[00:06] <moz_> guys do you know any usb wireless devices i cant go wrong with, with ubuntu-server
[00:13] <Kamping_Kaiser> Could someone give me the name of the command line version of ubiquity? I know i've seen it meantaioned on the ubuntu-server list, but I cant find anything about it on the net
[00:14] <twb> Kamping_Kaiser: there isn't one.
[00:14] <twb> Kamping_Kaiser: however you may be thinking of debian-installer (d-i), which is completely different software that is both more powerful and more flexible.
[00:14] <Kamping_Kaiser> twb, i'm pretty sure there is. it wasnt an ubuntu project, but was meantioned on the list.
[00:14] <Kamping_Kaiser> and no, its not d-i
[00:15] <twb> OK, perhaps I am not possessing all the facts.
[00:16] <Kamping_Kaiser> I'll lurk a while, and if no one knows, I'll have to grave dig my maildir :l
[04:07] <tommy_nmw> hi
[04:07] <tommy_nmw> let me ask about ubuntu
[04:07] <tommy_nmw> hello
[04:08] <tommy_nmw> who knows how to set up active directory like Ubuntu domain controller ?
[04:25] <twb> tommy_nmw: that would be OpenLDAP and Samba, I guess
[04:25] <twb> I don't know much about it
[04:29] <tommy_nmw> so who knows well?
[04:29] <treykee> good evening all
[04:41] <tommy_nmw> yes
[04:58] <llragsll> I want sshd to check the ssh_known_hosts file for all connections and no the .ssh/known_hosts file that is created for each user...
[05:32] <rags> can I have two gateways configured...with 2 different metrics?? I want it to switch whenever the other one drops..Is this possible?
[05:37] <Kamping_Kaiser> yes, and theres tools packaged for that purpose
[05:37] <Kamping_Kaiser> cant give personal experiance though
[05:51] <Noah0504> Anyone around?
[07:22] <_ruben> hmm .. im starting to experience more and more "issues" with cronjobs lately .. some work as expected, others down run at all, or dont do anything .. guess it's time to do some more investigating
[07:24] <rst-uanic> _ruben: have you tried to start those jobs manually?
[07:24] <_ruben> manually they work fine
[07:25] <_ruben> could be an $ENV problem i guess
[07:25] <rst-uanic> maybe, try exporting needed variables in the job script
[07:26] <rst-uanic> also, did you run those jobs as the user specified in the crontab or as yourself?
[07:30] <_ruben> both as root, which does have some effects on the $ENV (sudo versus cronjob)
[07:34] <rst-uanic> _ruben: and you use full paths in the cronjobs?
[07:39] <Alex_21> Hi, All,
[07:39] <Alex_21> In Ubuntu is .HTaccess access enabled by default? The release I am using is 8.04 Hardy. Please
[07:39] <_ruben> rst-uanic: yes
[07:40] <Alex_21> Odd, because I can't install Drupal from source for some reason
[07:40] <rst-uanic> _ruben: and what $ENV variables do you think you use?
[07:40] <_ruben> Alex_21: the allowoverride setting configures what stuff you can put in .htaccess sites
[07:40] <Alex_21> Does anyone have experiance with instaling Drupal from source?
[07:40] <_ruben> rst-uanic: none specifically .. i havent researched it very much yet .. was about to do so
[07:53] <Skaag> Can I safely remove console kit somehow? It wants to remove dbus as well
[07:53] <Skaag> Why do I need that stuff on a server that will never do gnome sessions?
[07:55] <_ruben> dbus depends on consolekit, for whatever reason
[08:20] <_ruben> it was a $PATH issue .. my cronjob calls /sbin/ipvsadm .. which turns out to be a shell script which then calls ipvsadm-1.24 and apparently /sbin/ isnt in the $PATH for (root) cronjobs
[08:42] <Skaag> _ruben: I know dbus depends on it, do I need dbus?
[08:43] <_ruben> Skaag: how should I know? :)
[08:44] <Skaag> is it required for proper functioning of an ubuntu server?
[08:44] <Skaag> ;-)
[08:44] <_ruben> Skaag: I dont know what apps you got installed which might need it
[08:44] <Skaag> Apache, MySQL...
[08:44] <_ruben> my servers dont have dbus
[08:44] <Skaag> ok great
[08:44] <Skaag> I assume if an app will need dbus it will say so
[08:45] <Skaag> I was just afraid it's required for the basic operation of the OS itself
[08:45] <_ruben> it should yeah
[08:46] <Skaag> ok removed that stuff...
[08:46] <Skaag> it's just that sometimes console kit takes 100% CPU for no reason
[08:46] <Skaag> and I noticed it's a known bug
[08:46] <Skaag> thanks :)
[08:47] <_ruben> nice
[08:49] <rst-uanic> _ruben: is it working now? :)
[08:49] <_ruben> rst-uanic: this specific one, yes .. i added a "PATH=/sbin:$PATH" to my script :)
[08:50] <rst-uanic> that's good
[08:52] <_ruben> i wonder if i should file a bug wrt the ipvsadm behaviour
[08:53] <rst-uanic> _ruben: my $PATH has /sbin
[08:54] <_ruben> rst-uanic: but also in a cronjob?
[08:54] <rst-uanic> don't know
[08:56] <rst-uanic> I think cron scripts declare their own parameters or just use full paths :)
[08:59] <_ruben> rst-uanic: well .. i did use a full path to ipvsadm, but that doesnt work, because /sbin/ipvsadm is a wrapper script, which doesnt use full pathts
[09:02] <rst-uanic> _ruben: maybe :)
[09:05] <_ruben> filed a bug for completeness sake .. at the very least it is now "documented" behaviour :)
[09:06] <rst-uanic> :)
[09:07] <infinity> rst-uanic: Crontabs can have a PATH declared at the top.
[09:09] <rst-uanic> infinity: yes they can
[09:10] <infinity> 02:56 < rst-uanic> I think cron scripts declare their own parameters or just use full paths :)
[09:11] <infinity> My point is that the crontab itself (not what you run from it) can have PATH and SHELL and other such things declared before the periodic entries.
[09:11] <infinity> man 5 crontab
[09:12] <rst-uanic> infinity: ok, and my point is that it is better to use full paths and have no problems
[09:12] <infinity> Generally not...
[09:13] <rst-uanic> infinity: why not
[09:13] <rst-uanic> ?
[09:13] <infinity> Scripts that use full paths (and worse, compiled code with full paths compiled in) aren't very resilient to the whole reason the UNIX filesystem hierarchy exists.
[09:14] <infinity>  /usr/local/bin:/usr/bin:/bin <-- That's arranged like that for a reason.  So you can override what's in /usr/bin with /usr/local/bin ... And if you don't want that behaviour, you remove local from your path, not recompile or hardcode.
[09:17] <rst-uanic> infinity: And if I want to write a small scripts for my system? I should think of how this will work on other systems too?
[09:18] <infinity> rst-uanic: Why would you bother hardcoding paths on your own system?  You know what you have installed, surely?
[09:19] <infinity> rst-uanic: But even small projects, I tend to write to the same standards as something I'd write for public release, based on the realisation that a good 75% of my "small projects" end up deployed to dozens of machines, at the least, and publically released to thousands/millions, at the worst.
[09:20] <rst-uanic> infinity: I am not saying that declaring $PATH is wrong. I saying that if I want to write some quick script using full path is quite ok.
[09:20] <infinity> rst-uanic: Well, sure, whatever you want to do is "okay for you", by definition.  It's your system.  I'm just arguing it's not actually "best practice" to hardcode full paths to binaries.
[09:21] <infinity> What if we move something from /bin to /usr/bin or from /usr/sbin to /usr/bin between releases, and your script breaks because of it?
[09:21] <infinity> No hardcoded binary locations, no problem.
[09:21] <rst-uanic> infinity: for sure :)
[09:24] <_ruben> good point
[09:25] <rst-uanic> infinity: thanks for your explanations :)
[12:25] <sparky_> I'm trying to mount my back-up disk to upload on to my server but it can't mount it using "sudo mount -t hfsplus /dev/sdf1" it kicks back saying "wrong fs type, bad option, bad superblock on dev/sdf1.  Any help?
[12:30] <soren> sparky_: Anything in dmesg?
[12:31] <sparky_> [  511.957477] hfs: unable to find HFS+ superblock
[12:31] <soren> Well, there you go.
[12:32] <sparky_> i don't know anything about superblock....
[12:33] <soren> The superblock is where the filesystem puts information about itself.
[12:33] <soren> It's created by mkfs.
[12:33] <sparky_> is there a way to fix without lossing my data
[12:34] <sparky_> oh so just run the mkfs and it will create the superblock?
[12:34] <soren> Nono!
[12:34] <simplexio> i dont think so
[12:34] <soren> no!
[12:34] <soren> sparky_: No!
