[01:30] <j3z> hi
[01:31] <j3z> i have an error while booting mount: mount point /dev/pts does not exist
[01:31] <j3z> when i create that dir by hand it works
[01:32] <j3z> but then, i have to mount rest vfses by hand, how to automate creation of that directory and what could be the source of that problem?
[01:32] <j3z> the same sytuation i have with /dev/shm
[01:33] <j3z> and yes, my /var/lib/udev/devices contents are correct (both directories exist there)
[01:53] <twb> j3z: this is 10.04?
[01:57] <j3z> no, it's even 9.10
[01:59] <twb> Is a program called "mountall" installed?
[02:00] <j3z> yes
[02:00] <j3z> when i invoke it, i get that error
[02:00] <twb> Pastebin the contents of /lib/init/fstab
[02:01] <twb> Wait, never mind
[02:01] <twb> mountall is correctly trying to mount it, but (say) udev has failed to create the mountpoint
[02:01] <j3z> stat: cannot stat `/lib/init/fstab': No such file or directory
[02:01] <twb> j3z: Are you running a stock Ubuntu kernel?
[02:03] <j3z> it's a VPS, kernel is provided by linode :/
[02:03] <twb> OK, I give up, talk to them about it
[02:03] <j3z> i thought that it could be some incompability between kernel and userland
[02:03] <twb> I was already just guessing
[02:04] <j3z> i thought that in some magic way i can create these dirs ;)
[02:04] <twb> I have had lots of problems with ubuntu's sucky init when using virtualization
[03:50] <zkirill> Should all my files in /var/www/ be owned by root or am I doing something wrong?
[04:07] <zkirill> Can I create an ssh key for a system user?
[04:15] <jeeves__> has anyone here come accross VMWare not restarting after a reboot?  I'm getting "VMware Server Host Agent: failed
[04:46] <twb> zkirill: you can; I advise against it.
[04:46] <zkirill> twb: Thanks, already figured it out. And I had no choice :(
[04:47] <twb> What are you actually trying to achieve?
[05:11] <gobbe> zkirill: if your apache/lighttpd-user has read access to www-files it's ok that root owns them
[05:11] <twb> www-data, IRIC
[05:11] <zkirill> gobbe i just chowned the folder in question to www-data
[05:12] <zkirill> seemed to solve the problem
[05:12] <twb> permissions on directories managed by dpkg are not preserved on package upgrade
[05:13] <twb> *I* would do it by telling apache to serve /foo out of /srv/foo or /home/foo/public_html or so, and simply tell the user foo to create files in there.
[05:13] <twb> "serve /foo" and is http://example.net/foo
[05:15] <gobbe> zkirill: yea
[05:15] <kees> ScottK, hallyn_: https://bugs.launchpad.net/ubuntu/+source/clamav/+bug/691414  not sure what's going on. maybe the db itself has gone weird?
[06:20] <munyua> Greetings
[06:20] <munyua> problem with server
[06:22] <munyua> how do I install ubuntu server on the dell poweredge r310 I cant get past the detect hard drive part, I cant get the right drivers for the Perc s300
[06:22] <munyua> hey there?
[07:12] <laen|work> Hello, good morning, how are you *tutututuuu*
[07:14] <xampart> how can i check the pid of sshd providing my current connection
[07:19] <laen|work> xampart: check your tty with command tty, and then ps -ef | grep 'sshd:.*<that tty you got>'
[07:20] <laen|work> I bet there's a better way, but oh well.
[07:20] <xampart> laen|work: k
[07:49] <ewook> so, who broke clamav-daemon? :)
[08:05] <laen|work> I wish it was just clamav ;).
[11:51] <marcriera> hello
[11:51] <marcriera> I would like to migrate to 10.04 from debian 4 and debian 5.
[11:51] <marcriera> hints?
[11:56] <cnu_> hi......
[11:57] <cnu_> can some one help me out....
[12:23] <RoyK> someone should add to the topic that if people don't ask with what to help them, well, they should leave at once
[12:23] <RoyK> like cnu_ here
[12:23] <RoyK> marcriera: why on earth would you want to do that?
[13:49] <ScottK> kees: Looks like it's time to push a microversion update for clamav (based on the discussion in the bug)
[13:50] <ScottK> RoyK: I find the security hardening features a reasonably compelling reason to be using Ubuntu instead of Debian for servers.
[13:50] <zul> Daviey: ping where are we with axis2c? ;)
[14:13] <consumerism> where are apache access logs located? i have a bunch of vhosts set up without individual log files but the only thing in /var/log/apache2 is error log
[14:18] <zul> /var/log/apache2
[14:24] <Daviey> zul: axis2c is untouched
[14:25] <zul> Daviey: ok thanks
[14:42] <JamesPage> hggdh: hey - sorry missed you yesterday - well done on getting the ltp test running in a vm through Hudson
[14:43] <hggdh> JamesPage: thanks :-)
[14:43] <hggdh> JamesPage: now I just have to find out why libvirt gets to own the ISOs
[14:44] <JamesPage> hggdh: that could be a tricky one (and is why I choose lucid for my two nodes)
[14:44] <JamesPage> hggdh:  maverick+ does some security related permissions changes when you start/stop a virtual machine
[14:45] <hggdh> JamesPage: yes... but it does not make sense to chown the ISOs, so I will open a bug for it
[14:46] <Daviey> cjwatson: Is bug #689343 a dupe of your recent finding?
