now
[21:43] <Sam-I-Am> the dhcp ldap schema is fairly well, formed, but in testing i've only found a few things that actually work in ldap... particularly host management, which is really what we want... although config would be a nice touch
[21:43] <Sam-I-Am> extra comma in there..
[21:44] <mathiaz> Sam-I-Am: right - my plan was to integrate dns+dhcp via a shared backend (openldap)
[21:44] <mathiaz> Sam-I-Am: and throw kerberos in the mix too
[21:44] <Sam-I-Am> yeah
[21:44] <Sam-I-Am> i havent deployed dns in ldap in production yet, but i have it in testing
[21:44] <mathiaz> Sam-I-Am: cool.
[21:44] <Sam-I-Am> yeah
[21:44] <Sam-I-Am> everything in ldap!
[21:45] <mathiaz> Sam-I-Am: If you could document this in a wiki page (to start)
[21:45] <Sam-I-Am> btw, i'm one of the dhcp maintainers :)
[21:45] <mathiaz> Sam-I-Am: then we can start on packaging all of this to create an OOTB experience
[21:45] <Sam-I-Am> i have several dozen wiki pages here at work... once i'm done documenting, i plan to post them publically somewhere once i clean up the company-specific hoopla
[21:46] <Sam-I-Am> btw, speaking of OOTB, should probably put a post-install note for slapd on how to access cn=config with .17 :)
[21:46] <Sam-I-Am> since it doesnt ask for a password anymore, it might confuse people who dont know about sasl external
[21:46] <mathiaz> Sam-I-Am: right - that's a good idea and can be documented after FeatureFreeze
[21:47] <mathiaz> Sam-I-Am: all the script I posted above use the new EXTERNAL method
[21:47] <mathiaz> Sam-I-Am: the missing part is that kerberos cannot used sasl EXTERNAL
[21:47] <Sam-I-Am> because you're using it for config?
[21:47] <mathiaz> Sam-I-Am: I should file a feature request with the upstream folks to support that
[21:48] <mathiaz> Sam-I-Am: when you run krb5_ldap_utils to create the realm you need to give a DN and a password
[21:48] <mathiaz> Sam-I-Am: is there another way to create a realms beside using krb5_ldap_utils?
[21:49] <Sam-I-Am> well, to be honest i'm using heimdal here... i just call kadmin to generate a realm after editing krb5.conf
[21:49] <mathiaz> Sam-I-Am: another point I'd like to investigate is to see if the two users that the krb5kdc use could use SASL external to connect to the slapd daemon over ldapi
[21:50] <Sam-I-Am> yeah, ldapi is a bit limited right now
[21:50] <mathiaz> Sam-I-Am: right - does heimdal kadmin support SASL external?
[21:50] <Sam-I-Am> yes
[21:51] <Sam-I-Am> i'm almost sure heimdal only supports ldapi
[21:53] <Sam-I-Am> it of course comes in as root...
[21:53] <Sam-I-Am> you can have other ldapi users and the proper authzregexp for them... but all the kerberos stuff runs as root
[21:53] <mathiaz> Sam-I-Am: right - support SASL EXTERNAL would mean running the different kdc under different accounts
[21:53] <mathiaz> Sam-I-Am: and then map this via AuthzMap to the correct dn
[21:53] <Sam-I-Am> yeh
[21:54] <Sam-I-Am> i'm still torn between mit and heimdal...
[21:54] <mathiaz> Sam-I-Am: IIUC this is not supported by heimdal?
[21:54] <Sam-I-Am> heimdal right now has the added benefit of updating samba hashes in ldap as it finds them
[21:54] <Sam-I-Am> heimdal supports ldapi and sasl external
[21:54] <mathiaz> Sam-I-Am: I'll post a feature request to the MIT folks as I'm in contact with the dev team
[21:55] <Sam-I-Am> does mit not support sasl external?
[21:55] <Sam-I-Am> its on my list of things to configure in the testbed...
[21:56] <Sam-I-Am> i design and plunk infrastructure systems down all the time and think in some cases heimdal or mit would better suit a particular customer
[21:56] <julius> hi
[21:56] <Sam-I-Am> howdy
[21:56] <julius> how can i close a bug reported by me on launchpad?
[21:56] <Sam-I-Am> just say its been fixed and the maintainer will close it as needed...
[21:56] <Sam-I-Am> post some detail about how it was fixed..
[21:58] <Sam-I-Am> i wonder if mit supports tls or just ssl...
[21:58] <julius> so i as the owner cant close it?