[12:34] <sparky_> ok
[12:34] <soren> mkfs is what you use to *create* the filesystem to begin with.
[12:35] <simplexio> sparky_: are you 100% sure that your device name is right ?
[12:35] <soren> So if you run it on a block device that already has a filesystem, you'll delete the current contents.
[12:35] <sparky_> yes, it is the only one that is 1Tb in size
[12:35] <soren> sparky_: Try "sudo vol_id /dev/sdf1"
[12:35] <soren> ...and tell me what it says.
[12:37] <sparky_> http://www.pastebin.ca/1369065
[12:38] <soren> Why do you think it's an hfs+ filesystem?
[12:39] <sparky_> I thought that is what i created it as on my mac, but it looks like its fat32 correct?
[12:40] <soren> Yes.
[12:41] <sparky_> so i would us -t fat32
[12:41] <sparky_> use*
[12:41] <soren> Just mount it.
[12:43] <sparky_> it's not showing anything in the dir
[12:54] <soren> sparky_: What did you expect to see? Did it mount correctly?
[12:54] <Giddion> hello
[12:54] <sparky_> i expected to see all my back-up'd files over 600 Gb worth
[12:55] <Giddion> ive got a spare server laying around and im thinking of putting ubuntu server on there, im curious to know what i could use a box which has 32gb HDD space, P4 1.66 and 512mb ram for?  im currently running a mail server, web server, file server and mssql. but this is all windows based so i couldnt transfer it to linux.   any ideas?
[12:59] <orudie> ivoks, hi thanx for helping me last night i got working it looks decent
[13:04] <soren> Giddion: Not sure. Perhaps you could make it a mail, web, file and mysql server? :)
[13:06] <ivoks> soren: have you seend apache directory studio?
[13:06] <ivoks> seen
[13:07] <soren> ivoks: Nope.
[13:07] <ivoks> soren: it's ldap browser
[13:07] <ivoks> soren: it even has scheme editor and ldif editor
[13:08] <ivoks> it's a perfect tool for managing ldap
[13:08] <ivoks> the only bad thing - it's for eclipse
[13:08]  * soren really wishes we didn't have to deal with JAva.
[13:09] <ivoks> soren: it's a workstation-side app
[13:09] <ivoks> soren: on server, we wouldn't have java
[13:09] <soren> ivoks: Right, right, I was just looking at the other Apache Directory stuff.
[13:09] <ivoks> soren: ah, apache directory - ldap implenetation
[13:10] <ivoks> grr... keyboard
[13:10] <ahasenack> hey, they have triggers :)
[13:11] <soren> ahasenack: I'm almost sure that doesn't outweigh the Java thing.
[13:11] <ahasenack> sommer: correct :)
[13:30] <sparky_> soren, thx for the help I got it going
[13:31] <soren> Cool.
[13:38] <AlexC_> g'morning
[13:40] <AlexC_> we've just sent a server off to a data centre, however we've ran into quite a problem that was not noticed before we sent it off. The partition that mounts as / is 250mb in size! Is there a way to somehow move the data that is current on the / partition somewhere else, re-mount that temporarely, increase the size of the partition, then move it back?
[13:42] <ivoks> you could create new partition
[13:42] <ivoks> of the same size
[13:42] <ivoks> dd data to it
[13:43] <ivoks> fix boot loader and fstab to boot from the other partition
[13:43] <ivoks> fix the first partition
[13:43] <ivoks> copy data back
[13:43] <ivoks> fix boot and fstab and reboot
[13:44] <ivoks> of course, be carefull and don't come looking for me if it doesn't work :D
[13:44] <AlexC_> :P
[13:46] <soren> AlexC_: Is it the only partition on the system?
[13:47] <AlexC_> ok, so we've 5 partitions - one each for /home /tmp /usr /var and /. Trouble is we have no more space left to make a new partition, so I'm going to have to kind of do this 'on-the-fly' /home is 213GB, so maybe I could move all partition data into there and only mount that, then delete all others, fix sizes, copy data back over?
[13:47] <soren> AlexC_: Can I see you partition layout, please?
[13:47] <ivoks> well, then copy everything to /home
[13:47] <AlexC_> soren, sure
[13:48] <giovani3> oh LVM
[13:48] <giovani3> why we love the
[13:48] <AlexC_> soren, http://paste2.org/p/169698
[13:48] <giovani3> thee*
[13:48] <soren> AlexC_: Where's sda1 and sda2?
[13:49] <giovani3> and sda4
[13:49] <ivoks> swap?
[13:49] <ivoks> sda4 would be extended
[13:49] <scope006> when i create a new user in ubuntu server auto complete and bash history doesn't seem to work out of the box as opposed to the first created user
[13:49] <giovani3> and sda7
[13:49] <soren> giovani3: There doesn't have to be a sda4.
[13:49] <AlexC_> soren, they do exist, just not mounted. They are some recovery partitions that ideally we could get rid of (I want to), though boss wants them to be kept there
[13:49] <scope006> anything special i should look at in /etc/profile or something like that
[13:49] <giovani3> why was this not LVMed?
[13:50] <soren> AlexC_: That's why I wanted to see your partition layout. The output of fdisk -l or parted print or something.
[13:50] <AlexC_> soren, ok
[13:50] <ivoks> well, having / on 260MB partition in this case isn't that bad
[13:50] <domas> when will people learn to use single big partition for system, and then big big partition for service data :)
[13:50] <domas> I just have 10G / and huge /a ;-)
[13:50] <AlexC_> soren, http://paste2.org/p/169702
[13:50] <ivoks>  /usr is separate, /var is separate...
[13:50] <soren> domas: Hopefully never :)
[13:51] <AlexC_> ivoks, it sure is, as I can't do any updates to the server due to it running out of disk space
[13:51] <ivoks> i guess kernel is creating problems
[13:52] <soren> AlexC_: Shame. I was hoping we could just extend your root partition.
[13:52] <ivoks> having boot bind-mounted to directory in /var would help :)
[13:53] <AlexC_> soren, same here. If I was to delete sda1 and 2, that would be possible?
[13:53] <AlexC_> though they are very small, but it would give us something to play around with
[13:53] <soren> AlexC_: Technically, yes. I wouldn't, though.
[13:53] <ivoks> AlexC_: if i were you, i would delete sda1, sda2, sad3 and sda5
[13:53] <soren> AlexC_: sda2 is several GB.
[13:53] <domas> soren: why? :)
[13:54] <ivoks> AlexC_: and then create / on top of all of them
[13:54] <ivoks> i'm not sure that's possible, since three of them are primary and one extended
[13:54] <soren> domas: Why on Earth would you do something like that?
[13:54] <Deeps> sda1 is marked type dell utility, sda2 is fat32, so i'm guessing server's recovery partition?
[13:55] <Deeps> oh, hmm, was scrolled up
[13:55] <Deeps> ignore me
[13:55] <AlexC_> Deeps, correct
[13:55] <domas> soren: why on earth would you constrain yourself in /var, /tmp, etc ;-)
[13:55] <AlexC_> though I see no need for them (recovery partitions)
[13:55] <domas> soren: i understand ubuntu has specific philosophy (like storing essential database files in /var/run ;-))
[13:55] <ivoks> right...
[13:55] <ivoks> that's very stupid :)
[13:55] <ivoks> we should fix that
[13:56] <soren> domas: Eh?
[13:56] <domas> I can give you 10 things that are idiotic about mysql package on ubuntu
[13:56] <domas> I filed one of them, and nobody even looked at the bug :)
[13:56] <ivoks> domas: pid in /var/lib/mysql? :)
[13:56] <giovani3> is there a reason you can't just dump all the data on a drive, and start over?
[13:56] <domas> ivoks: binlogs in /var/run
[13:56] <ivoks> domas: that's right
[13:56] <domas> ivoks: what is right in that?
[13:57] <AlexC_> giovani3, server is physically few hundred miles away from us now, though with the RAC we have it *may* be possible, however I would not like to do this
[13:57] <domas> anyway, mysql writable ~mysql is as bad =)
[13:57] <ivoks> nothing, i ack on that bug
[13:57] <domas> "CHECK TABLES" at restart is _extremely_ annoying on terabyte tables, you know.. ;-)
[13:57] <domas> well, even on 100GB databases it is quite bad. ;-)
[13:58] <domas> soren: see, people put arbitrary limits where they don't need them, and don't have proper setup for anything else
[13:58] <AlexC_> soren, what are the technical reasons for not deleting sda2 and extending sda3 into that space?
[13:58] <giovani3> AlexC_: your datacenter doesn't offer kvm?
[13:58] <giovani3> you could ask to borrow a usb drive, backup your stuff
[13:58] <ivoks> domas: could you summarize all those pitfalls?
[13:58] <AlexC_> giovani3, I don't know tbh
[13:58] <ivoks> domas: so that we can fix them all
[13:58] <giovani3> then ask them to put some install media in the cd drive, and start over with LVM!