[14:49] <cjwatson> Daviey: hard to tell, have followed up
[14:49] <Daviey> cjwatson: Yeah, i wasn't quite sure either - i thought it *might* be. :/
[14:50] <Daviey> thanks
[14:52] <JamesPage> hggdh: I temp disabled xango3 this morning as it was chewing through test failures v fast....
[14:52] <JamesPage> hggdh: and you can now access without the port  - http://hudson.qa.ubuntu-uk.org/
[14:53] <hggdh> JamesPage: yes, due to xango3 was barfing on being unable to update the ISOs -- bllody libvirt-bin got them
[14:53] <hggdh> JamesPage: COOL!
[15:01] <Daviey> JamesPage / hggdh: What is the process for using the results from hudson?
[15:02] <hggdh> Daviey: no firm stance so far, AFAIK. There is the qadashboard that can summarise the data
[15:02] <hggdh> Daviey: otherwise, right now, it is undefined
[15:02] <Daviey> hggdh: Interesting, where is qadashboard?
[15:04] <bobboau> hello, I'm having an issue with GD, for some reason it just stopped working, has there been an update recently that might have broken it?
[15:04] <Daviey> hggdh / JamesPage: What caused the recent failures?  Looks like an issue not with the ISO's. :/
[15:06] <Daviey> bobboau: Possibly... need to see an error message or traceback :/
[15:08] <bobboau> I'm not getting any error messages, the image just doesn't show up on the client. when I don't set the headers it looks like it is trying to send binary data, but I don't know anything about the jpeg file format to try and trouble shoot from that.
[15:10] <bobboau> so, it acts as though everything is fine, but for some reason the stream it sends out is invalid in some way. when I look at the raw dump I don't see any error messages embedded int the binary data or anything.
[15:11] <hggdh> Daviey: http://reports.qa.ubuntu.com/reports/qadashboard/qadashboard.html
[15:12] <hggdh> Daviey: akt-tbench barfed on the ISO
[15:12] <Daviey> bobboau: Check there isn't an error message at the top, in the jpeg output stream
[15:12] <Daviey> or deprecation warning
[15:12] <Daviey> hggdh: oh
[15:13] <bobboau> I have looked for that, I see nothing.
[15:13] <Daviey> bobboau: url?
[15:14] <hggdh> Daviey: I will look at the other failures soon; some of them may be due to differences on d-i/pressed on Natty
[15:14] <Daviey> hggdh: do you still need an extra box for hudson?
[15:14] <hggdh> Daviey: we certainly do :-)
[15:14] <bobboau> the server is internal and fire-walled but a copy of the jpeg it produces is here: http://www.hexellent.com/files/80/fuct.jpeg
[15:14] <hggdh> the more the merrier
[15:14] <Daviey> hggdh: gimme some instructions :)
[15:15] <hggdh> Daviey:  on the box, install ubuntu-server-iso-testing and hudson-slave (from JamesPage's PPA, just a sec)
[15:15] <Daviey> bobboau: the problem is that you have a white space line in the output stream
[15:17] <Daviey> bobboau: http://erk.daviey.com/fuct.jpeg
[15:17] <bobboau> at the beginning or end? also why would it suddenly break, shouldn't it have been broken the whole time?
[15:17] <Daviey> bobboau: Without seeing code, i can't comment :/   .... there is an empty line at the top of the file
[15:18] <Daviey> bobboau: Perhaps an empty echo statement, adding for debugging?
[15:19] <hggdh> Daviey: add ppa:james-page/usit to your apt sources list
[15:19] <hggdh> Daviey: then install ubuntu-server-iso-testing and hudson-slave from there
[15:19] <hggdh> Daviey: then we will have to go into Hudson and add your new slave
[15:20] <bobboau> I did find a bit of whitespace at the end of one of the include files, but after getting rid of that it still doesn't work...
[15:20] <hggdh> Daviey: so tell me when you are ready, and I will step you through
[15:21] <JamesPage> hggdh, Daviey: sorry I missed that - stepped out for a coffee :-)
[15:21] <Daviey> bobboau: add an echo 'foo'; statement just before the GD stream, and see if the whitespace is before or after the GD output.
[15:21] <Daviey> JamesPage: slacker :P
[15:22] <Daviey> hggdh: Well... my goal is to preseed setting up a slave, so i can take it out of rotation if i need it for other things ; but easily put it back in.
[15:22] <Daviey> (especially in times of demand)
[15:22] <hggdh> Daviey: this is the same I will do on two new machines I will be installing locally in a few days
[15:22] <hggdh> and the same I currently do with my laptop
[15:23] <Daviey> hggdh: you got the new hardware \o/
[15:23] <JamesPage> Daviey: that should be pretty easy now; the hudson-slave package now includes an upstart script so everything can be automatically started
[15:23] <hggdh> Daviey: not yet, not yet, PO has been submitted
[15:23] <Daviey> ah
[15:23] <Daviey> JamesPage: I am moderately excited!
[15:23] <JamesPage> Daviey: let me just check that
[15:24] <bobboau> before foo, it seems. it is the newline character you are seeing right? 0x0A?
