[01:53] <alyoshka> for some reason I'm not able to access anything from a web server I just set up. I'm using Brightbox's version of Nginx for Ruby on Rails. I've set up the same thing before on another server and everything worked fine, but here it just gives me a gateway timeout error after a few minutes. if I try wget http://localhost from the actual server, it works, but it just won't serve to remote machines for some reason. anybody know w
[01:53] <alyoshka> that?
[02:00] <alyoshka> it's like as if Nginx is listening only to requests from localhost or something. anybody know why?
[02:02] <alyoshka> hmm, if I shut Nginx down, then remote machines fail to connect completely, but when it is running, it has a gateway timeout after like 10 minutes
[02:03] <alyoshka> so I guess it is listening, but failing to respond or something. why?
[02:26] <alyoshka> doesn't anybody know what to do about Nginx giving a gateway timeout error on absolutely everything when the request is from a remote machine?
[02:33] <bc> alyoshka: listening on a port other than 80?
[02:33] <alyoshka> bc: no, it's on 80
[02:34] <alyoshka> if I shut down nginx, the remote machine will just say connection refused, if nginx is running, it'll hand for about 10 minutes and then give a gateway timeout error
[02:34] <alyoshka> so nginx is getting the request, but it's not serving anything
[02:34] <alyoshka> no static files, no rails, nothing
[02:35] <bc> alyoshka: I know very little about nginx, but what do the logs say, if anything?
[02:39] <alyoshka> 2010/06/16 01:09:31 [alert] 2282#0: worker process 2297 exited on signal 9
[02:39] <alyoshka> 2010/06/16 01:19:38 [error] 2361#0: *5 upstream timed out (110: Connection timed out) while reading response header from upstream, client: xx.xxx.xx.xxx, server: myserver.com, request: "GET / HTTP/1.1", upstream: "unix:/tmp/passenger.2314/master/helper_server.sock:", host: "myserver.com"
[02:39] <alyoshka> I have a bunch of those in the error logs
[02:40] <alyoshka> but those are only on the rails site
[02:40] <alyoshka> the static site doesn't seem to have logged any errors
[02:43] <bc> alyoshka: I'm wondering if you have a syntax error or loop or but that is causing it to run longer than normal. However, there is an absolute ton on google about 'gateway timeout error rails nginx' -- have you checked those sources? I see a lot of references to a proxy timeout configuration directive that may apply, I'm not sure.
[02:43] <bc> s/but/bug/
[02:44] <alyoshka> I tried some of the suggestions there, but nothing worked. another thing is that most of the problems people had were only with the rails websites or with large files, but static files worked for them
[02:44] <alyoshka> for me, nothing is working
[02:45] <alyoshka> might have to go back to Apache, but it brings my small virtual servers to their knees
[02:49] <acajou> Hi all. I'm having an issue with a universe/admin package that works on my Kubuntu desktop but appears broken on a new Ubuntu server install...
[02:49] <acajou> The package is btrfs-tools.
[02:50] <bc> alyoshka: is that thing using fast-cgi?
[02:50] <acajou> When you install it, it's supposed to (amongst other things) create a /dev/btrfs-control file. It doesn't do that, which means nothing works, basically.
[02:51] <acajou> But I'm not sure if it's just my box or if the package is actually broken...
[02:51] <bc> alyoshka: does your config specify this directive anywhere? -> fastcgi_pass
[02:52] <acajou> Would it be best to open a bug report against the package?
[02:52] <alyoshka> bc: no, it's using passenger for rails and well, static files are served by the server itself
[02:56] <enav1> hi i need the wiki of this channel.... thanks
[02:56] <enav1> woops nv
[02:57] <bc> alyoshka: based on some pages I just looked at, your static files load fine?
[02:58] <acajou> if someone could try "sudo aptitude install btrfs-tools" on a 10.04 server install and tell me if they see /dev/btrfs-tools, that would be really wonderful...
[02:58] <acajou> I'd know if it's my box that has a problem, or the package itself.
[03:01] <bc> alyoshka: I think I'm in over my head. However, include helper_server.sock in the query I mentioned earlier.
[03:02] <alyoshka> bc: my computer crashed so I didn't get the last few messages u sent
[03:03] <bc> alyoshka: only one you missed besides the one I just sent was <bc> alyoshka: based on some pages I just looked at, your static files load fine?
[03:03] <bc> alyoshka: after that: <bc> alyoshka: I think I'm in over my head. However, include helper_server.sock in the query I mentioned earlier.
[03:03] <alyoshka> bc: the answer to that is no, the static files don't load either; nothing loads at all
[03:04] <bc> alyoshka: I saw a good bit of info related to the error log snippet you pasted + passenger.
[03:04] <bc> That is, with 'gateway timeout'
[03:05] <alyoshka> yeah, but most ppl who have problems with passenger don't have the problem with static files
[03:06] <alyoshka> I'll set up apache for now because I have a client that needs their site up today and I'll play with nginx later, I guess
[03:33] <cwillu_at_work> acajou, poke poek
[03:33] <cwillu_at_work> acajou, did you ever get answers re: btrfs on a server install?
[05:03] <hallyn> jdstrand: sweet, thanks
[05:31] <Phlogistique> Hi
[05:31] <Phlogistique> I need help
[05:31] <Phlogistique> Does anyone here have an Ubuntu Jaunty?
[05:32] <Phlogistique> If someone has this and could send me his /etc/{init.d,rc*} in a tar, I would be eternally grateful
[05:35] <twb> Phlogistique: you can get that from packages.ubuntu.com/jaunty/sysvinit-rc or so.
[05:35] <twb> dpkg -S /etc/init.d/ should tell you what packages you need.
[05:36] <twb> Note also that you can "aptitude reinstall <package name>"
[05:36] <twb> ...although for conffiles that probably won't help.
[05:36] <twb> In future, I also recommend running etckeeper to place your /etc/ under version control, which would allow you to simply revert the deletion.
[05:38] <Phlogistique> twb: I didn't delete anything
[05:39] <Phlogistique> it's far worse
[05:39] <twb> Go on...
[05:39] <Phlogistique> well I installed a Jaunty with the /etc of a Lucid
[05:39] <Phlogistique> and after like one month it's the first time I try booting it
[05:39] <twb> In that case I suggest you reinstall from scratch.
[05:40] <Phlogistique> haha
[05:40] <twb> You will not get /etc back to the way it was
[05:40] <Phlogistique> but for this I have to save everything
[05:40] <twb> Well, this is why you put /home and / on different filesystems
[05:40] <Phlogistique> I know. If I can get the system to boot and to connect, I will be satisfied
[05:40] <Phlogistique> twb: not everything is in /home
[05:40] <twb> Granted.
[05:41] <twb> When installing I would normally make a squashfs of / and put it in /root/<old hostname>.squashfs or so, so that I can easily restore data from it as needed.
[06:06] <Phlogistique> WOOOOOHOOOOOO IT BOOOT
[06:06] <Phlogistique> ZS
[06:07] <Phlogistique> *S
[06:07] <Phlogistique> YAY YAY YAY
[06:07] <Phlogistique> YES
[06:07] <Phlogistique> IT WORKS
[06:26] <a3ist> anyone know if there's a channel for freenx related questions, or know how to enable compositing for nx-hosted sessions on ubuntu-server?
[06:30] <twb> a3ist: I imagine that would require AIGLX to work over NX
[06:30] <twb> Unless you mean "work unbearably slowly"
[06:31] <a3ist> I'll check it out, thanks for the heads up
[06:31] <a3ist> I'm using gnome sessions hosted via nx to help some family members learn ubuntu
[06:32] <a3ist> awn needs compositing though
[06:32] <twb> Yes, well.
[06:33] <twb> IMO these new-fangled desktops that assume the they get direct rendering should FOAD
[06:33] <twb> If I wanted a network-opaque windowing system, I'd use glitz
[06:35] <a3ist> i mean hell, I could always set up x forwarding to the osx clients but the nomachine nxclient was a lot more user-friendly for the non-CLI-comfortable crowd
[06:36] <ruben23> hi how do i terminate a running application, not responding at all
[06:37] <a3ist> find its pid, use -kill
[06:37] <a3ist> or if theres only one instance of it you can do "killall <processname>"
[06:38] <twb> FSVO killall = pkill
[06:38] <a3ist> err sorry that's "kill <pid>", not "-kill"
[06:42] <scar> if that doesn't work, 'kill -9 <pid>' usually will
[08:49] <ttx> Daviey: should I try to reproduce the instance run / alpha1 issue ?
[09:07] <Daviey> ttx: If you have a spare cycle!
[09:09] <ttx> Daviey: installing a UEC rig as we speak.. repdocuing should not take very long
[09:11] <Daviey> ttx: thanks
[09:23] <huats> morning
[09:27] <ttx> huats: o/
[09:27] <huats> hello ttx !
[09:48] <ttx> Daviey: reproduced
[09:49] <ttx> Daviey: also seen elsewhere and also linked to 2.6.34: http://open.eucalyptus.com/forum/decrypting-image-exception
[09:49] <ttx> Should definitely be escalated to Eucalyptus for them to test on 2.6.34
[09:51] <Ng> kirkland: hrm, how come that qemu-kvm upload you sponsored doesn't show up on https://launchpad.net/ubuntu/+source/qemu-kvm ? can I monitor its progress into -proposed somewhere?