[13:58] <giovani3> AlexC_: well I'd be asking if I were you
[13:58] <domas> ivoks: as long as someone will listen, sure
[13:58] <ivoks> domas: i will listen
[13:59] <domas> ivoks: lately I find out that people who pay me listen to me way more than whose who don't ;-))
[13:59] <domas> ivoks: ok!
[13:59] <soren> AlexC_: The fact that sda2 is located *before* sda3.
[13:59] <soren> You can't extend backwards.
[13:59] <AlexC_> ah
[14:00] <soren> domas: The problem isn't that people contrain themselves. The problem is that it's too tedious to move stuff around.
[14:00] <ivoks> domas: if you file a bug on LP, add me (ivoks) as surbscribed
[14:00] <ivoks> domas: or, if you'd preffere to send an email, ivoks at ubuntu com
[14:00] <domas> ivoks: oki
[14:01] <AlexC_> soren, would formatting sda2, then copying all data from sda3 to 2, then delete 3, extend 2 into 3 work?
[14:01] <domas> ivoks: e.g. https://bugs.launchpad.net/ubuntu/+source/mysql-dfsg-5.0/+bug/293258 - is there for quite some time already
[14:01] <soren> AlexC_: Yes.
[14:02] <AlexC_> soren, is that a yes, it is a good idea, or a technically possible yes - but shouldn't really, yes?
[14:02] <domas> ivoks: it isn't critical, but quite bad practice (used by many.. ;-)
[14:02] <ivoks> i know
[14:02] <soren> AlexC_: Probably a good idea, yes.
[14:02] <ivoks> i have to fix conf on couple of my installs
[14:02] <ivoks> and i forgot to fix all that
[14:03] <ivoks> so, once you have a list, we'll fix all of them in one go
[14:03] <domas> oki!
[14:03] <AlexC_> soren, ok cool, could dd handle the difference in partition sizes if I was to do a direct copy of sda3 to sda2?
[14:03] <domas> puts some pressure on me now!
[14:03] <AlexC_> or, should I format sda2, resize that to the same as sda3 - then copy, then extend?
[14:03] <ivoks> bbl
[14:04] <domas> ivoks: there're also few bugs, mysqld_safe loops at 100% after mysqld restart, etc
[14:04] <soren> AlexC_: Sort of.
[14:04] <soren> AlexC_: It won't break. :)
[14:04] <ivoks> domas: i've seen that one, too :/
[14:04] <AlexC_> but ... ? :P
[14:04] <domas> ivoks: I just really have to sit down and review that, maybe should ask my managers to approve this to do that on my payroll
[14:06] <domas> hehe, I have separate big /tmp only on boxes that have LVM snapshot roundrobin configured
[14:06] <domas> does 50G /tmp sound ok?
[14:07] <domas> ivoks: kudos though for having apparmor profile
[14:07] <domas> ivoks: I'm giving 'mysql security' talk soon at MySQL Conference, and it will be one of hilights in it =)
[14:07] <soren> AlexC_: Sorry, I was on the phone..
[14:07] <AlexC_> soren, no worries
[14:08] <soren> AlexC_: Just dd it over. The filesystem will still be those 200-300 MB, so once you've joined the partitions, you should resize2fs it.
[14:08] <soren> ...as that will extend it to be the size of the underlying block device.
[14:09] <ivoks> domas: well, mysql's apparmor isn't my work :)
[14:10] <AlexC_> soren, ok, so: 1) Format sda2, 2) dd sda3 to sda2, 3) delete sda3, 4) extend sda2 5) resize2fs ?
[14:10] <soren> AlexC_: don't bother formatting sda2.
[14:11] <AlexC_> oh really?
[14:11] <AlexC_> would I not then end up with half of sda3, and half of sda2 on the sda2 partition?
[14:11] <soren> dd'ing sda3 onto it will overwrite that anyway.
[14:11] <soren> You would anyway.
[14:11] <ivoks> bbl
[14:11] <soren> mkfs doesn't go through every bit of the block device zeroing it out.
[14:11] <soren> It just puts a superblock there.
[14:11] <AlexC_> ahh ok
[14:11] <domas> ivoks: hehe, now that hardy will fix my apparmor bug soonish, I'll be able to deploy it widely \o/
[14:12] <AlexC_> soren, ok, I'll start on it =) This could be fun, since if I can't boot into the server I'm fecked =3
[14:12] <soren> AlexC_: Be careful. Make sure grub know where to look for stuff.
[14:12] <soren> Or lilo or whatever you're using.
[14:13] <AlexC_> soren, yes, indeed. Just to confirm: 'dd if=/dev/sda3 of=/dev/sda2' is correct?
[14:13] <soren> That looks about right, yes.
[14:13] <AlexC_> ok, cool - thanks for all this btw
[14:14] <AlexC_> and it is ok to do while sda3 is mounted?
[14:16] <soren> I would probably remount it readonly.
[14:17] <soren> ...make sure you have a root terminal open ready to remount it rw.
[14:21] <AlexC_> soren, hum, and the server would still boot up ready for me to SSH in if I was to edit fstab, change to readonly, and restart server?
[14:22] <soren> Oh, you can remount while running.
[14:22] <soren> sudo mount -o remount,ro /
[14:22] <AlexC_> mm, interesting =)
[14:24] <AlexC_> soren, I'm so hesitant to do this, as it's a long trip to London if I mess up. I am currently SSHed in, so I remount, dd, then do same command once finished (though changing 'ro' to 'rw') ?
[14:25] <AlexC_> if you've not got time just say and I'll poke someone else
[14:26] <soren> AlexC_: Again: Make sure you've got a root prompt somewhere else, ready to remount rw.
[14:27] <soren> ..in case something goes bad.
[14:27] <AlexC_> soren, I'm not quite sure I understand you, I can only login via SSH - would that do?
[14:28] <Deeps> have another ssh session open at a root prompt
[14:28] <soren> Open up another terminal, ssh into the server, "sudo su -".
[14:28] <AlexC_> ok
[14:28] <soren> Which boot loader are you using?
[14:29] <AlexC_> grub
[14:30] <soren> Alright. Tell grub that your root partition is /dev/sda2
[14:30] <soren> ...since sda2 and sda3 will have identical UUID's, so I'm not sure which one will be chosen.
[14:30] <AlexC_> ok, I shall do that now. One thing I have just thought, we'll need to make sda2 bootable, or will dd copy that data across?
[14:31] <soren> dd won't copy it across, no.
[14:31] <AlexC_> ok
[14:31] <soren> It's a flag in the partition table.
[14:31] <soren> Not in the filesystem superblock.
[14:32] <AlexC_> ok that makes sense, and is something we'll need to do once dd has completed then
[14:34] <AlexC_> ok, I did the remount - though 'mount' says it is still rw
[14:34] <soren> mount looks at /etc/mtab
[14:34] <AlexC_> ahh ok
[14:34] <AlexC_> time to dd
[14:37] <AlexC_> right, dd finished
[14:38] <AlexC_> mount / back as rw, mount sda2 just to see if data has been copied correctly is next step?
[14:41] <AlexC_> ok that appears to have worked, though I am a bit concerned fdisk still says it is a FAT32 partition - is that because it just hasn't updated, or has everything been copied to fat32 (and more importantly, will that be 'ok')?
[14:41] <soren> fdisk doesn't know about the contents of partitins.
[14:42] <soren> It just sees that the partition's type is that. You can just change it to "Linux".
[14:42] <soren> (type 83)
[14:43] <AlexC_> ah ok, so that is just a ... 'label'/name for it?
[14:43] <soren> Sort o, yes.
[14:46] <AlexC_> soren, is it worth changing it to 83? Also, how do I make this partition bootable? Btw I shall send you a donation once I have done all this, as your help is extremely appreciated
[14:46] <soren> AlexC_: I would probably change it to avoid confusion.
[14:47] <soren> I don't know if the bootable flag really makes any difference, but it's something you do from fdisk as well.
[14:47] <AlexC_> I thought bootable was quite essential, other wise grub couldn't boot it?
[14:48] <soren> I doubt grub cares, to be honest, but seeing as there's very little margin for error, you should probably set it :)
[14:52] <AlexC_> ok, awesome - that's done
[14:53] <AlexC_> technically I should now be able to reboot and everything *should* work, on the new partition? Then after that I simply delete sda3 and increase sda2 size?
[15:00] <soren> That's the idea, yes.
[15:30] <rgreening> anyone here familiar with openerp? trying to run the server in jaunty and it fails. comp
[15:30] <rgreening> complains about xpath
[15:31] <rgreening> ERROR: Import xpath module
[15:31] <rgreening> ERROR: Try to install the old python-xml package
[15:31] <rgreening> ScottK: ^ any ideas?