[15:24] <JamesPage> (did it on Tuesday PM while in pre Flu stupor)
[15:25] <Daviey> JamesPage: heh
[15:25] <JamesPage> yep - all up to date.  it does need a manual configure at the moment but its pretty easy /etc/default/hudson-slave
[15:25] <Daviey> bobboau: Ah, you are seeing it before foo... that means that GD is certainly not responsible
[15:26] <Daviey> bobboau: I think you need to move that foo, further back into the code and see when it appears before the output :)
[15:27] <Daviey> JamesPage: If that is the only change, i can happily wget that as a post install.
[15:27] <JamesPage> Daviey: OK - I might manage to get it setup using debconf but not pre xmas....
[15:28] <Daviey> JamesHarrison: meh.. not sure the rewards is worth it :)
[15:28] <Daviey> err, JamesPage
[15:28] <JamesPage> Daviey: suggest you use lucid as a few nuances with qemu/permissions on maverick+
[15:28] <blackxored> hey, I woke up sleepy today, how can I cancel and purge a postfix queue, i can't remember?
[15:28] <JamesPage> Daviey; thought that you had the wrong window for a moment
[15:28] <Daviey> JamesPage: Oh, ok
[15:30] <bobboau> ... ok, it seems php or apache is prepending the output with a newline, I get it if echo 'foo;'; is the first instruction in the script or if simply foo is at the very start of the file before the <?php tag
[15:31] <bobboau> so this is probly not related to gd at all and is some sort of config issue...
[15:31] <hggdh> Daviey: the "few nuances" is bloody libvirt chown to itself the ISOs -- so when you later try to zsync them it fails with a errno=13
[15:32] <Daviey> blackxored: I seem to remember last time i needed to do that, i used a loop ... output of postqueue -p  into postsuper -d $MAIL-ID
[15:32] <bobboau> so, /*thinking out-loud*/ what would cause apache to prepend all output with a newline...
[15:32] <qjcg> does anyone know if a do-release-upgrade (8.04 -> 10.04) will work if the server is a xen domU?
[15:32] <blackxored> thanks I'll try that, but there was an easier way ;)
[15:32] <qjcg> or are there any problems I should be aware of?
[15:32] <Daviey> bobboau: Interesting!  I can't help further without smelling the code :(
[15:33] <Daviey> zul might be the best person to speak to qjcg.
[15:33] <Daviey> (AKA ping zul)
[15:34] <qjcg> Daviey: thanks!
[15:34] <zul> qjcg: there shouldnt be any problems if you are updating the domU
[15:34] <bobboau> ok, well at least I'm not chasing that turkey anymore, you'd probably need the whole codebase to do anything, and I'm not quite desperate to make that sort of an upload yet
[15:35] <Daviey> bobboau: heh
[15:35] <bobboau> this is a cakephp based app, I should probably poke my head in their channel
[15:35] <qjcg> zul: and on the dom0 side, I'm guessing the only thing that needs to be done is a kernel upgrade
[15:35] <qjcg> (ie the kernel the domU boots from)
[15:35] <Daviey> bobboau: bake it so.
[15:36] <zul> qjcg: you are basically on your own dom0 side
[15:36] <JamesPage> hggdh: FYI jibel has taken the ubuntu-server-iso-testing project and adapted it for desktop testing using the same model
[15:36] <JamesPage> hggdh: we agreed to spend some time at the platform rally consolidating the two pieces of work.
[15:37] <hggdh> JamesPage: nice -- -server-iso-testing is already overloaded
[15:39] <qjcg> zul: ok, I'll go ahead and see what happens... thanks!
[15:44] <consumerism> is there a way to see a live output of the size of a directory? i'm copying a large directory and want a better monitoring option than running du -hs  every minute or two
[15:46] <hggdh> JamesPage: I opened bug 691590 for ISO and libvirt
[15:47] <JamesPage> hggdh: excellent; I'm setting up a couple of tests the try to compare the install speeds between natty and maverick as well.
[15:47] <Daviey> consumerism: I'd use rsync :)
[15:48] <hggdh> heh
[15:49] <Daviey> hggdh / JamesPage: Hmm... i just had a *CRAZY* idea.
[15:49]  * hggdh likes crazy ideas
[15:49] <Daviey> hggdh / JamesPage: Is it possible to easily get the queue size, so i can do WoL if demand warrants it?
[15:49] <JamesPage> Daviey: fire away
[15:51] <hggdh> hum
[15:51] <lunaphyte_> i'm having trouble with aptitude search.  the man page says "the command "aptitude search '~N edit'" will only show "new" packages whose name contains "edit" - but it seems like i'm seeing different behavior : http://dpaste.com/287735/ - what am i doing wrong?  i want to return all installed packages whose names begin with dhcp
[15:51] <JamesPage> Daviey: let me take a look
[15:51] <JamesPage> Daviey: maybe - http://hudson.qa.ubuntu-uk.org/queue/api/
[15:52] <Daviey> interesting..
[15:54] <Daviey> JamesPage: eeek, installing apt-cacher-ng ?
[15:55] <JamesPage> Daviey: yep - apt is configured to use a local cache on the host to improve speed/limit download sizes
[15:55] <Daviey> JamesPage: Hmm, this box is probably not ideal, it's i386 install :/
[15:55] <JamesPage> Daviey: we can limit it to running i386 jobs;
[15:56] <hggdh> with a tag
[15:56] <Daviey> oh cool
[15:57] <Daviey> JamesPage: What change to rc.local did you want to change?
[15:58] <Daviey> I already pressed 'no' as i was about to give up due to the arch.