[09:53] <Daviey> ttx: Agree, seems they are a aware of it.. Really it's an issue for the kernel/mulesoft to resolve
[09:54] <Daviey> ttx: Will mention it in the weekly call.
[09:54] <ttx> I don't think Mule is in the path
[09:54] <ttx> It's a bouncycastle exception
[09:54] <Daviey> On another note, anyone else notice we don't have any daily cd images?
[09:55] <Daviey> ttx: ah yes
[09:57] <ttx> Daviey: I'll have a look at the crypto code in WalrusImageManager.java to check it's sane
[09:58] <Daviey> ttx: Sounds good to me, thanks
[09:59] <Daviey> ttx: There are two bugs that seem to have the same behaviour on the bouncycastle tracker
[09:59] <Daviey> However, both seem to be user issue.. but this could be also..
[10:00] <Daviey> (i doubt it)
[10:00] <ttx> Daviey: link ?
[10:00] <Daviey> http://www.bouncycastle.org/jira/browse/BJA-275
[10:00] <Daviey> http://www.bouncycastle.org/jira/browse/BJA-251
[10:01] <ttx> yes, those don't look like the same issue
[10:01] <Daviey> i agree.. but there are two potential pointers to check for a work around.
[10:02] <Daviey> ttx: I'm going to compose an email to euca upstream shortly regarding this (so they have more info ready for the call), any extra insight would be appreciated.
[10:03] <ttx> Daviey: yes, just make sure to point out that forum post as being the same issue
[10:03] <Daviey> will do
[10:03] <ttx> Daviey: the issue is either a kernel regression, something that needs to be changed in Bouncycastle to accomodate recent kernel changes, or in theur implementation of crypro decoding
[10:04] <Daviey> agreed
[10:04] <ttx> though that last one looks sane to me
[10:04] <Daviey> ttx: Hmmm.. how big is the image it's try to decrypt?
[10:04] <ttx> big.
[10:05] <Daviey> >2GB big?
[10:05] <ttx> hmm
[10:05] <Daviey> http://www.bouncycastle.org/jira/browse/BJA-276 could perhaps be related
[10:06] <Daviey> Although.. i fail to see why lucid kernel works.. if that is the issue.
[10:06] <Daviey> i would expect it to be a JVM version issue.
[10:06] <ttx> shouldn't be > 2Gb
[10:06] <ttx> the image before encryption is 1.4 Gb uncompressed
[10:06]  * ttx tries to find the size
[10:08] <ttx> hmm, looks like a red herring too
[10:08] <Daviey> yeah
[10:08] <ttx> not PGP, and I don't see why 2.6.34 would factor in
[10:08] <Daviey> Yeah, i noticed it wasn't PGP - but i would imagine it shares the same functions.
[10:09] <ttx> can be tseted by registering a verysmall image
[10:09] <Daviey> ttylinux
[10:09] <ttx> yep
[10:11] <Daviey> ttx: When you predouced, did you test the Alpha1 kernel - or the current maverick kernel?
[10:11] <Daviey> reproduced*
[10:11] <ttx> alpha1
[10:11] <ttx> I can upgrade that.
[10:11] <ttx> Daviey: looking at the logs, it decrypts the kernel alright
[10:11] <ttx> but not the image
[10:11] <Daviey> ttx: Okay, whilst you have the enviroment of A1, can you upgrade the kernel and try again?
[10:12] <ttx> So it's just not "decryption fails", it's "some decryption fails"
[10:12] <Daviey> ttx: Oh aye.. which made me wonder if it was a file size issue
[10:12] <ttx> yep
[10:12] <Daviey> ttx: Can you apt upgrade to maverick kernel?
[10:12] <ttx> will try upgrading first, then a ttylinux image
[10:12] <ttx> you know where I can find one ?
[10:13] <Daviey> cool.. That should answer if this is fixed in the latest kernel for the ubuntu-kernel team
[10:13] <Daviey> (/me suspects it'll still be there)
[10:15]  * Daviey pokes his lab
[10:25]  * ttx retries regsietring/running on latest kernel
[10:28] <ttx> same fail. Registering is sufficient to trigger the decryption error... no need to try to run it
[10:30] <nimrod10> is there a documentation where I can read about how ubuntu-server 10.04 handles pam and passwords for users ?
[10:30] <nimrod10> I found out I can't create users with empty passwords in 9.10 and 10.04 server
[10:31] <nimrod10> This was so that the users would login via ssh key and then set their own password
[10:34] <ttx> Daviey: where can I find a ttylinux image ready to register ?
[10:41] <Daviey> ttx: Hmm, is it not on the UEC web store?
[10:41] <ttx> I don't think so... looking again
[10:43] <Daviey> wget http://eucalyptus.cs.ucsb.edu/downloads/euca-ttylinux.tgz ?
[10:43] <ttx> will try that
[10:43] <Daviey> scrub that
[10:43] <Daviey> 40
[10:43] <Daviey> 404
[10:44] <remix_tj> no one got problems while installing ubuntu 10.04 server on IBM eServer 225? I get kernel panic while detecting hardware...
[10:44] <ttx> fwiw, registering an emi with an artificial .img the same size as the kernel is working
[10:44] <ttx> Daviey: ^
[10:44] <Daviey> interesting
[10:44] <ttx> So it's not a question of different codepaths between kernel and image
[10:44] <ttx> it's either size or luck
[10:45] <Daviey> remix_tj: I would recommend registering a bug against the package "linux"
[10:45]  * Daviey bets on size
[10:47]  * ttx will try with the smallest image in store, a karmic i386
[10:49] <Daviey> ttx: IIRC the ttylinux is shipped in the euca deps, where the WSDL stuff is (so unpackaged)
[10:51] <ttx> I can test the size issue by randomly making .img files.
[10:52] <ttx> they don't really need to be runnable
[10:52] <Daviey> ttx: ah, using dd?
[10:52] <ttx> yep
[10:53] <Daviey> good thinking that man
[10:54] <ttx> So far 4Mb -> ok, 1.5Gb -> fail
[10:56] <Daviey> interesting
[10:56] <ttx> the karmic image should test 1.3 Gb
[10:56] <ttx> i'll try 200Mb
[11:00] <ttx> 200 Mb -> fail
[11:01] <ttx> karmic (1.3Gb -> fail
[11:01] <Daviey> 100MB?
[11:01] <ttx> i'm on it :)
[11:01] <Daviey> sorry
[11:01]  * ttx dichotomizes
[11:02] <Daviey> That is the first use of dichotomise this week!
[11:04] <remix_tj> Daviey: but i can't read all the kernel panic... now i try with alternate, else i'll file a bug
[11:04] <ttx> Daviey: 75Mb -> pass
[11:06] <ttx> 140Mb -> pass
[11:07] <ttx> beh
[11:08] <ttx> 205Mb -> pass
[11:08]  * ttx retests 200Mb
[11:10] <ttx> arh, seems like its not just size
[11:16] <ttx> 1Gb -> fail
[11:17] <Daviey> remix_tj: This clearly sounds like a bug, so even if you can;'t provide enought detail at the moment - raising a bug in regards to the hardware make/model would be useful
[11:17] <Daviey> ttx: Bah.. was hoping it was a clear cut size thing
[11:17] <ttx> I'm testing by trimming images now
[11:18] <ttx> the one that fails is a blantant cp X.tar.gz X.img
[11:19] <remix_tj> ok Daviey
[11:20] <ttx> 500 Mb -> fail
[11:20] <Daviey> ttx: There doesn't seem to be any obvious consistency, unless it's related to free cache space at the time.
[11:21]  * ttx tries to refine nevertheless
[11:22] <ttx> 270Mb -> pass
[11:24] <ttx> 390 Mb -> pass
[11:26] <ttx> 490 Mb -> fail
[11:28] <ttx> 451Mb -> fail
[11:29] <ttx> 410Mb -> pass
[11:30] <ttx> I think it's accumulative, the more you add size the more likely you're to fail
[11:30] <ttx> but not failing on a precise size whatever the contents.
[11:30] <ttx> 430Mb -> pass
[11:32] <ttx> I've one file that fails at 200 Mb, and another that fails around 440 Mb.
[11:32] <ttx> I think it's just a question of when you hit that "bad block padding"
[11:32]  * ttx -> lunch
[11:33] <Daviey> yeah
[11:54] <ruben23>  hi guys anyone have tried implemeting thin client..? in a network environment..?
[11:54] <ruben23> using thin client server-ubuntu-server
[12:05] <SpamapS> oi... 22G of mirror downloaded.. 8G to go...
[12:06] <jpds> SpamapS: You're using mirror.anl.gov as a source?
[12:07] <BuZZ-dEE> hello, which package can i install to provide an openid server?
[12:20] <remix_tj> Daviey: bug 595040
[12:25] <SpamapS> jpds: no, uoregon.edu..  why?