[15:31] <ivoks> install python-xml package?
[15:31] <giovani3> heh
[15:31] <giovani3> the instructions are a clue
[15:31] <rgreening> ivoks: its installed, obviously.
[15:31] <giovani3> jaunty is also ... not stable
[15:31] <rgreening> giovani3: Im not a n00b :P
[15:32] <giovani3> rgreening: file a bug report?
[15:32] <ScottK> rgreening: By design python-xml is not in the default path.
[15:32] <ScottK> giovani3: I'd rather he fix it.
[15:32] <rgreening> :)
[15:32] <rgreening> ScottK: and I am capable, with a hint...
[15:32] <rgreening> as you know
[15:33] <ScottK> rgreening: I don't recall what you have to add to the path to see the old python-xml stuff, but that's what's needed.  The python-xml package should yield sufficient clues.
[15:33] <rgreening> So ScottK, does that mean, I need to mod the openerp server script to update the PYTHONPATH?
[15:33] <ivoks> /usr/share/pyshared/oldxml/
[15:33] <ScottK> rgreening: I think so.  As ivoks says.
[15:33] <ivoks> just ignore me :)
[15:33] <rgreening> cool.
[15:34] <rgreening> ScottK: I'll take a look
[15:34] <rgreening> thanks ivoks
[15:34] <ivoks> dpkg -L python-xml might help
[15:34] <ScottK> Great.  IIRC it's in Universe so you can upload it.
[15:34] <ivoks> i'm not sure my answer is correct
[15:34] <rgreening> cool
[15:34] <rgreening> np, I'll look around see if I can work it out.
[15:36] <AlexC_> soren, still alive? It's booted up into sda2 =) So far so good, what did you say I needed to do to make sda2 as big as I needed?
[15:36] <AlexC_> just use fdisk to change it? (well, mount sda2 as ro first I would assume?)
[15:37] <soren> AlexC_: No need.
[15:37] <soren> AlexC_: Just remove sda2 and sda3, recreate sda2 with the start sector of the old sda2 and end sector of the old sector3.
[15:37] <soren> Err... s/sector3/sda/3/g
[15:38] <AlexC_> soren, I can't remove both, if I do that I loose all data on them. Currently sda2 is the copy of sda3
[15:38] <soren> Yes, you can.
[15:38] <rgreening> ScottK: prob this path... /usr/share/pyshared/oldxml
[15:38] <soren> Removing the partition does not touch the data on it.
[15:38] <soren> AlexC_: ^
[15:39] <ScottK> rgreening: Since that's the one ivoks suggested, I'd tend to think yes.
[15:39] <rgreening> hmm. tried exporting it. no go..
[15:39] <AlexC_> soren, ok, that makes sense. And I do this all with fdisk, right?
[15:40] <soren> AlexC_: Yes.
[15:43] <domas> ivoks: heh, I see intrepid has a bit better scripts than hardy :)
[15:43] <ivoks> domas: still, we should fix it in hardy too
[15:43] <soren> domas: For what?
[15:43] <ivoks> mysql
[15:48] <AlexC_> soren, ok, that is done. I know have '/dev/sda2   *          12         307     2377620   83  Linux'. However, 'df -h' is still showing /dev/sda2 as 260mb, what is the next step after this?
[15:50] <ScottK> rgreening: sys.path.append('/usr/share/pyshared/oldxml/_xmlplus') will get you started, but then import xpath dies in progress leading to more clues.
[15:51] <rgreening> ScottK: yeah. They import boolean (which is part of python 2.6.. and other mess...)
[15:51] <ScottK> So I think that leads to more work worth doing.
[15:52] <rgreening> hehe
[15:52] <ScottK> rgreening: It's actually xpath that imports boolean.
[15:52] <rgreening> I wanted to demo this software to someone today... no go for Jaunty...
[15:52] <ScottK> Intrepid will be similar probably as python-xml is out of the path there too.
[15:54] <domas> ivoks: damn you people, quite a few problems have been fixed in intrepid :)
[15:54] <domas> ivoks: I was looking at all init scripts and saw "oh, they changed this, they changed that" :)
[15:55] <domas> ivoks: I'm tempted to set up PPA for hardy with intrepid packages
[15:55] <rgreening> ScottK: right from the python intrepreter, importing xpath is broken, so it leads me to believe there's either some voodoo needed or the oldxml stuff is just broken..
[15:56] <ScottK> I would tend to suspect an incomplete python-xml update for Python 2.6.
[15:57] <AlexC_> am I right in thinking,  resize2fs /dev/sda2  will resize the filesystem to  take up all the partition
[15:58] <rgreening> oh my, whats a rats nest i've uncovered....
[15:58] <rgreening> ScottK: Bug 337759
[15:59]  * ScottK was looking for In Progress -> rgreening....
[15:59] <rgreening> ScottK: its an existing bug report
[16:00] <ScottK> Yes, and?
[16:00] <AlexC_> g'bai
[16:00] <ScottK> rgreening: You're allowed to fix ones you didn't report.
[16:00] <rgreening> I know, I was just indicating I found an open one to work against :)
[16:02] <rgreening> ScottK: seems to me, modifying the oldxml to not load boolean may be whats necessary.
[16:03] <ScottK> rgreening: From what little I know about this, that seems likely.
[16:03] <ScottK> It's dead upstream, so don't feel bad about patching it.
[16:03] <rgreening> though what breaks outside of that, is unknown
[16:03] <rgreening> ok
[16:03] <rgreening> I'll give that a shot
[16:09] <rgreening> ScottK: I expect a big patch to remove boolean may fix this. It's in 44 places in the package.
[16:09] <ScottK> That's not so bad.
[16:09] <rgreening> :) just time consuming.. but hopefully this is all thats required to fix openerp
[16:23] <broheem> hello everyone. noob question: i need to add php-imap support to a current ubuntu server setup. if i run "sudo apt-get install php5-imap" will that leave the current setup intact and just add on imap? or do i need to run that command with all the modules/addons i currently have enabled?
[16:24] <giovani3> broheem: it will install the module, it doesn't remove your other modules
[16:24] <broheem> giovani3: perfect, thank you! i wasn't sure if that was replacing php5 with the new module or just adding it on. that answer my question. take care.
[16:25] <giovani3> broheem: no problem -- that goes for all modules in debian/ubuntu -- in some cases, though, there will be multiple packages that replace one another such as xxx-mysql versus xxx-pgsql
[16:25] <giovani3> because only one kind of support is built in at a time, for example
[16:26] <broheem> giovani3: ah, yes, that makes sense. thanks for the info. much appreciated!
[16:26] <giovani3> np
[16:27] <ivoks> domas: sorry, i wasn't here
[16:27] <ivoks> domas: if you are really eagar to do it, of course, you can set up ppa
[16:27] <ivoks> domas: we could then just pull the diff and review it
[16:28] <domas> ivoks: nah, I mean, I'd just pull package from intrepid and make it compiled for hardy
[16:28] <domas> ivoks: intrepid resolves most of bad stuff
[16:29] <ivoks> hehe ok
[16:29] <domas> thats good
[16:29] <domas> except that we run mostly LTS :)
[16:29] <ivoks> well, LTS isn't sealed
[16:29] <ivoks> fixes are always welcome
[16:53] <mathiaz> soren: did you push your opennebula modifications to my iso_testing scripts somewhere?
[17:20] <soren> mathiaz: No, I didn't, sorry.
[17:34] <rgreening> ScottK: I think I figured out how to make the rdepend on python-xml for openerp-server go away :)
[17:35] <ScottK> rgreening: That's an even better solution for openerp-server.
[17:35] <rgreening> :)
[17:35] <rgreening> 1 file to fix
[17:36] <rgreening> a couple of small changes to osv/orm.py ScottK
[17:36] <ScottK> Ideally we want python-xml to go away ....
[17:36] <rgreening> and it then uses internal python xml (xpathEval)
[17:37] <ScottK> Perfect.
[17:37] <rgreening> Im testing here. Will let you know how it all goes.
[17:37] <MHS_> Hello. I am setting up a server for a number of computers. The server will hand out an IP address however I am having trouble in getting it to accept the server as a router to the internet. Can anyone help me?
[17:44] <cemc> MHS_: what do you mean 'accept the server as a router' ?
[17:45] <MHS_> The computers can't access the internet through the server
[17:48] <cemc> MHS_: is ip forwarding enabled? check cat /proc/sys/net/ipv4/ip_forward
[17:49] <MHS_> I think you meant type in that to the terminal. If you did then after I typed it in it said 0
[17:50] <MHS_> I also went directly to the file and it also shows nothing
[17:50] <giovani3> MHS_: then that means that ip forwarding isn't enabled
[17:50] <giovani3> so your server isn't acting as a router
[17:51] <MHS_> How can I change this?
[17:51] <giovani3> did you ever try to set it up as a router?