[16:04] <JamesPage> Daviey: it does a 'fix' to tidy up the mess that libvirt-bin and dnsmasq make of each other
[16:04] <JamesPage> not pretty
[16:04] <Daviey> JamesPage: Have a diff handy? :)
[16:07] <JamesPage> Daviey: http://paste.ubuntu.com/544910/
[16:08] <JamesPage> Daviey: sorry - just re-read - you mean't rc.local right
[16:09] <JamesPage> http://paste.ubuntu.com/544911/
[16:10] <Daviey> yup
[16:15] <Daviey> JamesPage: Has vector shown up?
[16:15] <JamesPage> Daviey: not yet - let me just create his entry in Hudson (hostname vector?)
[16:15] <Daviey> ack
[16:16] <JamesPage> he has now  - and its trying to run something that won't work :-(
[16:17] <JamesPage> how fast is your Internet connection?
[16:18] <hggdh> JamesPage: I have found that, on my laptop, rc.local does not seem to work
[16:18] <Daviey> JamesPage: it is zsync'ing atm
[16:19] <JamesPage> hggdh: yeah - its an in-elegant solution.
[16:19] <hggdh> i.e., I still have to stop/restart libvirt, pkill dnsmasq and service dnsmasq start again
[16:19] <Daviey> JamesPage: sucking down at 880kB/s
[16:19] <JamesPage> hggdh: and it does not really work with upstart
[16:19] <hggdh> JamesPage: heh
[16:20] <hggdh> will look into it more
[16:20] <JamesPage> hggdh: what we really want is a fix for the bug (i'll dig it out) which means dnsmasq and libvirt-bin won't co-exist on the same server
[16:21] <Daviey> JamesPage: I assume dnsmasq is listening on localhost only?
[16:21] <hggdh> JamesPage: agreed -- or, perhaps, have all apps using dnsmasq start using the dbus interface
[16:21] <hggdh> Daviey: no
[16:22] <Daviey> !
[16:22] <JamesPage> Daviey: dnsmasq listens on the bridge so that it can do PXE/TFTP kungfoo with the guests
[16:22] <Daviey> ah, that is fine
[16:22] <Daviey> just wanting to make sure it wasn't going to take over the network :)
[16:22] <JamesPage> this conflicts with the one that libvirt-bin sticks up by default (even if its not configured to todo DHCP)
[16:23] <hggdh> Daviey: we work by steps, first the local machine, then the world ;-)
[16:23] <Daviey> lol
[16:23] <JamesPage> listen-address=192.168.122.1
[16:24] <JamesPage> don't worry!
[16:24] <Daviey> JamesPage: Hmm,, can i kill this zsync ?
[16:24] <Daviey> it's grabbing amd64
[16:24] <JamesPage> I can - not much use is it.
[16:25] <JamesPage> let me kick off a i386 test
[16:26] <Daviey> JamesPage: great
[16:26] <JamesPage> Daviey: so there is a sea of red at the moment
[16:26] <Daviey> i noticed :)
[16:26] <JamesPage> issue 1 is hggdh's qemu/permissions bug on xango3
[16:27] <JamesPage> basically it fails fast so while bowlan and mercury are working through their test cases;
[16:27] <JamesPage> xango3 speeds through the other 22 and fails them all.
[16:27] <JamesPage> (now disabled.....)
[16:27] <JamesPage> The other one is that tests are timing out; there is a 40 minute limit on the vm execution at which point the script gives up.
[16:28] <JamesPage> terminates the vm and fails the test.
[16:28] <JamesPage> we are hitting this *alot* with natty
[16:28] <hggdh> yes
[16:28]  * JamesPage worries that natty installs seem to take a lot longer than maverick
[16:28] <hggdh> resons unknown -- in my case I noticed (but did not confirm yet) that VBox is faster than KVM (!?)
[16:29] <JamesPage> http://hudson.qa.ubuntu-uk.org/job/natty-server-amd64_default/buildTimeTrend
 Daviey: otherwise, right now, it is undefined <- sorry :(
[16:31] <hggdh> UndiFineD: I am really sorry...
[16:31] <UndiFineD> :P
[16:32] <hggdh> JamesPage: and -- everything going fine -- soon-to-be xango4 and xango5 :-)
[16:32] <Daviey> JamesPage: I suspect it might be the sync's with ext4... there is now an option in dpkg to avoid these, making install faster... might be a good idea to check with cjwatson.
[16:33] <hggdh> Daviey: being correct, we can add it easily in the base preseed
[16:33] <hggdh> Daviey: do you know the option? I can add it locally, and test
[16:35] <cjwatson> there's no point changing it now
[16:35] <cjwatson> there's going to be another dpkg release in a couple of days which changes things around some more
[16:35] <hggdh> cjwatson: thanks, we wait, then :-)
[16:35] <cjwatson> (1.15.8.7 - I'll merge it when I see it in unstable)
[16:36] <cjwatson> chat on #debian-dpkg suggested Sunday
[16:36] <hggdh> cjwatson: a Q: if I select "manual package selection" on d-i, is aptitude expected to start?
[16:36] <hggdh> under tasksel
[16:36] <cjwatson> yes
[16:37] <hggdh> and -- after I select the packages, I should 'g' on aptitude, correct?
[16:37] <cjwatson> IIRC
[16:37] <hggdh> I did that, and had some of the packages fail to install (but seems it recovered later)
[16:37] <hggdh> will have to repeat
[16:42] <Daviey> JamesPage: Can you fire the same test again?