[12:27] <acajou> cwillu_at_work: Nope. Also tried #btrfs. https://bugs.launchpad.net/ubuntu/+source/btrfs-tools/+bug/594910
[12:30] <acajou> You'd think this package is especially well-suited to Ubuntu Server... ZFS-like bit-rot proof filesystem. I'm surprised that it seems no one else has tried it on Ubuntu Server.
[12:32] <acajou> If I don't solve it soon, I'll have to use a mirrored ext3/4 partition instead...
[12:37] <Daviey> remix_tj: That is a good bug, thanks
[12:37] <Daviey> remix_tj: I wouldn't expect it to be fixed for the next 2 weeks, at least.
[12:38] <SpamapS> acajou: btrfs is experimental
[12:38] <SpamapS> acajou: but it works fine on lucid
[12:38] <SpamapS> acajou: give it another year, btrfs will be the default FS on Ubuntu. :)
[12:38] <Daviey> ttx: Have you returned?
[12:39]  * SpamapS should not be awake. :-P
[12:39] <ttx> Daviey: yes
[12:39] <Daviey> SpamapS: a true geek, never sleeps :)
[12:39] <Daviey> ttx: Not having much joy with maverick atm
[12:39] <Daviey> ttx: I tried pinning back librampart, but that isn't it
[12:39] <Daviey> Most of the other deps seems to be the same as lucid.
[12:39] <ttx> Daviey: start with alpha1 and upgrade one at a time ?
[12:40] <Daviey> ttx: Times like this debian style snapshot archive would rock.
[12:40] <Daviey> ttx: I'm seeing something you have experienced before, axix2c.log [error] http_transport_utils.c(2557) Service or operation not found
[12:40] <tydeas> I am having some issues with the bind9 dns server
[12:42] <Daviey> !ask | tydeas
[12:42] <tydeas> sorry.
[12:42] <cwillu_at_work> acajou, try building from git, it's straightforward and recommended by upstream at this point anyway
[12:42] <cwillu_at_work> btrfs-tools at least
[12:42] <tydeas> I have set the ubuntuserver machine will call it "7" from know on the bind9 dns server
[12:43] <cwillu_at_work> I'm not certain btrfs-tools is responsible for creating entries in /dev though; at the very least, /dev/btrfs-control doesn't show up in the listed files on my 64bit server
[12:43] <tydeas> i set him as my primary dns server in an other machine but can resolve names
[12:43] <tydeas> firewall is fine
[12:43] <LukAtMeee> Please help.  Setting up a Linux Xubuntu based server on windows network.  Have installed Samba and configured as per the "Ubuntu 10.04 Help Server Windows Networking" page, and it says "From a windows client you should be able to browse the Ubuntu file server and see the shared directory"..  Well I can't.  I can ping the Linux box from a windows box on the network, but can't ping any windows box from the Linux box.
[12:43] <LukAtMeee> I can see the windows boxes from each other.
[12:44] <remix_tj> Daviey: no, problem, i try to install hardy, for the moment
[12:44] <cwillu_at_work> acajou, yep.  "modprobe btrfs" is what creates /dev/btrfs-control, not btrfs-tools (just tested on my desktop)
[12:44] <remix_tj> or karmic
[12:45] <Daviey> tydeas: Hmm.. have you tried the tool "dig"?
[12:47] <Daviey> ttx: Do you remember what fixed the axis2 issue?

[12:47] <LukAtMeee> Anyone, any ideas.  I've tried uninstalling and reinstalling samba to get a fresh start, and the Xubuntu is a brand new install from scratch (10.04).  I have read through the help guide, but when I can't even get it to do the most basic thing, I think there must be a problem, hence why I'm here!!
[12:48] <ttx> Daviey: doesn't ring a bell... why do you say I've hit this one before ? (I might have, I hit so many...)
[12:48] <Daviey> heh, one mo
[12:48] <Daviey> ttx: https://lists.ubuntu.com/archives/ubuntu-devel/2009-September/029089.html
[12:49] <tydeas> Daviey: brb
[12:49] <Daviey> ttx: it looks like the issue may be the same/similar
[12:49] <ttx> hm
[12:49] <ttx> that one was due to upstartification missing a few steps... lemme check the commits
[12:50] <LukAtMeee> Also, if I run smbclient -L it tells me that the linux box is the only box on the workgroup, which it blatantly isn't.  So where is the problem??!
[12:50] <Daviey> ttx: it can't be apache/libapache2-mod-axis2c related as they are the same version as lucid
[12:50] <acajou> cwillu_at_work: What's surprising is that /dev/btrfs-control was created without my intervention on my desktop...
[12:51] <cwillu_at_work> acajou, does btrfs show up in lsmod on your desktop?
[12:51] <Daviey> LukAtMeee: If you leave it a few hours, zul might be the best person to ask about that :)
[12:51] <cwillu_at_work> acajou, it disappears if you rmmod btrfs, and comes back when you modprobe btrfs, seems pretty conclusive to me :p
[12:52] <cwillu_at_work> LukAtMeee, if you "restart nmbd; restart smbd", does it magically start working?
[12:52] <cwillu_at_work> I think the samba upstart jobs in lucid are slightly broken if you have more than one network interface
[12:55] <LukAtMeee> cwillu_at_work: nope.  Now I can't see the server in the workgroup either!
[12:55] <LukAtMeee> Daviey: okay, thanks.
[12:56] <cwillu_at_work> oh, sorry.  You can connect from windows to the server, but you can't connect from the server back to any windows machines?
[12:57] <LukAtMeee> So now I can't see the windows machines from the Linux box, and the Windows boxes can't see the Linux box and the Linux box can't see the Linux box when I run smbclient -L servername
[12:57] <halvors> What is the best administrationpanel for my serve?
[12:57] <halvors> server*
[12:58] <cwillu_at_work> LukAtMeee, can the windows machines current ping the server (even though the server doesn't show in the workground)?
[12:59] <LukAtMeee> yes
[12:59] <cwillu_at_work> okay, the workgroup should fix itself then in a few minutes
[12:59] <LukAtMeee> and they do it by IP address or by servername
[13:00] <cwillu_at_work> this should definitely work, but what does "net lookup <windows host name>" do?
[13:01] <LukAtMeee> gives me the IP assigned by my router for each of the windows machines....I feel like there's a light at the end of the tunnel, finally (I;ve been trying for 3 days!)
[13:01] <ttx> Daviey: ISTR the issue was some unset AXIS2C env variables in the upstart script
[13:02] <Daviey> ttx: hmm.. don't think that is this issye
[13:02] <cwillu_at_work> LukAtMeee, I'm not 100% certain of this, as my desktop doesn't seem to need it (but that could be for other reasons);  but try adding "wins" to the hosts line in /etc/nsswitch.conf (after files)
[13:02] <Daviey> issue*
[13:02] <ttx> Daviey: check /var/run/eucalyptus/httpd-cc.conf
[13:03] <Daviey> ttx: I've increaced verbosity of axis2 and apache, and here are all the logs
[13:03] <Daviey> http://pastebin.daviey.com/yVVp/
[13:03] <Daviey> ttx: looks like axis is doing it's thing, but for a non-obvious reason apache is being killed
[13:03] <cwillu_at_work> LukAtMeee, you may also need to install "winbind", but try it without first.  Should take effect immediately after modifying the file
[13:03] <LukAtMeee> I don't have a line like that in my smb.conf  What should it say?
[13:04] <cwillu_at_work> hosts:          files wins mdns4_minimal [NOTFOUND=return] dns mdns4 is mine
[13:04] <cwillu_at_work> it's not smb.conf though
[13:04] <cwillu_at_work> /etc/nsswitch.conf
[13:04] <cwillu_at_work> (i.e., it's not a samba setting)
[13:05] <LukAtMeee> ok....brb
[13:05] <cwillu_at_work> no reboot necessary
[13:06] <ttx> Daviey: confirmed -- my error was solved by appropriately setting Set EUCALYPTUS and AXIS2C_HOME environment variables in eucalyptus-cc upstart job
[13:06] <LukAtMeee> okay, do I need to do a service restart or anything?
[13:06] <ttx> so probably not the same root error
[13:06] <Daviey> nah
[13:07] <Daviey> *sigh*
[13:07] <ttx> Daviey: but I would really try to fix one issue at a time
[13:08] <ttx> Daviey: and concentrate on the already-identified alpha-1 critical issue
[13:08] <_tydeas_> hi back on the dns server issue.
[13:09] <_tydeas_> Do i have to allow icmp also from firewall
[13:09] <cwillu_at_work> LukAtMeee, no, it takes effect immediately
[13:09] <Daviey> ttx: Hmm.. when i spoke with Jos.. we seemed to think euca not starting at all was more pressing
[13:09] <_tydeas_> i run the wireshark on server with dns filter
[13:09] <LukAtMeee> okay, done the above, but windows still can't "see" linux, nor the other way around....any other ideas???
[13:10] <ttx> Daviey: then you should start from A1, gradually upgrade and check which upgrade breaks it
[13:10] <Daviey> ttx: ok
[13:11] <cwillu_at_work> LukAtMeee, ignore the windows not seeing linux for now
[13:12] <cwillu_at_work> LukAtMeee, as I said, that should go away within 30 minutes or so
[13:12] <cwillu_at_work> LukAtMeee, with that change, can you "ping <windows host>"?