[17:51] <MHS_> No.
[17:51] <Tuxist> hi
[17:51] <giovani3> ok -- it's not the most simple of tasks
[17:52] <Tuxist> i have problem with nfs, pam and kerberos
[17:52] <giovani3> I wouldn't advise it if you don't feel pretty comfortable with linux
[17:52] <Tuxist> when i ount the nfs volume i can't login via pam kerberos
[17:52] <MHS_> I'm comfortable enough that I believe I can at least try. If you could possibly help me that is.
[17:52] <giovani3> MHS_: https://help.ubuntu.com/community/Router
[17:53] <giovani3> that's ONE guide -- there are many
[17:53] <giovani3> I don't know if it's complete
[17:53] <MHS_> Thank you
[17:53] <Tuxist> have anybody solutions for this problem
[17:53] <giovani3> google is your friend
[17:53] <Tuxist> i think nfs overwrite the krb5cc_0
[17:53] <MHS_> Thank you for all the help. I know. An I've looked but I just couldn't find the right wording I suppose.
[18:50] <rgreening> ScottK: Im having a diff time getting postgres properly setup to test out the openerp changes...
[18:51]  * ScottK hasn't set up postgres in a very long time.
[18:51]  * ScottK would just say ask here ....
[18:52] <rgreening> I have set up in the past... but never had much difficulty...
[18:52] <rgreening> I think its openerp related.
[19:18] <Hans67521> hello all
[19:18] <Hans67521> how do i remove the login warranty message for each user?
[19:20] <_ruben> edit /etc/motd
[19:20] <ivoks> err
[19:21] <ivoks> edit /etc/motd.tail
[19:21] <Nafallo> edit /dev/ivoks
[19:21] <ivoks> :)
[19:21] <_ruben> oh .. right .. forgot about that one :p
[19:28] <Hans67521> thanks
[19:28] <Hans67521> dont have a /dev/ivoks
[19:30] <giovani3> Hans67521: 'twas a joke, disregard
[19:55] <Jeeves_> Hi all!
[20:02] <moz_> my system is having trouble booting
[20:02] <moz_> it seems to hang after "loading hardware drivers.."
[20:02] <moz_> there is a message saying [8.8xxxxxx] header request #8
[20:03] <moz_> then there is no hard drive activty, anyone know whats happened?
[20:03] <Jeeves_> moz_: Try booting without 'quiet' at the bootloader
[20:03] <Jeeves_> That will give you more output
[20:04] <moz_> [8.847701] header request #1 is actually the message
[20:05] <moz_> Jeeves_: how do I do that sorry?
[20:05] <Jeeves_> moz_: When it's starting grub
[20:05] <Jeeves_> hit escape
[20:05] <moz_> yup
[20:05] <Jeeves_> and than you can edit the boot-options
[20:05] <Jeeves_> someone figured that it is better to boot a server silent
[20:08] <Hans67521> can i run fetchmailconf from terminal without X
[20:09] <giovani3> Hans67521: you can't ... fetchmailconf is a GUI
[20:09] <giovani3> it just edits .fetchmailrc though, so you can do that manually
[20:10] <slops17> hey all
[20:10] <Hans67521> i have a fetchmailrc file for a user
[20:10] <Hans67521> how do i run it
[20:10] <Hans67521> or if i have 10 users how do i run it
[20:10] <moz_> Jeeves_: it seemed like when i changed that it loaded ok, i guess that is just coincidence it couldnt have fixed it ?
[20:10] <slops17> i am trying to do an install of ubuntu server on my server but after the install it takes me to a grub command line but i cant do anything
[20:11] <giovani3> Hans67521: I don't know what you mean by "run it" -- the .fetchmailrc file is a preferences file
[20:11] <Hans67521> ok
[20:11] <slops17> i am using the server cd iso
[20:11] <Hans67521> but how to i run to to download mail for a users
[20:11] <Hans67521> user
[20:14] <User777> hello
[20:14] <cemc> Hans67521: you run fetchmail, which uses the .fetchmailrc file for options
[20:15] <Jeeves_> moz_: hmm, might be a coincidence
[20:15] <User777> okay so i just downloaded and installed ubuntu server with LAMP..(its actually a turnkey Joomla appliance) and although Everything seems set up correctly in interfaces and resolve i cant resolve..any ideas?
[20:15] <Jeeves_> only one way to find out :)
[20:15] <Hans67521> mmm.. i am lost
[20:15] <Jeeves_> User777: cat /etc/resolv.conf
[20:16] <Hans67521> i a put in set daemon 60 in the users fetchmailrc file will it then automaticaly check mail every 1min?
[20:16] <User777> cat /etc/resolve.conf list the 2 nameservers i put and they are correct
[20:17] <oly562> question - does the ubuntu - server run gui-less by default?
[20:17] <Jeeves_> User777: And they are reachable?
[20:17] <Jeeves_> oly562: Of course
[20:17] <oly562> why of course?
[20:18] <slops17> any one?
[20:18] <User777> they are reachable
[20:18] <Jeeves_> slops17: type enter
[20:18] <slops17> i did
[20:18] <oly562> why does the ubuntu server take so long to boot?
[20:18] <slops17> i am trying to do an install of ubuntu server on my server but after the install it takes me to a grub command line but i cant do anything
[20:18] <oly562> vs desktop and gui, actually the desktop and gui load faster than the gui-less server
[20:19] <oly562> is there a way to fix that?
[20:19] <Jeeves_> oly562: That depends what the reason is
[20:19] <oly562> im speaking from a default install no real services added yet...
[20:19] <oly562> no reason, its default settings im concerned with
[20:19] <oly562> try to follow what im saying, it will make more sense
[20:19] <moz_> my wireless USB stick is disconnect very frequently, sometimes after 5 mins of use sometimes 20 mins, is it possible is it over heating, the device is still being recognized on my laptop that is running unbuntu-server, but it will not ping anything, could this be overheating
[20:19] <moz_> or with overheating should it not be recognized
[20:20] <User777> Jeeves_ ?
[20:20] <moz_> if i replug it in, it will usually work again for a while
[20:20] <slops17> jeeves_ i am trying to do an install of ubuntu server on my server but after the install it takes me to a grub command line but i cant do anything
[20:20] <oly562> ok, i will try this again... the gnome ubuntu desktop loads 10x faster than the ubuntu - server os. why is this? note, i haven not loaded any services on the server, other than the defaults... and.... is there a way to fix this? if you know how to speed up the default server great, if not, i dont need to type extra kystrks.
[20:21] <Jeeves_> Ok
[20:21] <Jeeves_> so now there are three peope asking me 6 questions
[20:21] <oly562> im asking the room
[20:21] <Jeeves_> User777: are you firewalling stuff>
[20:21] <User777> im only asking one question as such
[20:21] <Jeeves_> ?
[20:22] <User777> i havent touched any firewall settings
[20:22] <Jeeves_> User777: That's not the question
[20:22] <Hans67521> does anyone know fetchmail
[20:22] <Hans67521> i need help
[20:22] <Jeeves_> oly562: Try finding out what needs time during booting
[20:22] <Jeeves_> Hans67521: man fetchmailrc
[20:22] <oly562> ill wait for a few, then i will try back later, nothing i can find on docs tell me why it takes longer than a window 2000 server to boot up
[20:22] <henkjan> oly562: you didnt install an tasksel like 'mailserver' or something?
[20:22] <oly562> no henkjan
[20:23] <oly562> i did not
[20:23] <oly562> although i would like to load services next
[20:23] <henkjan> oly562: i remember sendmail taking a longtime to start if not configured correct
[20:23] <User777> sorry, its address is behind a firewall yes, but other VM's with windows work fine out of the box
[20:23] <oly562> but if the server is going to take that long, i will simply use freebsd for my server needs
[20:23] <oly562> im trying out ubuntu this year
[20:23] <Jeeves_> oly562: Sure. Blame Ubuntu for your cluelessnes
[20:23] <oly562> well debian i mean
[20:23] <User777> other vm's on the same subnet
[20:23] <User777> i mean^
[20:24] <Jeeves_> User777: Still not answering the question
[20:24] <oly562> im not running a freeradius server Jeeves.
[20:24] <henkjan> oly562: in the grub menu you can edit the boot parameters
[20:24] <Jeeves_> User777: Hint: sudo iptables -L -n -v
[20:24] <oly562> or apache ssl that i can automajically get to boot with out halting
[20:24] <oly562> yesh....
[20:24] <henkjan> oly562: remove the 'splash' line
[20:24] <oly562> dont assume anyone is cluess from a few lines in a chat room
[20:24] <oly562> henkjan:  oh? you think that will work?
[20:24] <henkjan> oly562: er,, remove the 'slash' word and boot
[20:25] <oly562> i dont see any splash, are you saying in grub?