[21:58] <Sam-I-Am> i... dont think so.
[21:58] <julius> thx
[21:59] <Sam-I-Am> mathiaz: whats your priority on a dhcp4 package w/ ldap patch?
[21:59] <Sam-I-Am> i have dhcp4 built against sid/karmic with the patch... just not packaged yet since they more or less replaced the entire build internals
[22:00] <Sam-I-Am> probably something for 10.04
[22:00] <mathiaz> Sam-I-Am: not a target for 9.10
[22:00] <mathiaz> Sam-I-Am: having ldap+krb5 integrated was a low priority target for 9.10
[22:01] <mathiaz> Sam-I-Am: but things haven't moved as quickly as expected
[22:01] <Sam-I-Am> that seems to be a common case
[22:01] <mathiaz> Sam-I-Am: (mainly blocked on administration tools)
[22:01] <Sam-I-Am> if i could spend every day working on open source stuff i would... but it doesnt work that way
[22:01] <mathiaz> Sam-I-Am: well - you're *already* spending more time on open source stuff than most of the people :)
[22:02] <Sam-I-Am> the best i can do is make my stuff as generic as possible, push docs back out, and get things patched
[22:02] <mathiaz> Sam-I-Am: that's one of the best approach
[22:03] <Sam-I-Am> and working closely with the upstream developers for what i use
[22:03] <Sam-I-Am> it literally took 5 minutes to get the nssov patch from howard
[22:04] <Sam-I-Am> now if only gnutls worked that quickly... :/
[22:04] <Sam-I-Am> i found a bug in that which essentially breaks typical certs with subjectaltname...
[22:06] <Doonz> heya all
[22:06] <Sam-I-Am> howdy
[22:07] <Doonz> need some help with dns on my server
[22:07] <Doonz> i set the 2 nics to static ip's
[22:07] <Doonz> but now i cant resolve host names ie: cant ping google.com
[22:07] <Sam-I-Am> is there a default route?
[22:08] <Doonz> i changed my /etc/resolv.conf to show my isp dns servers
[22:08] <Doonz> Sam-I-Am: sorry im new with this
[22:08] <Sam-I-Am> can you ping those dns server IPs?
[22:08] <Doonz> hmm never thought of that
[22:08] <Doonz> one sec
[22:09] <luxos> buenas
[22:09] <luxos> como van
[22:09] <luxos> pregunta
[22:09] <luxos> quiero instlar una impresora en el servidor linux y quiero compartirla en la red
[22:09] <mathiaz> Sam-I-Am: well gnutls may be a bit buggy but it's the only option we have from a licensing point of view
[22:09] <luxos> como hago eso
[22:09] <Doonz> Sam-I-Am: i cant ping the nameservers either
[22:10] <mathiaz> !es | luxos
[22:10] <Doonz> i can only ping my router and other pcs on this network
[22:11] <Sam-I-Am> can you get an ip of google.com from another machine and try to ping it from the machine in question?
[22:11] <KillMeNow> Doonz, did you check to make sure you have a good gateway set?
[22:12] <Doonz> the gateway is set to 192.168.1.1 wich it is
[22:12] <Doonz> but i cant ping outside of my network
[22:12] <KillMeNow> can you ping outside from your gateway?
[22:13] <Sam-I-Am> are other machines using that gateway working?
[22:13] <Doonz> i cant ping everything from this laptop
[22:13] <Doonz> can*
[22:13] <KillMeNow> what is your gateway?  is it a linux box or a little router?
[22:13] <Doonz> its a router
[22:13] <Doonz> but i have my resolv.conf file set up with the ips of my isp dns servers
[22:13] <Sam-I-Am> same as the laptop which works?
[22:14] <KillMeNow> run a route command
[22:14] <KillMeNow> from the box that is having the issue
[22:14] <KillMeNow> can you ping those?
[22:14] <Sam-I-Am> you also mentioned something about having 2 nics or at least two IPs on this machine in question
[22:15] <Sam-I-Am> depending on how those are configured you could have a problem
[22:15] <Doonz> http://pastebin.com/m59313df1
[22:15] <Doonz> thats the output from route
[22:15] <Doonz> both nics have different ip's
[22:15] <KillMeNow> anyone else seeing 2 gateways?
[22:15] <KillMeNow> do a ifconfig please
[22:16] <Doonz> ok
[22:16] <Sam-I-Am> yeah i'm seeing two
[22:16] <Sam-I-Am> you shouldnt have two default gateways...