[13:13] <LukAtMeee> sadly not....
[13:13] <cwillu_at_work> okay, install winbind and try again
[13:14] <LukAtMeee> okay, tried it, but no joy.  should I reboot?
[13:14] <cwillu_at_work> can you pastebin your /etc/nsswitch.conf?
[13:15] <LukAtMeee> # /etc/nsswitch.conf
[13:15] <LukAtMeee> #
[13:15] <LukAtMeee> # Example configuration of GNU Name Service Switch functionality.
[13:15] <LukAtMeee> # If you have the `glibc-doc-reference' and `info' packages installed, try:
[13:15] <LukAtMeee> # `info libc "Name Service Switch"' for information about this file.
[13:15] <LukAtMeee> passwd:         compat
[13:15] <LukAtMeee> group:          compat
[13:15] <cwillu_at_work> ugh
[13:15] <LukAtMeee> shadow:         compat
[13:15] <LukAtMeee> hosts:          files wins mdns4_minimal [NOTFOUND=return] dns mdns4
[13:15] <cwillu_at_work> use  a pastebin
[13:15] <LukAtMeee> networks:       files
[13:15] <LukAtMeee> protocols:      db files
[13:15] <LukAtMeee> services:       db files
[13:15] <LukAtMeee> ethers:         db files
[13:15] <LukAtMeee> rpc:            db files
[13:15] <LukAtMeee> netgroup:       nis
[13:15] <LukAtMeee> ah, how do I do that?
[13:15] <cwillu_at_work> LukAtMeee, never paste more than a couple lines directly into the channel
[13:15] <LukAtMeee> sorry
[13:15] <cwillu_at_work> http://pastebin.com
[13:15] <cwillu_at_work> that said, the file looks right :p
[13:18] <LukAtMeee> okay....any more ideas?
[13:19] <LukAtMeee> you're being very helpful, thanks!
[13:19] <LukAtMeee> brb
[13:20] <_tydeas_> The problem with bind is that i can dig from the dnsserver machine.
[13:20] <_tydeas_> But if i set this server machine as dns server to a local area machine it can not get the addresses.
[13:21] <_tydeas_> i am trying to dig from local machine and nothing
[13:21] <_tydeas_> the firewall of the server allows 53
[13:21] <_tydeas_> any ideas?
[13:21] <_tydeas_> how can i debug this issue?
[13:21] <cwillu_at_work> apt-get install pastebinit; strace -o /tmp/ping.trace ping windowshostname; pastebinit /tmp/ping.trace
[13:21] <cwillu_at_work> _tydeas_, you're allowing udp and tcp?
[13:22] <cwillu_at_work> LukAtMeee, that apt-get line is for ou
[13:22] <cwillu_at_work> you
[13:23] <LukAtMeee> Thx, it's running now.
[13:23] <_tydeas_> i used wireshark and check that ICMP protocol with source local and dest server is port unreachable.
[13:23] <cwillu_at_work> LukAtMeee, when it's done, it'll spit out a link;  paste that link here
[13:23] <acajou> (sorry - power failure)
[13:23] <cwillu_at_work> _tydeas_, here's a question, if you turn off the firewall, does it work? :)
[13:25] <LukAtMeee> http://pastebin.com/38SeBZrD
[13:26] <_tydeas_> root@sindos2:~# ufw status
[13:26] <_tydeas_> Status: inactive
[13:26] <LukAtMeee> okay, I hope you can unserstand that, because most of that looks....interestingly complex(?) to me!
[13:26] <cwillu_at_work> LukAtMeee, it's a list of all the system calls made by "ping"
[13:26] <_tydeas_> cwillu_at_work: still does not work
[13:26] <cwillu_at_work> LukAtMeee, you can see which files it opens, etc
[13:26] <cwillu_at_work> _tydeas_, okay, not a firewall problem then :/
[13:27] <LukAtMeee> okay, thx, now it makes more sense!
[13:27] <LukAtMeee> never realised ping was so complex...
[13:27] <cwillu_at_work> it's not :p
[13:27] <cwillu_at_work> but it calls out to libraries which are somewhat
[13:27] <cwillu_at_work> figuring out hostnames is serious business, just ask _tydeas_
[13:28] <_tydeas_> :)
[13:28] <_tydeas_> any other idea?
[13:28] <cwillu_at_work> What interfaces is the dns daemon listening on?
[13:29] <acajou> cwillu_at_work: You're right modprobe btrfs creates /dev/btrfs-control.
[13:29] <cwillu_at_work> acajou, I usually am :p
[13:29] <sommer> morning
[13:30] <cwillu_at_work> acajou, you may not have noticed it on your desktop if you made and mounted a btrfs filesystem before you started experimenting with the tools
[13:30] <hggdh> kirkland: morning, and a Q about eucalyptus-network.conf and $IFACE
[13:32] <cwillu_at_work> _tydeas_, check "lsof -iTCP -iUDP"
[13:32] <cwillu_at_work> _tydeas_, make sure it's not listening on only localhost or something silly like that :)
[13:32] <acajou> cwillu_at_work: first thing I did was convert an ext3 filesystem... (btrfs-convert, then mount -t btrfs)
[13:33] <cwillu_at_work> acajou, okay; that would have resulted in btrfs being modprobed
[13:33] <acajou> (that would have auto-created /dev/btrfs-control?)
[13:33] <cwillu_at_work> acajou, yes, modprobing the driver is what triggers the creation of btrfs-control
[13:33] <acajou> Hmmmm. Seems to me modprobing should be part of the package install?
[13:33] <cwillu_at_work> acajou, no
[13:33] <cwillu_at_work> acajou, this has nothing at all to do with btrfs-tools :p
[13:34] <cwillu_at_work> the tools which require btrfs-control require a mounted filesystem anyway
[13:34] <acajou> Well, if you start by mkfs.btrfs, thinhgs fail.
[13:34] <cwillu_at_work> eh?
[13:34] <cwillu_at_work> mkfs.btrfs should work regardless, and then when you mount it, everything works, no?
[13:35] <acajou> On the server, instead of starting by converting an ext3 FS I tried to mkfs, and that failed.
[13:35] <acajou> With the missing /dev/btrfs-control error.
[13:35] <cwillu_at_work> let me test, one sec
[13:36] <cwillu_at_work> acajou, works here with btrfs not loaded
[13:37] <cwillu_at_work> mkfs -t btrfs ./btrfs-test
[13:37] <cwillu_at_work> fs created label (null) on ./btrfs-test
[13:37] <cwillu_at_work> 	nodesize 4096 leafsize 4096 sectorsize 4096 size 256.00MB
[13:38] <acajou> I believe I used: mkfs.btrfs -m raid1 -d raid1 /dev/sdc /dev/sdd
[13:38] <cwillu_at_work> k, let me try that
[13:38] <zul> LukAtMeee: are you using opendns?
[13:39] <LukAtMeee> @zul erm, I don't know.  How do I find out?
[13:39] <acajou> I don't know how to undo the modprobe, so I can' t come back to my previous state...
[13:39] <cwillu_at_work> acajou, "rmmod btrfs"
[13:40] <zul> LukAtMeee: check your /etc/resolv.conf
[13:40] <LukAtMeee> and how do I highlight that a message is aimed at a particular person on the irc?(new to this too!)
[13:40] <acajou> cwillu Thanks. :)
[13:40] <cwillu_at_work> LukAtMeee, include their name on the line (use the tab key to complete nicks)
[13:41] <LukAtMeee> zul: no, looks like I'm using the virgin dns
[13:41] <LukAtMeee> cwillu_at_work, thanks\1
[13:41] <cwillu_at_work> acajou, looks like it's specific to raids
[13:42] <acajou> cwillu_at_work May not be fatal, it just says " failed to open /dev/btrfs-control skipping device registration"
[13:42] <cwillu_at_work> acajou, it doesn't work though, tested
[13:42] <cwillu_at_work> acajou, you can create a btrfs fs with a single device as raid1 (which is actually "dup" in that case), and then add the second device after mounting it
[13:43] <cwillu_at_work> but that won't give you the same result I don't believe
[13:43] <cwillu_at_work> there's raid1, and then there's dup.
[13:46] <acajou> Hmmmm. If I modprobe first, it seems to work, but tells me the FS created is 2.73TB, and it should not be that large for two 1.5tb drives in raid1....
[13:46] <cwillu_at_work> yep
[13:46] <cwillu_at_work> acajou, no, that's correct
[13:46] <cwillu_at_work> you have to remember that "used space" is a complicated notion in a filesystem as flexible as btrfs or zfs
[13:47] <cwillu_at_work> acajou, you have a 2.73tb pool which can be allocated to particular trees
[13:47] <cwillu_at_work> i.e., the metadata tree, or the data tree, or the system tree
[13:47] <acajou> Hmmm. First brushes with raid...