[20:25] <Jeeves_> henkjan: oly does not understand what he is doing
[20:25] <Jeeves_> I give up :)
[20:25] <henkjan> oly562: by removing splash your server wil show you whats starting up
[20:25] <oly562> Jeeves_:  is a dick
[20:25] <oly562> ;)
[20:25] <Jeeves_> oly562: Indeed!
[20:25] <oly562> hmmm
[20:25] <Jeeves_> But at least i'm a smart dick!
[20:25] <User777> chain input (policy accept 0 packets, 0 bytes)
[20:25] <User777> err..sorry thats incomplete
[20:25] <oly562> where can i remove splash...
[20:25] <henkjan> reboot the server
[20:26] <User777> pkts bytes target  prot opt in  out  source    destination
[20:26] <oly562> it starts up in cmdline, there is no splash pic, gui, notta. i dont see where the splash line would be
[20:26] <henkjan> in the grub boot menu you should be able to edit the parameters
[20:26] <Jeeves_> !paste
[20:26] <User777> is the same for FORWARD and OUTPUT
[20:26] <oly562> the server is not on at this time, im just asking for furture ref..
[20:26] <Jeeves_> User777: See !paste
[20:26] <henkjan> oly562: if the servers is starting up in cli. on which line is it 'hanging' ?
[20:26] <User777> ah, yes sorry, thanks
[20:27] <oly562> yah, i need to find that out... henkjan
[20:27] <oly562> its been a month since i messed with it
[20:27] <oly562> ill do that. and return later
[20:27] <henkjan> oly562: okay
[20:27] <oly562> i sorta remember, something, but not definitive
[20:27] <Jeeves_> henkjan: that's about the same I said before :)
[20:27] <henkjan> oly562: and don't call jeeves a dick again!!!!!
[20:27] <Jeeves_> henkjan: It seems to help when you repeat stuff I say :)
[20:28] <oly562> but one more thing, where would i remove this splash entry? in grub.conf?
[20:28] <Jeeves_> oly562: #kopt in /boot/grub/menu.lst
[20:28] <oly562> i call'em as i see'em
[20:28] <oly562> ic.. menu.list got it
[20:28] <henkjan> Jeeves_: its my authority!
[20:28] <Jeeves_> henkjan: Ah!
[20:28] <Jeeves_> Enlighten me!
[20:29] <Jeeves_> oly562: No, menu.lst
[20:29] <oly562> one last little question, about wireless ;) oh and Jeeves how bout just in grub dir
[20:29] <henkjan> Jeeves_: recoverd from an busy saterdayafternoon? ;)
[20:29] <Jeeves_> henkjan: Almost :)
[20:29] <Jeeves_> oly562: Try writing a phrase
[20:29] <User777> here it is    http://paste.ubuntu.com/136238/
[20:30] <Jeeves_> That might help people understand what you try to say
[20:30] <oly562> ok,  can i manually change my wireless nic signal, such as 2.4. to say, something higher, manually? i do not know if the nic is 5.x ghz but i do know its 2.4. its a wireless G nic
[20:30] <Jeeves_> User777: Ok, so no firewalling
[20:30] <oly562> im thinkin i can change the freq
[20:30] <Jeeves_> User777: Now paste what's in /etc/resolv.conf
[20:30] <oly562> any suggestions? if not, not big
[20:31] <Jeeves_> oly562: No clue
[20:31] <oly562> i will want to set up a wireless net on my server, is why
[20:32] <User777> ahhhhh!
[20:32] <oly562> henkjan: ? ur thoughts on wireless change of nic strength
[20:32] <User777> lol..Thanks so much Jeeves!
[20:32] <User777> i dont know what file name i typed in but nano must have created something elese
[20:32] <oly562> ok, ill be back at a later time.. Njoy... l8
[20:32] <User777> that resolve was blank so i filled it in and now reolving is happening
[20:32] <Jeeves_> User777: coolio
[20:32] <Jeeves_> have fun
[20:33] <henkjan> olcafo: no thoughts on that. wireless is for homi!
[20:34] <Jeeves_> :)
[20:34] <henkjan> hmm, sorry olcafo. that wasnt for you
[21:05] <domas> hi!
[21:08] <Jeeves_> Hi!
[21:08] <_ruben> ho!
[21:13] <Jeeves_> Wat nou!
[21:19] <rgreening> ScottK: ping
[21:19] <ScottK> rgreening: Pong
[21:19] <rgreening> ScottK: I really hate postgres
[21:19] <mathiaz> kirkland: you mentionned that kvm 84 supported kvm-in-kvm.
[21:19] <mathiaz> kirkland: is this also true for the hardy backport?
[21:19] <rgreening> I still cant get postgres + openerp running
[21:19] <kirkland> mathiaz: true, if you have an amd64 cpu
[21:19] <ScottK> rgreening: Don't hate postgres.  It's not mysql after all.
[21:19] <kirkland> mathiaz: and if you're running the kvm-source module
[21:20] <mathiaz> kirkland: is there any specific options to be passed to kvm?
[21:20] <kirkland> mathiaz: yes, i actually done that on my hardy backport test machine
[21:20] <kirkland> mathiaz: yes, there is ...
[21:20] <rgreening> ScottK: I have no idea where to go... I've read everything I can to see about getting this running.
[21:20] <mathiaz> kirkland: I've been some beta testing and there is the last test case that deals with a virtual host - http://testcases.qa.ubuntu.com/Install/ServerWhole
[21:20] <kirkland> mathiaz: -enable-nesting enable support for running a VM inside the VM (AMD only)
[21:20] <mathiaz> kirkland: ah - and libvirt in hardy doesn't support that?
[21:21] <kirkland> mathiaz: oh, hmm, probably not
[21:21] <ScottK> rgreening: There are probably people here who can help you with the postgres stuff.
[21:21] <kirkland> mathiaz: i got a post on my blog that makes me think that we might need to backport libvirt too, if we go this route
[21:21] <mathiaz> kirkland: ok - I'll give a try using the command line then.
[21:22] <mathiaz> kirkland: that wouldn't suprise me - we should think about the whole stack
[21:22] <mathiaz> kirkland: and backporting the same version as in jaunty
[21:22] <mathiaz> kirkland: so that we can be a more confident that things are stable in hardy
[21:22] <kirkland> mathiaz: agreed
[21:22] <rgreening> ScottK: this is the bug I am running into... http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=516348
[21:23] <mathiaz> kirkland: at least we know that the combination of the different packages (version wise) work on Jaunty.
[21:23] <ScottK> So at least it's not just you.
[21:23] <kirkland> mathiaz: right
[21:23] <mathiaz> kirkland: anyway - thanks for the pointer to the kvm option. I'll give it try later.
[21:25] <ScottK> rgreening: Were the last comments helpful?
[21:25] <kirkland> mathiaz: sure thing
[21:25] <rgreening> ScottK: nope. not really...
[21:26] <ScottK> rgreening: I've got no great ideas for you.
[21:26] <rgreening> he
[21:26] <rgreening> heh
[21:27] <rgreening> stupid package being packages and not properly tested!
[21:27]  * rgreening needs to kick someone
[21:27] <Jeeves_> rgreening: ask someone to bend over
[21:27] <rgreening> :)
[21:27] <rgreening> volunteering?
[21:27] <rgreening> :P
[21:27] <Jeeves_> Nope :)
[21:28] <Jeeves_> btw
[21:28] <User777> hello again
[21:28] <Jeeves_> why do we boot servers without any usefull output?
[21:28] <Jeeves_> Why not remove splash and quiet from the grub commandline?
[21:28] <Jeeves_> User777: Hi
[21:28] <mathiaz> kirkland: the RAID bug (boot from degraded raid) is still there?
[21:29] <kirkland> mathiaz: yes, that's what i'm working on
[21:29] <User777> so..on this ubuntu server..i just installed and (i think) configured correctly proftpd....however i cant connect from filezilla on a windows machine....any ideas?
[21:29] <kirkland> mathiaz: the latest mdadm release in debian solves the problem
[21:29] <kirkland> mathiaz: i'm trying to isolate the fix
[21:29] <kirkland> mathiaz: otherwise, i have a merge prepared
[21:29] <Jeeves_> User777: Did you select inetd or standalone?
[21:29] <User777> inetd
[21:29] <kirkland> mathiaz: what do you think would be preferred?
[21:29] <Jeeves_> User777: Is inetd installed?
[21:30] <Jeeves_> User777: By default, it isn't
[21:30] <User777> ah
[21:30] <User777> i did not know that
[21:30] <Jeeves_> which is kinda dumb in the proftpd package
[21:30] <Jeeves_> User777: it's because of the 'no open ports by default' policy, i believe
[21:30] <User777> will installing inetd now make it work..or will i have to then redo proftpd?
[21:30] <Jeeves_> User777: The install should do the trick
[21:31] <User777> inetd isnt in repository?