[22:16] <Doonz> http://pastebin.com/mebc428a
[22:17]  * Doonz is a noob
[22:17] <KillMeNow> remove the gateway statement from one of your interfaces
[22:17] <KillMeNow> or set one statically
[22:17] <Doonz> ...
[22:17] <KillMeNow> but i guess you see the problem now
[22:17] <Doonz> NO I DONT
[22:17] <Doonz> bah sorry
[22:18] <KillMeNow> ok give me the config for each of your eth interfaces...  are they DHCP or static?
[22:18] <Doonz> static
[22:19] <Doonz> http://pastebin.com/me4fae7
[22:20] <KillMeNow> pick one of your interfaces to be the default and then edit your config and remove one of the gateway addresses
[22:20] <KillMeNow> the whole gateway statement
[22:20] <Doonz> so un eth1 just romve the gateway part?
[22:20] <KillMeNow> yea i updated your pastebin
[22:20] <Doonz> ok
[22:20] <KillMeNow> using eth0 as the main interface
[22:21] <KillMeNow> what are you trying to do?  you trying to bind them together for better bandwidth?  like get 2 gbps by joining 2 nics?
[22:21] <Doonz> no
[22:21] <KillMeNow> ok
[22:21] <Doonz> just 2 seperate ips
[22:22] <KillMeNow> once you ahve it edited, restart your network
[22:22] <Doonz> want to keep media traffic on one nic and the rest on the other nic
[22:22] <KillMeNow> by issuing:  /etc/init.d/network restart
[22:24] <Doonz> hmm
[22:24] <Doonz> it doesnt like that command
[22:24] <hggdh> or 'service network restart'
[22:24] <Doonz> @server:/etc/init.d# service network restart
[22:24] <Doonz> $network: unrecognized service
[22:25] <KillMeNow> sorry
[22:25] <KillMeNow> use networking
[22:25] <KillMeNow> s/network/networking/
[22:25] <Doonz> oh ok
[22:25] <hggdh> yeah. Same mistake I made ;-)
[22:26] <Doonz> hehe
[22:26] <KillMeNow> nah, i work in RHEL / Ubuntu
[22:26] <KillMeNow> get the two mixed up sometimes
[22:26] <Doonz> YAY
[22:26] <Doonz> thanx all good now
[22:26] <KillMeNow> congrats
[22:27] <Doonz> brb
[22:28] <Doonz> yaya
[22:28] <Doonz> woohoo
[22:28] <Doonz> all good
[22:29] <KillMeNow> i'm glad
[22:29] <Sam-I-Am> mathiaz: so whats the idea for security regarding all this stuff talking to ldap... like dhcp, dns, etc... most of them don't support ldapi, and even if they did, they'd come in as the root user.  i dont think all of these services would need write access to the whole ldap tree.
[22:43] <VirtualDisaster> and you dont want to make dhcp/dns dependant on ldap
[22:43] <VirtualDisaster> use ldap for auth and use radius for items like dhcp clients
[22:45] <mathiaz> Sam-I-Am: right. One of the idea in Ubuntu is to try to run most of daemons as non-root
[22:45] <mathiaz> Sam-I-Am: so given that goal, having the services running under their own account would solve the problem
[22:45] <Sam-I-Am> yes
[22:46] <mathiaz> Sam-I-Am: IIRC the dhcp server doesn't run as root
[22:46] <Sam-I-Am> nope
[22:46] <Sam-I-Am> nor does bind to some extent...
[22:46] <Sam-I-Am> they do at one point or another
[22:47] <Sam-I-Am> reading through your todo...
[22:47] <mathiaz> Sam-I-Am: another idea wrt to security is to try to use SASL External + kerberos to authenticate and encrypt communication
[22:47] <mathiaz> Sam-I-Am: from my openldap-dit branch?
[22:47] <Sam-I-Am> yeah
[22:47] <mathiaz> Sam-I-Am: the first lines may no be relevant anymore - they're related to another DIT
[22:48] <mathiaz> Sam-I-Am: my branch is actually base on the trunk from the openldap-dit LP project
[22:48] <mathiaz> Sam-I-Am: which was started by andreas
[22:48] <mathiaz> Sam-I-Am: it's based on the Mandriva Directory Service
[22:48] <Sam-I-Am> ah
[22:48] <Sam-I-Am> noticed the heimdal branch question
[22:48] <mathiaz> Sam-I-Am: andreas used to work for Mandriva
[22:48] <mathiaz> Sam-I-Am: yes - heimdal was the choice then
[22:49] <Sam-I-Am> i store heimdal machine/service credentials in ou=computers with the samba machines
[22:49] <Sam-I-Am> and person creds in ou=people
[22:49] <mathiaz> Sam-I-Am: I've removed a lot of the original work to reduce the scope
[22:49] <mathiaz> Sam-I-Am: right - I moved everything under accounts
[22:49] <Sam-I-Am> yea, guess thats irrelevant now if we're using mit heh
[22:50] <mathiaz> Sam-I-Am: as from the kerberos perspective they're all principals
[22:50] <Sam-I-Am> yeah
[22:50] <Sam-I-Am> mit clients talk to heimdal kdcs just fine
[22:50] <Sam-I-Am> its just kadmin that doesnt work
[22:50] <mathiaz> Sam-I-Am: does heimdal make a difference between host principals, service principals and user principals?