[13:47] <acajou> OK
[13:48] <cwillu_at_work> normal "df", "btrfs filesystem df /mount/point", and "btrfs-show" will all give different views on that
[13:49] <cwillu_at_work> "btrfs filesystem show /dev/device" instead of btrfs-show, actually
[13:50] <cwillu_at_work> _tydeas_, can you pastebin the actual output from ping?
[13:51] <cwillu_at_work> _tydeas_, it looks like it's resolving fine from the strace dump, and it's just the actual ping packets that aren't getting through
[13:51] <blsmith> Anyone here?
[13:52] <cwillu_at_work> no.
[13:52] <ttx> jasonb: please see bug 591802 and bug 594989 and comment on them if the proposed solution looks good to you
[13:52] <cwillu_at_work> blsmith, pro-tip:  don't ask to ask, just ask.
[13:53] <blsmith> I'm setting up a 10.04 server and need setting up a usb wireless adaptor.
[13:54] <cwillu_at_work> blsmith, if you set up the wireless under a desktop session, hit the "Available to all users" option in the connection editor
[13:54] <LukAtMeee> cwillu_at_work, so did that ping.trace help at all?
[13:54] <cwillu_at_work> blsmith, then it should automatically connect even with nobody logged in
[13:55] <cwillu_at_work> LukAtMeee, sorry, I sent that to the wrong person
[13:55] <cwillu_at_work> LukAtMeee, (not _tydeas_), it looks like it's resolving fine from the strace dump, and it's just the actual ping packets that aren't getting through
[13:55] <cwillu_at_work> LukAtMeee, can you pastebin the actual output from ping?
[13:55] <blsmith> cwillu, I'm under server, should I install a desktop version?
[13:56] <LukAtMeee> if I ping, it just sits there and does nothing until I Ctrl-C
[13:57] <cwillu_at_work> LukAtMeee, what does "host <hostname>" say?
[13:58] <LukAtMeee> portable.cable.virginmedia.net has address 192.168.1.118
[13:58] <acajou> cwillu_at_work: Ya, df -> 2.93TB. Gonna put 60 odd gb to see what that does. So I guess the trick is just to modprobe first...
[13:58] <cwillu_at_work> yep
[13:58] <cwillu_at_work> LukAtMeee, okay, it's working fine then
[13:59] <cwillu_at_work> LukAtMeee, you probably just have a firewall on the windows machines that's blocking ping and so forth
[13:59] <LukAtMeee> but how do I get them to see the machines...remove the firewall?
[13:59] <cwillu_at_work> LukAtMeee, basically;  do you really need to be able to see them from the server though?
[14:00] <blsmith> cwillu_at_work, Should I install ubuntu desktop in order to set up the newtork then, disable the GUI?
[14:01] <cwillu_at_work> bladernr, a moment
[14:03] <cwillu_at_work> blsmith, I'm not sure if networkmanager is used in a bare server install;  it'd be straightforward to install that though if it's not.  That still doesn't help you configure wireless though.
[14:03] <cwillu_at_work> blsmith, is this just a home server?
[14:03] <cwillu_at_work> if so, then yes, I'd say just install ubuntu-desktop to configure it, simpler that way
[14:04] <YouCanHelpMe> damn.  don't know what happened, but I am LukAtMeee.   cwillu_at_work: not really, that's a fair point, but I do need to be able to see the server from the windows machines.
[14:04] <cwillu_at_work> I know you can configure it via dbus calls (dbus-send from a command line), but I don't have the commands on hand
[14:04] <cwillu_at_work> YouCanHelpMe, okay, you'll have to figure out the firewall rules on those machines then
[14:05] <YouCanHelpMe> okay.  I'll look into it.  I have to go now, but THANKYOU so much for helping out.
[14:05] <cwillu_at_work> one option might be to set up those machines to use the samba wins server (which you'd enable from /etc/samba/smb.conf)
[14:05] <remix_tj> Daviey: problems affecting also karmik, maybe due to raid controller
[14:05] <YouCanHelpMe> okay... cwillu_at_work, how do I do that?
[14:06] <cwillu_at_work> YouCanHelpMe, you're gonna have to dive into the documentation a little ;p
[14:07] <YouCanHelpMe> no worries.  Will do.  Thanks again!
[14:07] <cwillu_at_work> my consulting rate is 105$/h :p
[14:07] <YouCanHelpMe> ouch!
[14:10] <blsmith> cwilu_at_work, thanks for the help.
[14:10] <cwillu_at_work> blsmith, that works for you?
[14:10] <cwillu_at_work> blsmith, I'm sure there's some instructions somewhere to configure it by hand, it's just that most servers aren't connected over wireless
[14:12] <blsmith> cwilu_at_work, yeah, i have a copy of 9.10 desktop, I'm going to set up the wireless network and then upgrade that to 10.04.  I understand and thats why i'm not even going to spend anymore time looking for instructions.
[14:13] <bogeyd6> wireless on a server isnt so bad if the server has a definite need to be
[14:13] <YouCanHelpMe> cwillu_at_work, I don't believe it.  The windows machines and Linux server just started talking to each other without me doing anything else!  So it must have been what you suggested!   Thanks again!
[14:13] <cwillu_at_work> ;)
[14:13] <ruben23> any problem on the ubuntu-server lucid repo, im installing packages now but its really so slow..downloading..
[14:13] <bogeyd6> cnc program relay servers work great over wireless in a factory setting
[14:13] <blsmith> bogeyd6, can you help me set this up?  I think its being recongnized but its just not turning on.
[14:13] <ruben23> any issue on the repo site
[14:13] <cwillu_at_work> bogeyd6, that's why I said "usually aren't" and not "it's stupid to think about it"
[14:14] <bogeyd6> cwillu_at_work, thank you for the unnecessary explanation
[14:14] <bogeyd6> ruben23, you in the midwest US?
[14:14] <cwillu_at_work> bogeyd6, and thanks for the unnecessary correction :)
[14:14]  * bogeyd6 's neckbeard radar as went full alert
[14:15] <ruben23> bogeyd6: Asia
[14:15] <bogeyd6> ruben23, if you are in the midwest area of the US you need to switch all of your repos to anl.gov
[14:16] <bogeyd6> blsmith, i can try but i wasnt here fr the original problem
[14:16] <cwillu_at_work> bogeyd6, wants to configure wireless from the console
[14:16] <blsmith> bogeyd6, basically i'm trying to set up a 10.04 server with a dlink dwa 150 wireless adaptor. Did a google search nothing came up  so I came here.
[14:18] <bogeyd6> blsmith, are you using it USB style?
[14:18] <blsmith> bogeyd6, yes it is usb, I'm not sure what other style it could be though?
[14:19] <bogeyd6> blsmith, https://help.ubuntu.com/community/HardwareSupportComponentsWirelessNetworkCardsDlink   << suggest your card is not supported
[14:20] <blsmith> bogeyd6: dmesg output after unplug and replug.  Registered led device: rt2800usb-phy1: :radio
[14:22] <bogeyd6> oh thats ok blsmith, linux does lots of dumb stuff like that. ask anyone using 9.10 how they felt about everything the broadcom chipset looking like everything was kosher and would connect to the AP but no traffic could be transmitted/received
[14:23] <bogeyd6> bcm43*, it was a peach, thankfully it works in 10.04
[14:23] <blsmith> bogey6: ouch.  so i'm actually getting no where with this device.
[14:23] <bogeyd6> blsmith, your comments and the documentation do in fact suggest that
[14:24] <bogeyd6> however if you want to continue might i suggest this
[14:24] <bogeyd6> !wireless | blsmith
[14:26] <blsmith> bogeyd6, i don't understand, what would you like me to try?
[14:26] <remix_tj> is there a way to disable hw autodetection?
[14:27] <bogeyd6> blsmith, visit that website and follow the step by step instructions and when you reach the end if it works, great, if not then move on
[14:27] <blsmith> thanks
[14:27] <bogeyd6> https://help.ubuntu.com/community/WifiDocs/WiFiHowTo
[14:27] <bogeyd6> its for the cli
[14:31] <blsmith> bogeyd6: this didn't help. sorry, but thanks for the support. moving on to attempt desktop setup.
[14:31] <elb0w> why do I see /usr/bin/dpkg --status-fd 31 --unpack --auto-deconfigure /var/cache/apt/archives/mysql-server-5.1_5.1.41-3ubuntu12.3_amd64.deb
[14:31] <elb0w> is that an auto-update?
[14:32] <bogeyd6> elb0w, only if you have the updates set to install ALL and not just security
[14:32] <acajou> cwillu_at_work: Thank you! Things seems to be working peachy fine now.
[14:32] <elb0w> bogeyd6, how can I check
[14:33] <cwillu_at_work> acajou, marked the bug as invalid
[14:33] <elb0w> root     10995  0.0  0.0  50868   780 ?        S    Jun10   0:00 /usr/bin/perl -w /usr/share/debconf/frontend /var/lib/dpkg/info/mysql-server-5.1.prerm upgrade 5.1.41-3ubuntu12.3
[14:33] <elb0w> root     10997  0.0  0.0   9236   928 ?        S    Jun10   0:00 /bin/bash -e /var/lib/dpkg/info/mysql-server-5.1.prerm upgrade 5.1.41-3ubuntu12.3
[14:33] <elb0w> I did not ask for these
[14:34] <cwillu_at_work> elb0w, ps axjf | less
[14:34] <cwillu_at_work> elb0w, you can use that to figure out what process started it
[14:35] <elb0w> which row is the starting process?