[21:32] <mathiaz> kirkland: hm - looking at the changelog.
[21:32] <mathiaz> kirkland: would the fix be included in 2.6.7.2?
[21:33] <kirkland> mathiaz: the changelog entry seems to indicate so
[21:33] <kirkland> mathiaz: i cherrypicked the one non-documentation patch in that release
[21:33] <kirkland> mathiaz: this did NOT solve the problem
[21:33] <mathiaz> kirkland: merging 2.6.7.2 would be easy since it's a bug fix only release
[21:33] <kirkland> mathiaz: so i merged 2.6.8-12-gb47dff6-2
[21:33] <kirkland> mathiaz: which did solve the problem
[21:34] <User777> okay su just installed inetutil-inetd and am rebooting
[21:35] <kirkland> mathiaz: this one looks tasty too: http://git.debian.org/?p=pkg-mdadm/mdadm.git;a=commit;h=2b4ca8f079335c1b3f345ec13da58699aaa0269d
[21:35] <kirkland> mathiaz: let me try a merge of 2.6.7.2
[21:38] <User777> thanks you so much
[21:38] <User777> inetd did it
[21:39] <User777> your the man (and or woman)
[21:41] <mathiaz> kirkland: hm - according to ANNOUNCE-2.6.8: Release 2.6.8 contains various bug fixes and minor enhancements.
[21:42] <mathiaz> kirkland: It seems like that fits the criteria for a FFexception.
[21:42] <mathiaz> kirkland: I wouldn't upload 2.6.8 with an Exception as 2.6.8 is not a bug-fix *only* release.
[21:42] <kirkland> mathiaz: right
[21:42] <mathiaz> kirkland: I wouldn't upload 2.6.8 *without* an Exception as 2.6.8 is not a bug-fix *only* release.
[21:42] <kirkland> mathiaz: i'm trying to merge 2.6.7.2
[21:45] <cellofellow> changing the system password doesn't change the courier password. How do I change the courier pop/imap password? I've tried restarting the courier daemons but to no avail.
[21:48] <cellofellow> never mind, I just didn't actually change my password :P
[21:50] <mathiaz> kirkland: hm - regarding the RAID bug - booting a degraded RAID5 array works
[21:50] <kirkland> mathiaz: okay, got a 2.6.7.2 merge building
[21:50] <kirkland> mathiaz: lemme test
[21:50] <mathiaz> kirkland: the RAID5 array boots with only 2 disks active.
[21:50] <kirkland> mathiaz: with the mdadm that's in the archive?
[21:50] <mathiaz> kirkland: however a RAID1 array doesn't boot
[21:50] <mathiaz> kirkland: yes.
[21:53] <kirkland> mathiaz: \o/
[21:53] <kirkland> mathiaz: sweet, a merge of 2.6.7.2 works
[21:53] <kirkland> mathiaz: let me upload the source to p.u.c
[21:53] <kirkland> mathiaz: would you mind building/testing to confirm this?
[21:53] <mathiaz> kirkland: no problem.
[21:53] <mathiaz> kirkland: http://people.ubuntu.com/~mathiaz/raid5_boot.png
[21:54] <mathiaz> kirkland: ^^ this is the degraded raid5 array booting.
[21:54] <mathiaz> kirkland: the same error message is still there - however the system is able to boot.
[21:55] <kirkland> mathiaz: right, same here
[21:55] <mathiaz> kirkland: http://people.ubuntu.com/~mathiaz/raid5_boot_2.png <- the following messages
[21:55] <kirkland> mathiaz: i get the error message
[21:57] <kirkland> mathiaz: http://people.ubuntu.com/~kirkland/mdadm
[21:57] <kirkland> mathiaz: you should be able to dget the source
[21:57] <kirkland> mathiaz: note that i haven't sufficiently documented the merge in the changelog yet
[22:02] <mathiaz> kirkland: http://paste.ubuntu.com/136289/
[22:02] <mathiaz> kirkland: is the error message about MAKEDEV normal?
[22:03] <kirkland> mathiaz: hmm, that doesn't look right
[22:03] <kirkland> mathiaz: and i just ran it here, same thing
[22:04] <olcafo> is ethereal not in the repositories?
[22:05] <giovani3> olcafo: ethereal changed its name to wireshark in 2006
[22:06] <mathiaz> kirkland: 2.6.7.2 doesn't fix the RAID bug.
[22:06] <olcafo> oh, thanks. Odd that I could'nt find this info online.
[22:06] <mathiaz> kirkland: the system fails to boot from a degraded RAID1 array.
[22:06] <kirkland> mathiaz: fark
[22:07] <giovani3> olcafo: what were you searching for?
[22:07] <mathiaz> kirkland: http://people.ubuntu.com/~mathiaz/raid1_boot.png
[22:07] <giovani3> olcafo: the second google result for "ethereal" is wireshark
[22:08] <mathiaz> kirkland: do you have an idea why a degraded RAID5 array would boot while a degraded RAID1 wouldn't?
[22:10] <kirkland> mathiaz: i don't ...  but i almost want to throw this back to kernel space
[22:10] <kirkland> mathiaz: b/c we're still getting that gnarly kernel message
[22:10] <kirkland> mathiaz: which simply cant be right
[22:11] <kirkland> mathiaz: okay, i've verified that the /dev/MAKEDEV problem has been around for before the current changes
[22:11] <kirkland> mathiaz: not that it's right, but it's been there for a little while
[22:12] <ScottK> SDGathman: Right.  I understand that.  I don't think the pypolicyd-spf is packaged for Fedora.  That's likely easy enough.
[22:12] <ScottK> Sorry.  Wrong channel.
[22:13] <kirkland> mathiaz: try this ...
[22:13] <xyberpix> Hi all
[22:14] <kirkland> mathiaz: sudo dpkg-reconfigure mdadm
[22:14] <kirkland> mathiaz: choose the boot-degraded=yes
[22:14] <xyberpix> I really hope that someone can help me out here
[22:14] <kirkland> mathiaz: see if that gets you to boot degraded
[22:14] <xyberpix> I'm getting major slow responses when copying files to my 8.10 samba server
[22:14] <xyberpix> has anyone got any pointers?
[22:14] <xyberpix> I've added the default socket options as well
[22:17] <xyberpix> so no one's had problems with samba on Ibex then?
[22:18] <kirkland> xyberpix: https://bugs.edge.launchpad.net/ubuntu/+source/samba
[22:19] <xyberpix> kirkland: tnx
[22:20] <mathiaz> kirkland: nope - I'm still dropped to a busybox prompt.
[22:20]  * Jeeves_ is gone
[22:20] <Jeeves_> see ya
[22:20] <mathiaz> kirkland: that's with mdadm 2.6.7.2 and after having run dpkg-reconfigure mdadm with boot degraded set to yes.
[22:20] <kirkland> mathiaz: jeebus
[22:21] <mathiaz> kirkland: I didn't get the question on reboot and got dropped directly to the busybox prompt
[22:21] <mathiaz> kirkland: I can see the same error messages
[22:21] <kirkland> mathiaz: right
[22:21] <kirkland> mathiaz: that's the expected behavior when you answer the question ahead of time
[22:21] <kirkland> mathiaz: okay, try this ...
[22:23] <kirkland> mathiaz: install an upstream 2.6.29 kernel from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.29-rc8/
[22:24] <kirkland> mathiaz: in my experience, that made the error messages, as well as the problem go away
[22:24] <mathiaz> kirkland: ok.
[22:24] <mathiaz> kirkland: I also noticed in the mdadm 2.6.8 changelog entry mention about adding support for kernel 2.6.28
[22:24] <kirkland> mathiaz: oh really ....
[22:25] <kirkland> mathiaz: do you want to try my 2.6.8 merge too?
[22:25] <kirkland> mathiaz: before updating your kernel?
[22:25] <rgreening> ScottK: The docs for setting up openerp v5 are all wrong.
[22:25] <mathiaz> kirkland: sure -
[22:26] <mathiaz> kirkland: let's try 2.6.8
[22:27] <ScottK> rgreening: Sounds fun ....
[22:27] <rgreening> ScottK: not really. Just found the correct way, and it crashes in building the db
[22:27] <rgreening> go figure.
[22:27] <kirkland> mathiaz: http://people.ubuntu.com/~kirkland/mdadm
[22:35] <kirkland> mathiaz: any better luck with 2.6.8?
[22:37] <kirkland> mathiaz: fwiw, i just tested 2.6.29 and it boots degraded very cleanly, exactly as expected, no nasty error messages
[22:37] <mathiaz> kirkland: \o/
[22:37] <mathiaz> kirkland: 2.6.8 boots
[22:37] <mathiaz> kirkland: http://people.ubuntu.com/~mathiaz/raid1_boot_success.png
[22:38] <mathiaz> kirkland: there is the same error message though.