[16:43] <JamesPage> Daviey: yep - doing
[16:43] <JamesPage> Daviey: well I could if I knew what I was doing :-)
[16:44] <Daviey> JamesPage: curently idle :(
[16:44] <JamesPage> Daviey: yeah - I forgot about bowlan (also idle so picked up the job)
[16:44] <Daviey> awesome :)
[16:45] <Daviey> JamesPage: Fire *another* job? :)
[16:46] <JamesPage> Daviey: just taking bowlan offline
[16:46] <Daviey> JamesPage: cool
[16:49] <JamesPage> Daviey: there we go
[16:49] <JamesPage> Daviey: looks like its going for the full ISO again
[16:50] <hggdh> lol
[16:50] <Daviey> JamesPage: I have confidence it will work :)
[16:51] <Daviey> JamesPage: I think it might have been my fault
[16:51] <Daviey> :)
[16:51] <JamesPage> Daviey: what did you touch?
[16:52] <Daviey> I cleaned up the amd64 images ...  /var/lib/ubuntu-server-iso-testing/isos/ubuntu-server/*
[16:52] <Daviey> ... i386 started quicker than i expecte
[16:52] <Daviey> d
[16:52] <JamesPage> you little tinker!
[16:53] <Daviey> heh
[16:54]  * Daviey blames hggdh 
[16:54] <JamesPage> I remember a sys admin I once work with tidied up a production webserver once
[16:55] <JamesPage> Got the wrong root directory - /
[16:55] <JamesPage> not a great moment.
[16:55] <Daviey> eek
[16:55] <Daviey> JamesPage: Have you seen the "webscale" video?
[16:55] <JamesPage> Fortunately it was farmed :-) (and yes I have)
[16:56] <JamesPage> and we had automated rebuild
[16:56] <Daviey> http://pb.daviey.com/WwP2/raw/
[16:56] <Daviey> JamesPage: ^^
[16:56] <Daviey> reminded me of that.
[16:56] <smoser> JamesPage, still around ?
[16:56] <smoser> Daviey, did you use pastebinit to load that ?
[16:56] <smoser> i've heard it has support for pb.daviey.com now
[16:56] <Daviey> smoser: maybe :)
[16:56] <JamesPage> Daviey: nice
[16:57] <Daviey> smoser: Yes, much sought after support aiui
[16:57] <Daviey> smoser: The media is calling it an early xmas pressent
[16:57] <JamesPage> smoser: sure am
[16:58] <smoser> http://paste.ubuntu.com/544934/
[16:58] <Daviey> yeah, 29seconds after you said something :)
[16:58] <JamesPage> :-)
[16:58] <JamesPage> I type slowly....
[16:58] <smoser> JamesPage, ^ so that is the list of 23 libraries used by ec2-api-tools
[16:58] <JamesPage> smoser: thats a nice list of Java
[16:59] <smoser> so there are 23 libraries there, the majority of them packaged in ubuntu, but zero of them packaged and of the "correct" version
[16:59] <smoser> i'm guessing that you're aware that this is the state of the world.
[16:59] <JamesPage> so it seems to be.
[17:00] <smoser> i'm just kind of wondering what you think could be done about it
[17:00] <Daviey> *should* anything be done?
[17:00] <JamesPage> well some will be OK - most of the commons-* packages maintain backwards compat.
[17:00] <smoser> (in that list, the un-indended things are the ubuntu package and then version) the ones under them are the file that ec2-api-tools provides
[17:01] <consumerism> what's the quickest way to delete a large directory?
[17:01] <smoser> Daviey, i dont think anything shoudl be done explicitly to address ec2-api-tools.
[17:01] <smoser> rm -Rf dir &
[17:01] <JamesPage> others are pretty close; think we can find most of your unpackaged list as well
[17:01] <smoser> or, if you need to recreate it right away: mv dir dir.old && mkdir dir && rm -Rf dir.old &
[17:01] <smoser> JamesPage, oh ?
[17:02] <JamesPage> they seem to ring bells
[17:02] <smoser> i did'nt see anything when i went looking with apt-cache and apt-file search
[17:02] <consumerism> smoser: rm -rf is the only way i know to do it, there's not a quicker way? i have read somehwere emptying the directory first makes it faster
[17:02] <smoser> but... you're more aware than i
[17:02] <JamesPage> what part of the archive is ec2-api-tools
[17:02] <smoser> so basically, if i can, it seems "right" to use packaged versions, and depend on the java packages, right ?
[17:03] <smoser> JamesPage, do you agree with that ? or should i not bother.
[17:03] <Daviey> JamesPage: Seen the lovely failure?
[17:03] <smoser> multiverse. other than those libraries, its closed source.
[17:03] <bobboau> haza! fixed my problem, I had a newline at the end of a random model file!
[17:03] <Daviey> bobboau: \o/
[17:03] <JamesPage> Daviey: the couchdb instance might be in a bit of a twist.
[17:04] <Daviey> JamesPage: tips for fixoring?
[17:04] <JamesPage> Daviey: sometimes it does not startup on the bridge interface
[17:04] <smoser> consumerism, i dont know... it could be that rm is for some reason less than ideal in how it implements -R, but really, that would be a bug / feature request against it.
[17:04] <JamesPage> Daviey: you could try a restart of couchdb but a reboot might be better :-(
[17:04] <Daviey> hmm
[17:05]  * JamesPage thinks the install script needs to work better
[17:05] <Daviey> need to see if i can reboot this
[17:05] <Daviey> rebooting
[17:07] <JamesPage> smoser: good question re packaged/bundled versions....