[14:36] <elb0w> PPID is parent process I guess
[14:36] <acajou> cwillu_at_work: 10-4.
[14:36] <bogeyd6> elb0w, sudo cat /etc/apt/apt.conf.d/50unattended-upgrades
[14:36] <smoser> kees, around ?
[14:37] <cwillu_at_work> elb0w, that command would bring up a tree list, just follow it back
[14:37] <elb0w> I traced it back to 1
[14:37] <elb0w> sbin/init
[14:37] <elb0w> Unattended-Upgrade::Allowed-Origins {
[14:37] <elb0w> 	"Ubuntu lucid-security";
[14:37] <elb0w> thats all I have unchecked
[14:37] <elb0w> er commented
[14:38] <cwillu_at_work> okay, probably a cron job as bogeyd6 suggests
[14:38] <elb0w> im checking what bogey suggested. The only thing I have enabled is security
[14:39] <bogeyd6> ok then it must be a security update, but i distinctly remember have to run apt-get upgrade myself to get that particular update because it was causing utf-8 tables to be manually repaired
[14:43] <ElTonerino> I upgraded dovecot today, and am now having problems when trying to send emails.
[14:44] <ElTonerino> I get this error "warning: SASL: Connect to private/dovecot-auth failed: Connection refused"
[14:44] <ElTonerino> any ideas?
[14:45] <bogeyd6> Anyone know how to tell gnome terminal to stop capturing my fkeys?
[14:45] <bogeyd6> ElTonerino, seems like the service may not be running or maybe the ufw was enabled
[14:48] <ElTonerino> dovecot is running, I'll check ufw
[14:51] <ElTonerino> ufw seems the same today as it was yesterday: http://pastebin.com/xUSvz8RB
[14:52] <guhcampos> I need to create a local mirror. Lucid Server only, should I use debmirror or apt-mirror?
[14:55] <ElTonerino> netstat -ln: http://pastebin.com/rZHfjyN2
[14:56] <dassouki> sometimes I want to check my email when i'm working on the server. back in the day we used to use pico ... I'm wondering is there a google mail friendly terminal email client
[14:57] <bogeyd6> dassouki, mutt or pine
[14:58] <bogeyd6> https://help.ubuntu.com/community/mutt
[14:58] <bogeyd6> ElTonerino, i wish i could help but i use zimbra
[14:59] <ElTonerino> bogeyd6, ok.  I'm still stuck if anyone else wants to help. :)
[15:00] <bogeyd6> ElTonerino, using postfix?
[15:00] <bogeyd6> http://wiki.dovecot.org/HowTo/PostfixAndDovecotSASL
[15:01] <bogeyd6> during the upgrade a config file maybe screwed the pooch
[15:02] <ElTonerino> bogeyd6, Yeah. Thanks for the link, I'll go through it.  I think I might head to my backups and compare the configs to todays config.
[15:03] <bogeyd6> diff command will help big time
[15:04] <kirkland> hggdh: howdy, here now
[15:04] <kirkland> Ng: it has to be approved into proposed
[15:05] <kirkland> Ng: subscribe to the bug to monitor it's progress (i suspect you are already)
[15:07] <hggdh> kirkland: after applying the euca update yesterday, I noticed that euca (re)start would print out a message stating IFACE was an undefined variable
[15:08] <hggdh> kirkland: /etc/init/eucalyptus-network.conf uses it
[15:08] <Ng> kirkland: k, I'll have a look tonight
[15:18] <smoser> kirkland, you fixed the byobu-in-byobu behavior when jumping through systems and byobu-by-default, right ?
[15:18] <smoser> i thought you did, by checking TERM != SCREEN
[15:19] <jasonb> ttx: Alright.
[15:19] <bogeyd6> rotfl, the nagios3 documentation in the help guide doesnt include the install of nagios-cgi
[15:21] <ttx> jasonb: just emailed you about those
[15:21] <jasonb> ttx: Ahh.  Thanks!  I just got up over here in the Pacific time zone.
[15:23] <jasonb> ttx: "if Ubuntu's solution doesn't already implement this alternative to authbind" .. I meant Debian's solution (sorry).
[15:28] <ruben23> its really slow downlaoding on installing repo pagakaage for ubuntu-server lucid..any idea, is repo site having problems now
[15:29] <ruben23> my location si southeast asia
[15:30] <kirkland> smoser: i have for maverick
[15:30] <kirkland> smoser: i think that probably needs to be sru'd to lucid though
[15:30] <kirkland> smoser: ie, it was after rc-freeze
[15:31] <ElTonerino> bogeyd6, Yeah, there were changes in /etc/postfix/main.cf.  Fixing them resolved my problem.
[15:31] <ElTonerino> Pheww...
[15:31] <bogeyd6> thankfully
[15:32] <ruben23> taking 40 minuites to download a single package only..:'(
[15:45] <konam> hi
[15:46] <konam>  i installed a lamp server on my ubuntu desktop to learn php but it doesn't seem to be working, i even installed phpmyadmin and it works but i can't get my php scripts to work.
[15:46] <konam> can somebody tell me why
[15:46] <konam> or how to fix it
[15:51] <hallyn> Daviey: .
[15:51] <Daviey> hallyn: \o
[15:52] <hallyn> hey, i noticed your emails about virsh save failing.  how long have you been looking into that?
[15:52] <Daviey> konam: what happens if you open a php file in your browser?
[15:52] <Daviey> hallyn: When do those mails date to?
[15:52] <hallyn> hm, well i saw them yesterday, i think on qemu list
[15:53] <hallyn> i was just wondering whether  you got the impression anyone know the cause at this point
[15:53] <hallyn> it seems like some ppl think i twas the dd blocksize in virsh, but qemu ppl think not?
[15:53] <Daviey> ahh that
[15:53] <hallyn> right
[15:53] <konam> Daviey it opens a download file dialog....
[15:53] <hallyn> anyway i was hoping to ask jdstrand to reproduce with 0.8.whatever :)
[15:54] <Daviey> hallyn: It seems to be both, there is (was) a bug in libvirt and that has improved it somewhat
[15:54] <Daviey> but qemu think there is also a timing issue there
[15:54] <jdstrand> 0.8.1 is being a little grumpy, but I'll have something for people to play with
[15:54] <hallyn> because it still takes 56 seconds?
[15:54] <hallyn> Daviey: have you reproduced it yourself?
[15:54] <Daviey> konam: Hmm.. I think you might need to clear the cache and/or restart your browser
[15:54] <hallyn> jdstrand: rockin'
[15:54] <Daviey> hallyn: no
[15:55] <hallyn> ok
[15:55] <hallyn> (then i'll stop bugging you :)
[15:55] <Daviey> hallyn: well yes, i know it was really slow!
[15:55] <Daviey> but i never looked deeply into the cause
[15:55] <hallyn> Daviey: well the thing is i don't know if on my laptop it's just really slow, or completely hung :)
[15:55] <hallyn> i've never seen it finish
[15:55] <hallyn> and i don't see the savefile grow
[15:56] <Daviey> hallyn: iotop showing anything interesting?
[15:57] <hallyn> Daviey: i don't think so
[15:57] <hallyn> but i hadn't really looked
[15:58] <Daviey> :/
[16:00] <Daviey> ttx: I've found the cause of UEC not starting!
[16:00] <ttx> Daviey: \o/
[16:00] <ttx> what was it ?
[16:01] <Daviey> ttx: groovy!
[16:01]  * ttx cries
[16:01]  * Daviey wimpers
[16:01] <Daviey> ttx: Any debug pointers on that?
[16:01] <ttx> no, I'm completely groovy-ignorant
[16:02] <ttx> Daviey: we might have to have a groovy-1.6 package if Euca doesn't support 1.7
[16:02] <ttx> another one to push upstream.
[16:03] <konam> Daviey but i'm having this behavior since yesterday when i installed the lamp server (i installed phpmyadmin today though)
[16:03] <Daviey> konam: Yes, sometimes it's a client issue, rather than a server issue
[16:04] <Daviey> konam: If phpmyadmin is working, it sounds like php is working on the server
[16:04] <Daviey> konam: perhaps created a simple <?php echo "TEST"; ?> in /var/www/test.php
[16:04] <Daviey> then use "curl -v http://localhost/test.php" and see what is returned
[16:05] <remix_tj> Daviey: found, it was the motherboard's scsi controller
[16:05] <Daviey> ^^ from the server
[16:05] <Daviey> remix_tj: Oh, hardware failure?
[16:05] <remix_tj> no, it is not an hw failure. with windows works. maybe a driver bug
[16:07] <Daviey> remix_tj: Ahh, i see
[16:07] <Daviey> can you update the bug with info, and provide lspci -vvv info?