[22:50] <mathiaz> Sam-I-Am: ie can it store them in different sub-trees?
[22:51] <Sam-I-Am> it only allows you to configure one place to store principals in the config file, but i usually use a script to add mine so they go into the appropriate locations... and the heimdal service account can read the entire ldap tree
[22:51] <mathiaz> Sam-I-Am: I haven't played enough with the DIT where all principals are stored in the same OU
[22:51] <Sam-I-Am> it'll work all in the same ou, it just looks messy
[22:52] <mathiaz> Sam-I-Am: right - that comes back to the issue of administration tools
[22:52] <mathiaz> Sam-I-Am: right - the DIT isn't supposed to be read by end users
[22:52] <mathiaz> Sam-I-Am: one of the design principles I follow is to have a shallow tree with as little hierarchy as possible
[22:53] <Sam-I-Am> with heimdal, if your kerberos attributes are not in the same DN as your other user stuff, it wont also update your samba hash upon kerberos key change
[22:53] <Sam-I-Am> mine are pretty limited too... mostly nis-like with some extras for heimdal, dhcp, dns, and samba
[22:54] <mathiaz> Sam-I-Am: right - that would be covered by the krb5smb openldap overlay
[22:54] <Sam-I-Am> yes, which should probably get built along with nssov
[22:54] <mathiaz> Sam-I-Am: which would be responsible for maintaining all passwords in sync
[22:54] <mathiaz> Sam-I-Am: it doesn't support MIT kerberos though
[22:54] <Sam-I-Am> ahh yes, another problem...
[22:55] <Sam-I-Am> as of a few months ago it didnt like compiling against gnutls either
[22:55] <mathiaz> Sam-I-Am: and I know that howard is working on updating a couple of IETF drafts wrt to password policies and kerberos schema
[22:55] <Sam-I-Am> yeah, thats cool
[22:55] <mathiaz> Sam-I-Am: so it may worth waiting a bit to see what comes out of this
[22:56] <Sam-I-Am> someone in openldap-devel sent me his diffs for building smbk5pwd within the ubuntu package, so that might be useful either way...
[22:56] <mathiaz> Sam-I-Am: oh cool.
[22:56] <Sam-I-Am> i think that was part of the reason i chose heimdal
[22:56] <Sam-I-Am> that and its potential integration with samba4... although samba4 appears to run its own kdc
[22:56] <mathiaz> Sam-I-Am: yes - it makes sense.
[22:56] <mathiaz> Sam-I-Am: samba4 is being ported to MIT kerberos
[22:57] <Sam-I-Am> ha
[22:57] <mathiaz> Sam-I-Am: they've started with heimdal but work is done to support MIT as well.
[22:57] <Sam-I-Am> would be nice if it could just compile against one or the other... just like openldap and gnutls vs. openssl vs. moznss
[22:58] <Sam-I-Am> although that sort of thing makes the code a bit more complicated
[22:58] <Sam-I-Am> especially if you have to work around implementation specific bugs... which are fairly common with openldap-gnutls
[23:00] <Sam-I-Am> also, while i'm thinking about it, how does one get approved for the server team?
[23:00] <mathiaz> Sam-I-Am: hm - when I'll process pending the requests :D
[23:00] <Sam-I-Am> ah, ok
[23:01] <mathiaz> Sam-I-Am: I haven't done that for a while
[23:01] <Sam-I-Am> i should be one of them
[23:04] <Sam-I-Am> i sure wish there were more hours in the day...
[23:04] <Sam-I-Am> so many interesting things to do, so little time
[23:04] <Sam-I-Am> also saw your note about autofs... several schema for that :/
[23:04] <Sam-I-Am> depending on what kind of compatibility you want
[23:07] <Sam-I-Am> time to head home... bbl