[17:07]  * JamesPage thinks
[17:07] <Daviey> smoser: get to the back of the queue, JamesPage is mine atm :)
[17:07] <JamesPage> Daviey: we're on a break - you are rebooting
[17:07] <Daviey> JamesPage: wanna fire that job again?
[17:08] <Daviey> JamesPage: Doesn't make it ok to see other people!
[17:08] <smoser> JamesPage, so for the packaging i did in iamcli recently, I did depend on the ubuntu packages and did not install as part of the binary package the .jar files that were in the "source" package.
[17:08] <smoser> that seemed to work well, and in some cases means smaller download for someone with 'apt-get install iamcli'
[17:09] <smoser> i was going to try to do that here, but it seemed like so many near misses.
[17:09] <JamesPage> Daviey: going now
[17:09] <Daviey> watching :)
[17:10] <JamesPage> smoser: well it might but that would depend on how many deps they already had installed.
[17:10] <Daviey> JamesPage: couch still seems poorly
[17:10] <Daviey> (/me grumbles about couch)
[17:11] <JamesPage> smoser: as the ec2-api-tools is prop. it may be better to stick with bundled versions.
[17:12] <JamesPage> smoser: if something does break compat you can't fix is in ec2-api-tools
[17:12] <smoser> hm... it just doensn't seem right.
[17:12] <smoser> well, i can fix it by reverting to the included version in an updated package.
[17:12] <JamesPage> smoser: I guess
[17:12] <Daviey> JamesPage: Oddly, couch is running at port 5984 is working as expected
[17:13] <JamesPage> Daviey: check that is running on 192.168.122.1:5984
[17:13] <smoser> but i'm completely willing to accept your sage advice
[17:14] <JamesPage> smoser: I'm nervous about 'break compat you can't fix is in ec2-api-tools'
[17:14] <JamesPage> smoser: let me just look at the list again
[17:15] <Daviey> JamesPage: not getting response on that IP
[17:15] <Daviey> infact....
[17:15] <Daviey> a traceroute looks even worse
[17:15] <JamesPage> Daviey: can you check that you have a local config file in /etc/couchdb/local.d
[17:16] <Daviey> JamesPage: I don't have the br
[17:16] <MrPicard> Hello all!, Im running ubuntu on the EC2 cloud, ive used the offical Concanical Cloud AMI however it seems i cannot add new users and the server when ftping refuses new users and requires them to use an auth key file. is there anyway to get around this? i just want a working ubuntu LAMP server.
[17:16] <JamesPage> No bridge?
[17:16] <Daviey> nope!
[17:17] <JamesPage> OK; try stopping dnsmasq, restarting libvirt-bin and see if bridge returns
[17:17] <Daviey> nope
[17:18] <smoser> JamesPage, i'll be back in 30 minuts or so.
[17:19] <Daviey> JamesPage: it's bwoken. :(
[17:19] <JamesPage> smoser: sorry I will be gone by then
[17:19] <MrPicard> so, is there anything i can do?
[17:19] <JamesPage> Daviey: this is related to the race condition between libvirt dnsmasq and normal dnsmasq
[17:20] <JamesPage> for some reason this does not happen on my lucid servers.
[17:20] <Daviey> JamesPage: How did you confirm this issue?
[17:20] <MrPicard> i guess not <.<
[17:20] <JamesPage> Daviey - if you disable dnsmasq it libvirt should startup fine including the bridge.
[17:21] <JamesPage> Daviey: killing the dnsmasq instance is spawns and then restarting dnsmasq got me past this issue (hence rc.local stuff)
[17:23] <Daviey> :(
[17:23] <JamesPage> This is the bug 231060
[17:23] <Daviey> JamesPage: Is there any way for me to fire to jobs?
[17:23] <MrPicard> ..... its a shame no one can help me with my ec2 problem.
[17:23] <JamesPage> Daviey: yes - I can set you up with an account
[17:28]  * Daviey takes a break
[17:28] <zul> Daviey kirkland: im going to update the git snapshot
[17:29] <kirkland> zul: sounds good!
[17:29] <kirkland> zul: how's it looking?
[17:29] <zul> kirkland: lunch was good if that is what you were asking :)
[17:29] <kirkland> zul: :-)  did you have "cobbler" for lunch?
[17:29] <kirkland> mmm... blackberry cobbler mm....
[17:29] <zul> kirkland: nope kd macroni
[17:30] <kirkland> zul: i'm having mexican for lunch today ;-)  Mmmm.... Manuels....
[17:34] <JamesPage> hggdh: I've left bowlan and mercury disabled in hudson while Daviey tries to get his node working.
[17:34] <JamesPage> nothing for them todo at the moment so not really a problem.
[17:34] <hggdh> JamesPage: ack. I will also keep xango3 off
[17:34] <JamesPage> cool
[17:40] <ScottK> kees: I just uploaded clamav 0.96.5 to lucid/maverick-proposed so we should be able to get that sorted reasonably quickly.
[17:44] <kees> ScottK: okay, thanks
[17:44] <ewook> lamont: well, so do I.
[17:46] <ewook> I'm curious. Is it just my upgrades that has caused the "switch" to another startup-manager to fail and leave most of the services dead after a reboot? (from 8.04 to 10.04)
[17:47] <qjcg> so, after my do-release-upgrade on a xen domU (8.04->10.04), my machine no longer boots cleanly (root mounted ro, other partitions not mounted at all...)