[16:11] <konam> Daviey you were right, i tried with another file with a .php extension and it worked
[16:11] <Daviey> konam: cool
[16:12] <remix_tj> Daviey: after completion of the setup i'll do
[16:12] <Daviey> remix_tj: thanks, really appreciate it
[16:12] <remix_tj> but i do not know if the os will boot with that hw enabled...
[16:18] <kpettit> can anybody recommend a good search tool for a desktop/server that doesn't suck up alot of resources?
[16:19] <kpettit> locate just isn't doing the trick and a few others I've tried make the desktop slow when they are archiving
[16:19] <bogeyd6> man i spent 30 mins preparing a bug just to have to mark it invalid because of my own stupidity
[16:20] <konam> Daviey a question, should i put eclipse's workspace on the root of localhost? it doesn't seem to be working otherwise, it says that it can't find my php file on the localhost..
[16:21] <Daviey> konam: I would say not.. but i've never used eclipse for php work
[16:21] <Daviey> konam: How are you accessing the files in the server through eclipse?
[16:22] <konam> i just created a project and when i hit execute it gives me a NOT FOUND error
[16:23] <konam> the server is set to localhost i think
[16:23] <Daviey> konam: Is this eclipse ON the ubuntu server, or ubuntu desktop accessing php files on the server?
[16:24] <konam> Daviey no, i installed the lamp server on my desktop so eclipse is on the server
[16:25] <Daviey> ok.. in that case; it is probably easier to have your workspace in the webroot
[16:25] <Daviey> konam: if you "sudo a2enmod userdir" , restart apache
[16:25] <konam> i tried but it's readonly
[16:25] <Daviey> mkdir ~/public_html , you can put everything in the public_html folder in your home folder
[16:26] <Daviey> then you can access it via http://localhost/~konan/
[16:26] <Daviey> (for example)
[16:26] <konam> Daviey uhm, i don't understand, what command should i use first?
[16:26] <Daviey> konam: "sudo a2enmod userdir"
[16:26] <konam> userdir would be?
[16:26] <Daviey> konam: as i typed
[16:27] <Daviey> (userdir is an apache module)
[16:29] <Daviey> konam: then, "mkdir ~/public_html"
[16:30] <Daviey> then anything you put in ~/public_html (example, /home/konam/public_html/) will show in http://localhost/~konam/
[16:30] <Daviey> and eclipse will have read/write access there
[16:37] <lau> hello, I am trying to migrate from hardy to lucid
[16:37] <lau> I do-release-upgrade
[16:38] <lau> now I get the initramfs prompt on the server
[16:38] <amstan> hey guys, i'm getting an error from a 3ware card, apparently i have a degraded unit
[16:38] <Daviey> ttx: Confirmed, latest maverick with lucid groovy works.. attempting to investigate further.
[16:38] <amstan> i'm just installing the os, so no problems there, but what does degraded mean? 1 dead drive? or are they not in sync yet?
[16:39] <lau> libudev: udev_monitor_new_from_netlink: error getting socket: Invalid argument
[16:39] <lau> Segmentation fault
[16:39] <ttx> Daviey: I'd suggest pushing a groovy 1.7.3really1.6to a PPA in order to have an easy workaround
[16:39] <lau> ALERT! /dev/disk/by-uuid/23.... does not exist. Droping to shell!
[16:39] <lau> any idea ?
[16:40] <Daviey> ttx: End game solution?!
[16:40] <ttx> Daviey: I mean, to easily unblock testers, while waiting for a solution
[16:41] <Daviey> ahh, that makes more sense :)
[16:45] <SpamapS> lau: that sounds pretty serious.. hmm
[16:46] <lau> SpamapS: I chrooted from a live cd to the system then (re) installed grub but same issue
[16:46] <Daviey> ttx: urg! groovy is a horrible package
[16:47] <lau> now I am trying to replace uuid param by /dev/sdx param in menu.lst
[16:48] <lau> any idea ?
[16:48] <SpamapS> lau: that may work yes
[16:48] <SpamapS> lau: I'm sorry I'm sort of a novice when it comes to upgrades.. :-/
[16:49] <SpamapS> In my world you do upgrades by buying new servers and putting the new version on them. ;)
[16:49] <ttx> Daviey: agreed.
[16:50] <SpamapS> lau: http://ubuntuforums.org/showthread.php?t=1127779   did that help at all?
[16:53] <jdstrand> hallyn: hey, so are you planning another qemu-kvm upload any time soon?
[16:56] <Daviey> ttx: Uploading groovy_1.7.3really1.6-0ubuntu1 now
[16:57] <ttx> Daviey: I meant, in your PPA
[16:57] <ttx> Daviey: not necessarily to the archive.
[16:58] <ttx> Daviey: I think it makes sense to the archive as an endgame solution
 Daviey: I'd suggest pushing a groovy 1.7.3really1.6to a PPA in order to have an easy workaround
[17:04] <konam> Daviey and isn't there a way to especify the directory you want apache to access?
[17:07] <hallyn> jdstrand: when the next kvm version is released, yeah
[17:07] <hallyn> jdstrand: you mean for maverick, and newer than 0.12.4 right?
[17:07] <jdstrand> hallyn: yes for maverick. ok. I have a patch I need in qemu for libvirt 0.8.1 network device attach to work
[17:07] <Daviey> hallyn: Do you know KVM's release schedule, is it worth tracking the unstable version through maverick?
[17:08] <jdstrand> hallyn: I'll just upload it after testing
[17:08] <jdstrand> hallyn: fyi-- e53f27b9d9df73461308618151fa6e6392aebd85
[17:08] <jdstrand> hallyn: but I'm already working on it
[17:08] <hallyn> jdstrand: the SCM_RIGHTS one?
[17:09] <Daviey> ttx: yes, i pushed to my PPA
[17:10]  * hallyn tries to recall what aliguori told kirkland about the next kvm release
[17:10] <ttx> Daviey: ok
[17:10] <Daviey> ttx: https://edge.launchpad.net/~davewalker/+archive/groovy
[17:10] <jdstrand> hallyn: yeah
[17:10] <lau> SpamapS: thx for the link but not working yet, still trying the fix in your thread
[17:10] <jdstrand> hallyn: http://www.redhat.com/archives/libvirt-users/2010-June/msg00053.html
[17:12] <hallyn> jdstrand: the patch looks liek it should be safe with or without newer libvirt, wonder if i can just cherry-pick it into 0.12.4+noroms-0ubuntu1
[17:13] <jdstrand> hallyn: you cannot-- it doesn't apply clean. but I've already got the patch going and am building it
[17:13] <hallyn> oh, ok
[17:13] <jdstrand> hallyn: I just didn't want to step on your toes with an upload that you were already planning
[17:13] <hallyn> ok cool, thanks.  yeah i was gonna wait
[17:13] <jdstrand> hallyn: thanks
[17:13]  * ttx -> eod
[17:14] <Daviey> ttx: o/
[17:18] <kirkland> hallyn: not that that should be 0.12.4+noroms-0ubuntu2 ... i uploaded last night
[17:21]  * jdstrand is working off ubuntu2
[17:21] <kirkland> jdstrand: okay, good;  just wanted to make sure that hallyn was aware
[17:21] <hallyn> i was not
[17:21] <hallyn> kirkland: so do you recall when the next kvm release was expected?  in 2 weeks?
[17:24] <kirkland> hallyn: i think 0.13 is due around August 1st
[17:24] <kirkland> hallyn: there may be more 0.12.x releases though before then
[17:24] <kirkland> hallyn: i'd like to target 0.13 for Maverick
[17:25] <hallyn> k
[17:30] <hallyn> kirkland: oh, yeah, that one (qemu-system-i386)
[17:39] <mathiaz> zul: re puppet merge
[17:39] <mathiaz> zul: I'd rather wait a bit
[17:39] <mathiaz> zul: Debian is about to upload a new version - 0.25.5
[17:39] <mathiaz> zul: and we may be able to sync it
[17:41] <zul> cool
[17:42] <halem> nyone here using a USB microscope under linux and could tip me off about good and compatible brands etc..?
[18:01] <pmatulis> zul: any word?
[18:02] <zul> pmatulis: no i dont have a fix can you submit the bug upstream for me?
[18:02] <pmatulis> zul: yes
[18:02] <zul> pmatulis:thanks
[18:10] <Mateo_> Hey !
[18:10] <Mateo_> i didn't knew that there was a ubuntu server channel :)
[18:12] <Mateo_> maybe someone already encountered this problem: i'm trying to configure postfix (sudo dpkg-reconfigure postfix ) but at the end, i have a newaliases: fatal: bad string length 0 < 1: setgid_group =
[18:13] <lamont> what does main.cf have in it?
[18:18] <Mateo_> lamont: oh, a lot of line :)
[18:18] <Mateo_> but concerning the setgid_group =  line
[18:18] <Mateo_> i have nothing on this line after the =
[18:18] <lamont> delete it?
[18:19] <Mateo_> lamont: sure ?
[18:20] <lamont> it's not in the default main.cf, and it needs to be either the default or not there
[18:21] <Mateo_> lamont: ok, so if i have other lines like that, i comment them ?