[22:38] <kirkland> mathiaz: right
[22:38] <kirkland> mathiaz: but 2.6.8 is definitely an improvement
[22:38] <mathiaz> kirkland: but the system boots from the degraded RAID1 array.
[22:38] <kirkland> mathiaz: i can clean up the merge changelog entry and try to get an FFe approved
[22:38] <kirkland> mathiaz: can you also try the 2.6.29 kernel?
[22:39] <kirkland> mathiaz: you should see no error message at all there
[22:39] <mathiaz> kirkland: FYI I have oopses on dmesg
[22:39] <mathiaz> kirkland: probably related to the error message
[22:39] <kirkland> mathiaz: yeah
[22:39] <mathiaz> kirkland: I'm going to try 2.6.29.
[22:40] <mathiaz> kirkland: I though I tried your 2.6.8 merge
[22:40] <mathiaz> kirkland: hm - nm.
[22:41] <kirkland> ?
[22:41] <mathiaz> kirkland: yeah - I was reading the scrollback
[22:41] <kirkland> mathiaz: what version of mdadm did work?
[22:41] <mathiaz> kirkland: 2.6.8
[22:41] <kirkland> mathiaz: k
[22:41] <mathiaz> kirkland: the one currently in http://people.ubuntu.com/~kirkland/mdadm/
[22:41] <kirkland> mathiaz: k
[22:41] <kirkland> mathiaz: yeah, that's 2.6.8
[22:41] <kirkland> mathiaz: fwiw, 2.6.7.2 also worked for me
[22:42] <kirkland> mathiaz: but you reported that that one did not work for you
[22:42] <mathiaz> kirkland: yes - 2.6.7.2 didn't work for me.
[22:42] <kirkland> mathiaz: okay, i'm glad you're around to double-check these
[22:42] <mathiaz> kirkland: what are you test cases?
[22:43] <mathiaz> kirkland: I'd like to know why your tests are successful while mine aren't
[22:43] <kirkland> mathiaz: boot with both disks
[22:43] <kirkland> mathiaz: make sure they're sync'd
[22:43] <kirkland> mathiaz: reboot once with both disks, make sure that works
[22:43] <kirkland> mathiaz: install some version of mdadm
[22:43] <kirkland> mathiaz: poweroff
[22:43] <kirkland> mathiaz: boot with one of the 2 disks attached
[22:44] <kirkland> mathiaz: rejoice on boot, cry on initramfs
[22:46] <mathiaz> kirkland: I've just rerun dpkg-reconfigure mdadm on a system that is already running a degraded RAID1 array.
[22:47] <mathiaz> kirkland: and ask to not boot from a degraded array.
[22:47] <kirkland> k
[22:47] <kirkland> mathiaz: now you'll get the prompt
[22:47] <mathiaz> kirkland: however on boot the system still boots without asking a question
[22:47] <kirkland> mathiaz: oh, right
[22:47] <kirkland> mathiaz: note the wording of the question ...
[22:47] <kirkland> mathiaz: "newly degraded"
[22:48] <mathiaz> kirkland: yes :) - so it's the expected behavior.
[22:48] <kirkland> mathiaz: once a system has booted a degraded raid, it expects that that is the current behavior
[22:48] <mathiaz> kirkland: ?
[22:48] <kirkland> mathiaz: we've discussed this before ....
[22:48] <mathiaz> kirkland: ok.
[22:48] <kirkland> mathiaz: this is why my test case includes this step:
 mathiaz: make sure they're sync'd
[22:49] <kirkland> mathiaz: so after i'm done with a round of testing, to "reset", i boot with both disks attached
[22:49] <kirkland> mathiaz: re-add to the array if necessary, and wait for the sync to complete
[22:49] <kirkland> then
 mathiaz: reboot once with both disks, make sure that works
[22:49] <mathiaz> kirkland: right.
[22:50] <mathiaz> kirkland: BTW 2.6.29 doesn't show any error message
[22:50] <kirkland> mathiaz: \o/
[22:50] <kirkland> mathiaz: would you confirm that in https://bugs.edge.launchpad.net/ubuntu/jaunty/+source/linux/+bug/334994 ?
[22:50] <kirkland> mathiaz: i *firmly* believe this is a kernel issue
[22:50] <mathiaz> kirkland: 2.6.8 is able to boot from degraded RAID1 array - with 2.6.28 there are messages in the kernel log, while 2.6.29 is quiet
[22:51] <kirkland> mathiaz: yes
[22:51] <kirkland> mathiaz: that's exactly my experience
[22:53] <mathiaz> kirkland: wrt to you tests - I'm not doing the reset step when the system was failing to boot from a RAID1 array.
[22:54] <mathiaz> kirkland: Multiple reboot would still lead to have the question asked.
[22:54] <mathiaz> kirkland: I guess this is the expected behavior.
[22:54] <kirkland> mathiaz: right, until you successfully boot with one disk
[22:54] <kirkland> mathiaz: which you hadn't really done before you used the 2.6.8 package
[22:54] <kirkland> mathiaz: once you've done that, you've told the system, "It's cool--we can boot with one disk"
[22:54]  * mathiaz nods
[22:55] <mathiaz> kirkland: so why 2.6.7.2 was working for you but not for me?
[22:55] <kirkland> mathiaz: and it says, "Okay, cool.  Degraded is now the known state of your array."
[22:55] <kirkland> mathiaz: that's a good question
[23:06] <mathiaz> kirkland: bug 334994 updated
[23:06] <mathiaz> kirkland: I've added the kernel oops from /var/log/kern.log
[23:09] <mathiaz> kirkland: hm - I may have found a regression with kvm 84 on hardy. It seems that the shutdown command doesn't send the proper ACPI signal to the guest
[23:09] <mathiaz> kirkland: I used to be able to do a clean shutdown on the guest with the command virsh shutdown vm-id.
[23:10] <mathiaz> kirkland: that doesn't seem to work anymore.
[23:10] <mathiaz> kirkland: (at least with a jaunty guest)
[23:11] <mathiaz> kirkland: are you aware of this?
[23:11] <kirkland> mathiaz: i am not
[23:12] <mathiaz> kirkland: I'll do more testing to make sure it's kvm 84 and not jaunty that is broken
[23:12] <mathiaz> kirkland: I'll file a bug against kvm if needed.
[23:12] <kirkland> mathiaz: okay
[23:29] <moz_> at the moment i am setting the key and essid of my wireless usb device by using "iwconfig" and dhclient to connect to it, is there a way fro mthe terminal I can set all those things to happen automatically instead of doing it each itme
[23:29] <moz_> *time
[23:31] <kirkland> mathiaz: https://bugs.launchpad.net/bugs/337763
[23:31] <kirkland> mathiaz: that might be your acpi issue?
[23:35] <mathiaz> kirkland: let me try with a hardy guest and see if the shutdown command is sucessful
[23:35] <mathiaz> kirkland: hm - seems like it.
[23:36] <mathiaz> kirkland: virsh shutdown hardy-base is able to sucessufuly shutdown the guest
[23:36] <kirkland> mathiaz: so hardy shuts down, but jaunty does not?
[23:36] <mathiaz> kirkland: yes.
[23:36] <mathiaz> kirkland: and this is with kvm 84 on a hardy host
[23:36] <kirkland> sounds like a bug in jaunty kernel, right?
[23:36] <olcafo> I have this problem with Intrepid host running 8.10 and 7.10 clients.
[23:37] <olcafo> havent' tried hardy guests.
[23:37] <mathiaz> kirkland: yes I don't think it's the same bug as reported by jdstrand
[23:37] <mathiaz> olcafo: do you have the acpid daemon running in your guests?
[23:38] <mathiaz> olcafo: and did you boot your guest with acpi enabled (from the host)?
[23:38] <olcafo> 7.10's is not. I'm not sure about 8.10
[23:38] <mathiaz> olcafo: you need acpid running in your guest other wise the ACPI shutdown signal is not process by your guest.
[23:39] <olcafo> that machine is not available to me at the moment. I'll have to check that later.
[23:40] <olcafo> mathiaz, that makes sense. I didn't really think of it as a problem that needed fixing though :P, just an inconvinience at the moment.
[23:58] <orudie> hi, anyone familiar with zoneedit? i need some help with their nameservers, its kind of confisuing
[23:58] <kirkland> mathiaz: http://git.debian.org/?p=pkg-mdadm/mdadm.git;a=blobdiff_plain;f=Assemble.c;h=3ee028b461619a1379827e1bfce84c245a1aebfc;hp=1d3787022adec681aec9ddbb3e3ef98e2dfcfb35;hb=4e9a6ff778cdc58dcc6897e74cf5ee1d3f73e1f7;hpb=22eba5121632c4b26541fbb04209a90b93c24bcc
[23:59] <mathiaz> kirkland: this is a very simple fix :)