[17:48] <qjcg> my domU uses a kernel on the dom0 (2.6.24...xen)
[17:48] <qjcg> anyone know of a place where i can find xen kernels for 10.04?
[17:49] <ScottK> ewook: I did several 8.04 -> 10.04 upgrades without problem.
[17:51] <qjcg> or, any ideas why my domU no longer boots correctly? nothing in my config changed, just a do-release-upgrade that finished with no errors
[18:02] <ewook> ScottK: mkay.. all services starts without problems ?
[18:02] <ScottK> Yes.
[18:03] <ewook> ScottK: also, postgrey simply fails on upgrade now..
[18:03] <ewook> lucky bastard :P
[18:03] <ScottK> What's the error?
[18:05] <ewook> ScottK: http://pastebin.com/B4m6YN11
[18:06] <ewook> ScottK: and no, I did not do a custombuild, almost everything on that machine is from the repos
[18:07] <ewook> (otherwise, how would apt recognize postgrey?)
[18:08] <ScottK> Let me give it a try.
[18:08] <ewook> sure
[18:08] <ewook> notice that it's an upgrade
[18:09] <ewook> alas : tcp        0      0 127.0.0.1:60000         0.0.0.0:*               LISTEN      32046/postgrey.pid
[18:09] <ewook> it gives an message about postgrey chaning ports to that exact port for compliance.. but yeah.. it's been "compliant" all the time :)
[18:10] <ewook> aaw, my membership with ubuntu-server has expired *_*.
[18:14] <ScottK> I tried installing the hardy postgrey in a lucid chroot and then upgrading it.  I got http://paste.ubuntu.com/544972/
[18:14] <ScottK> So that matches your experience.
[18:14] <ewook> oh.. darn. I was hoping it was just me :P
[18:15] <ewook> better check for a bugreport then
[18:15] <ewook> I forgot.. where those at again?
[18:16] <ScottK> bugs.launchpad.net.
[18:17] <ScottK> Also we get that package unmodified from Debian, so bugs.debian.org is worth a look too.
[18:17] <ewook> ScottK: aah. great. thanks
[18:18] <ewook> https://bugs.launchpad.net/ubuntu/+source/postgrey/+bug/537472
[18:18] <ewook> there we go
[18:19] <ewook> not really the right one, but still the same error
[18:19] <ewook> a fix commited for 4 months?
[18:19] <zul> kirkland Daviey: the new cobbler snapshot should be in the ppa in about 13 minutes
[18:22] <ewook> hrpmf. now to check out why init.d / service is breaking everything at boot
[18:23] <ewook> any pointers?
[18:25] <RoyK> ewook: is a service in init.d (or rc2.d) breaking things?
[18:25] <RoyK> if so, start reading that script
[18:25] <ewook> nono
[18:26] <ewook> since some services started to use "service" instead of the init-structure they simply will not start at boot
[18:26] <RoyK> oh
[18:26] <RoyK> that's upstart config
[18:26] <ewook> a
[18:26] <ewook> h
[18:27] <ewook> thank god ssh is still starting corretly
[18:27] <ewook> correctly
[18:27] <RoyK> iirc ssh doesn't use upstart :P
[18:27] <ewook> that explains it... :P
[18:29] <ewook> wow. I'm lost.
[19:00] <MrPicard> Friendly greetings, i have been having a minor issue with Ubuntu 10.4 LTS, sad to say when i type in sudo apt-get install mysql i get a error saying that ubuntu cannot find the package, any ideas?
[19:07] <_spacer_> hi room
[19:07] <_spacer_> sever newb here
[19:08] <_spacer_> i cannot access my server remotely via ssh, i could the otherday, but cannot today
[19:09] <zul> MrPicard: mysql-server is what you want
[19:09] <_spacer_> i cannot even ping my ip
[19:09] <MrPicard> Ta Zul :-)
[19:09] <_spacer_> but i am still serving webpages np
[19:12] <_spacer_> ports forwarded ok..
[19:16] <_spacer_> any idea what i can check?
[19:23] <Slyboots> God almight; Anyone here know much about getting WPAD configration pushing going?
[19:24] <Slyboots> I've configured DNSMasq to push the configuration by setting my Domain name (Then using a CNAME re-direct get get wpad.example.com to work; and using dhcp-option252,http://x.x.x.x/wpad.dat
[19:25] <Slyboots> But it seems that only the DNS push is working
[20:34] <ChmEarl> any estimate when natty will have linux-image-2.6.38-x-server out?
[20:34] <ChmEarl> weeks or months or after release?
[20:49] <latenite> Hi folks, my new ubuntu box can not go online. http://pastebin.com/E6kUYmT8 192.168..4.101 is my own DNS What did I do wrong?
[20:59] <patdk-wk_> not running a dns server on your own system?
[21:12] <soren> latenite: Your resolv.conf is wrong.
[21:13] <latenite> soren, why I have this on all my boxes?
[21:13] <soren> latenite: It's supposed to read "nameserver xxx.yyy.zzz.www", not just "xxx.yyy.zzz.www"
[21:13] <soren> latenite: I haven't the faintest idea.
[21:13] <latenite> soren, but doesnt ubuntu create the resolv.conf on startup?
[21:14] <soren> No.
[21:14] <teddymills> where are SSL certificates installed ?
[21:15] <hggdh> usually /etc/ssl
[21:15] <latenite> soren, no? like in No it not created on startup? why does it say : # Generated by NetworkManager ?
[21:16] <hggdh> resolv.conf is changed if using NetworkManager to manage your interfaces
[21:17] <hggdh> but I never saw NM writing this way
[21:17] <latenite> ah ok..it works with by adding the "nameserver" prefix. thanks guys :)
[21:21] <Slyboots> This crap is enough to break your will to live!
[21:22] <Slyboots> All I want to do; is block ads locally (including youtube video ads and that crap) and replace them with a 1x1 transparent pixel
[21:22] <Slyboots> So Id ont get huge freaking 503 DENIED! were the ads used to be
[21:22] <Slyboots> And yet I cnat seem to find an easy way to do this
[21:32] <OlsonCR> hi buddies
[21:32] <OlsonCR> i got problem with my 10.10 server
[21:33] <OlsonCR> that is giving me hard times
[21:33] <OlsonCR> any one there?
[21:34] <OlsonCR> dudes...?
[21:35] <teddymills> I have a Verisign SSL cert installed. I can view it via https:  but I cannot find the actual CERT file. It is ubuntu 10.10
[21:36] <teddymills> nothing in /etc/apache2/sites-available/default or default-ssl points to the CERT.
[21:36] <teddymills> It is the correct box, because I am using the same IP
[21:37] <teddymills> It is the correct box, because I am using the same public  IP
[21:38] <teddymills> there is an even an entire virtualhost container I cannot find.
[21:39] <teddymills> If I can find either the missing SSL cert or the missing VHOST container that is where the SSL is configured I am sure.
[21:39] <teddymills> i ahve search every line of /etc/apache2
[21:39] <teddymills> i ahve search every line of of most files in /etc/apache2
[22:00] <Olsoncr> hello?
[22:01] <sparc_> hello
[22:01] <Olsoncr> I got a problem with an external storage and 10.10 server
[22:01] <Olsoncr> hey sparc
[22:01] <Olsoncr> I got this situatiom
[22:01] <Olsoncr> I got an external USB drive on this 10.10 server
[22:02] <sparc_> there are other people that are much more skilled with ubuntu specifically, but maybe i can help
[22:02] <Olsoncr> there is where I mirror the backups of my company and keep them outsite
[22:02] <Olsoncr> dont worry
[22:02] <Olsoncr> let's try
[22:02] <Olsoncr> but..
[22:03] <Olsoncr> the ubuntu always disconnect the USB and reconnect the drive
[22:03] <Olsoncr> and the kernel assigns another letter..
[22:03] <Olsoncr> let's say /dev/sdb to /dev/sdc
[22:03] <Olsoncr> I have fixed that by making a udev rule..
[22:03] <Olsoncr> fine..
[22:04] <Olsoncr> no the drive is always (let's say) /dev/mydrive
[22:04] <sparc_> ok
[22:04] <Olsoncr> but since the drive is still getting disconnected I new have I/O errors on the mountpoint
[22:04] <Olsoncr> :(
[22:05] <sparc_> what kind of errors are you getting?
[22:05] <Olsoncr> I/O
[22:05] <Olsoncr> input output
[22:05] <sparc_> what do they say?
[22:05] <Olsoncr> ls: reading directory .: Input/output error
[22:06] <sparc_> oh
[22:06] <sparc_> darn that error doesn't say very much
[22:06] <Olsoncr> I'm pretty sure this is caused cuz the USB drive is disconnected by the system
[22:06] <Olsoncr> and connected again
[22:07] <sparc_> maybe someone knows right off the bat, but i would look for more verbose errors in dmesg and /var/log/messages
[22:07] <sparc_> i would also fsck the filesystem to see if there's anything wrong with it
[22:07] <sparc_> and use SMART checks to see if the device is healthy
[22:08] <sparc_> if those check out, then maybe there's something funny with udev, if you didn't have this problem before you wrote the rule?
[22:10] <Olsoncr> no..
[22:10] <Olsoncr> no problems with udev before
[22:10] <Olsoncr> actually this is happening on my desktop also.. when I reboot the HD loses his 'letter' and
[22:11] <Olsoncr> I have to remap all drives on /etc/fstab
[22:11] <Olsoncr> in this case
[22:11] <Olsoncr> on this server
[22:11] <Olsoncr> I am sure the FS is fine..
[22:11] <sparc_> i know you can put labels on ext filesystems
[22:11] <sparc_> and mount the filesystems based on labels instead of drive letters
[22:11] <Olsoncr> on messages I just have "usb disconneted" messajes
[22:12] <Olsoncr> udev is a lower lower than labels
[22:14] <Olsoncr> *lowe layer
[22:14] <Olsoncr> *lower
[22:14] <Olsoncr> sorry for my english
[22:15] <sparc_> your english is good, no problem
[22:16] <sparc_> i don't know what to say, i'm sorry.  maybe someone else has some ideas on how to get your disks to mount properly
[22:16] <sparc_> i still think checking the filesysm might not be a bad idea
[22:16] <Olsoncr> jus done buddy
[22:16] <Olsoncr> *just
[22:17] <sparc_> ok
[22:17] <Olsoncr> and problems still bothering me
[22:17] <Olsoncr> I may be a problem with the drive FS
[22:18] <Olsoncr> it's using fuseblk