[18:21] <Mateo_> like now i have the same error with sendmail_path
[18:21] <lamont> where did you get this file from?
[18:21] <lamont> because, yeah, nothing should be set to blank in that file
[18:22] <Mateo_> from the default install, but in my user directory
[18:23] <lamont> default package install doesn't add those variables
[18:23] <Mateo_> lamont: in fact, i have like a LOT of blank in this file :s
[18:24] <Mateo_> i didn't had any main.cf after the install, so i took the one from /usr/lib/postfix
[18:26] <Mateo_> lamont: how could i generate a brand new one with the good options ?
[18:27] <lamont> dpkg-reconfigure postfix after removing /etc/postfix/main.cf should do it for you
[18:29] <zul> smoser: ebsmount is uploaded
[18:29] <smoser> danke
[18:29] <smoser> pinged Riddell in -devel to ask him to review
[18:36] <Italian_Plumber> wow... I'm removing some directories, about 36GB, and my load average is at 1.5 ....
[18:37] <Italian_Plumber> Taking longer than I expected too...
[18:42] <Mateo_> lamont: the reconfigure tool keep on giving me blank line (sendmail_path, newaliases_path, mailq_path ...)
[18:44] <amstan> hey guys, how do i override key errors in ssh?
[18:44] <amstan> in many situations i reinstall a server and i can't reconnect anymore to it because of a key error, since they key changed
[18:44] <amstan> therefore i proceed to remove .ssh/known_hosts
[18:44] <amstan> which makes the whole system less secure
[18:45] <reisi> amstan: you can just remove the offending line, as stated in the ssh message
[18:45] <amstan> and how do i know which line offends it?
[18:45] <amstan> there is no ip addresses, nothing: U6HV2wqfAO2DiuEEZ+3K2WuXhbarpOj3SyFqargL12Iyi7xTQNgO1M5FEa5osq4Txn6PTRFLCpqzELJrq0oEfoHXiLVb7IStwGNBgGBPdkKHzkSLP0HZjwTkWLMkRydcP5vYFSmd0ub2VI8acPkEyUQ4rczftSlcvKsnqV0Q==
[18:46] <reisi> amstan: i remember there was text like "the offending key is at $HOME/.ssh/known_hosts:LINENUMBER"
[18:46] <reisi> amstan: there is actually identification data at the end of line afair
[18:46] <amstan> reisi: oh, thanks, i overlooked the line number
[18:47] <reisi> amstan: no problemos
[18:50] <sommer> mathiaz: had a chance to review the openldap-dit branch?
[19:08] <tasaif[work]> does anyone in here exist?
[19:09] <tasaif[work]> I'd like some assistance with an ubuntu server running kvm
[19:09] <jpds> tasaif[work]: Potentially.
[19:11] <tasaif[work]> I'm wondering if it's possible to install a guest and complete the installation using only the command line
[19:16] <cloakable> Why wouldn't it be?
[19:22] <tasaif[work]> cloakable: well when you run virt-install to create a guest, it starts up a graphic /win 4
[19:24] <tasaif[work]> woopsie
[19:24] <tasaif[work]> it starts up a graphic console
[19:25] <tasaif[work]> i'm wondering if it's possible to set up just a text console
[19:25] <cloakable> aha
[19:25] <cloakable> unsure
[19:25] <cloakable> ssh /w X forwarding?
[19:27] <tasaif[work]> i don't want any X involved =p
[19:27] <tasaif[work]> i'd like to be connect to the VM's output and send it my input
[19:27] <RoyK> tasaif[work]: it's in the virt-install docs - you can easily install with virt-install and just connect to that with vncviewer later
[19:27] <tasaif[work]> RoyK: Yes, but I don't want to do the installation graphically.
[19:27] <RoyK> tasaif[work]: don't you have a graphical workstation?
[19:28] <RoyK> iirc you can do with a serial console-thing as well
[19:28] <RoyK> so you can install with minicom or something
[19:29] <tasaif[work]> RoyK: I do but the key is automation
[19:29] <tasaif[work]> i can't automate a graphical installation
[19:29] <RoyK> tasaif[work]: see kickstart - that can be used to automate the install
[19:29] <RoyK> !kickstart
[19:29] <tasaif[work]> yes but that involves a premanipulated image doesn't it =p
[19:29] <RoyK> it does
[19:30] <RoyK> you don't really get automated installs for free
[19:30] <RoyK> that is, free from work
[19:30] <tasaif[work]> hmmm
[19:30] <RoyK> the installer wants to know language, timezone, how to partition the disks, a username and password for the new user etc
[19:31] <RoyK> vnc is the easy way
[19:31] <tasaif[work]> yes, but I require automation
[19:31] <RoyK> then see above
[19:31] <tasaif[work]> so vnc is out
[19:32] <tasaif[work]> RoyK: okie dokes tyvm ^__^
[19:39] <hallyn> vmbuilder?
[19:42] <hggdh> kirkland, smoser: for a bit of enlightement: on the UEC run (with Chris fixes) we had 51 failures (~5%). All these failures were metadata issues; all these instances did not get a public Ip address
[19:42] <smoser> how do you know they did not get a public ip address
[19:42] <hggdh> kirkland, smoser: now, I know very well that correlation is not causation
[19:43] <hggdh> smoser: by browsing the script log for every single one, and looking at the IP address assignments
[19:43] <smoser> "script log" ?
[19:43] <smoser> the test log output ?
[19:43] <hggdh> correct
[19:43] <smoser> hold on
[19:43] <hggdh> (running in debug mode)
[19:43] <smoser> where is "upstream" for that ?
[19:44] <hggdh> lp:~hggdh2/+junk/uec-qa
[19:44] <smoser> lp:...
[19:44] <hggdh> smoser: be prepared, some few 100s MB
[19:44] <smoser> oh
[19:44] <smoser> no
[19:44] <smoser> the test cases
[19:44] <smoser> where are those ?
[19:44] <hggdh> lp:uec-testing-scripts
[19:46] <kirkland> hggdh: okay, so this is progress, right?
[19:47] <kirkland> hggdh: what class of issues did this solve?  as it seems to have solved some, no?
[19:47] <smoser> "progress" in that its now harder to hit a race condition
[19:48] <hggdh> and progress in that I do not see any other failure apart from the metadata
[19:48] <smoser> what other failures did we see before ?
[19:49] <hggdh> please note that I am still to go thru the "not-tested" instances (which may have some 'failed to start'
[19:49] <hggdh> smoser: SSH failing, but with metadata correctly gotten
[19:50] <hggdh> so, right now, based on ~1,400 instances run, we ar at ~90% success, with ~5% failures, and another ~5% not-tested
[19:52] <smoser> hggdh, so, when you look at the logs, you see stuff like:
[19:52] <smoser> public IP was 0.0.0.0, now is X.Y.Z.T
[19:52] <smoser> for the "good" results
[19:52] <smoser> but not for the bad ones ?
[19:54] <hggdh> smoser: actually I was looking at the output of euca-describe-instances for the pub/priv addresses
[19:55] <smoser> i see that you recently added that code.  i didn't know it was there.
[19:55] <smoser> well, i think you need to be careful
[19:55] <smoser> because a terminated instance always has private == public
[19:55] <hggdh> indeed. But *not* running instances
[19:56] <hggdh> see, for example, http://pastebin.com/SHLXbnf8
[19:57] <hggdh> hum. No, this is not a good one, there is no console output for it.
[19:58] <hggdh> see http://pastebin.com/deELdKvC
[20:02] <smoser> maybe that output is from old logs ?
[20:03] <smoser> WARNING:INSTANCE i-4CDA092C:public IP was 172.19.2.24, now is 172.19.2.24
[20:03] <smoser> it seems that the current code should only output when a change is done
[20:05] <shanezilla> anyone pretty good with email server set up??
[20:05] <smoser> might be also useful to see the state of the instance there.
[20:05] <hggdh> smoser: the code wa wrong, I updated it after this run (last rev)
[20:05] <smoser> and that really should be splitting on tab, not white space
[20:05] <smoser> yeah. i thought so
[20:06] <hggdh> smoser: indeed, good idea. On this run the code would always print, even if no IP change
[20:06] <hggdh> hum. Tabs.
[20:06]  * hggdh notes it down
[20:06] <smoser> as it is, 'split()' will empty fields
[20:07] <smoser> which you dont want. i dont knwo if decscribe-instances ends up giving empty fields, but describe-images does have empity fields
[20:08] <shanezilla> looking for a yoda to help me out
[20:09] <hggdh> smoser: If sep is not specified or is None, a different splitting algorithm is applied: runs of consecutive whitespace are regarded as a single separator, and the result will contain no empty strings at the start or end if the string has leading or trailing whitespace. Consequently, splitting an empty string or a string consisting of just whitespace with a None separator returns [].
[20:10] <smoser> right.
[20:10] <smoser> which is bad
[20:10] <amstan> why does "sudo su" not make me root? i did 10 minutes ago
[20:10] <hggdh> only if the line is empty/whitespaced
[20:11] <smoser> you're supplying no separater
[20:11] <smoser> which means [[:blank]*]
[20:11] <smoser> (i think)
[20:11] <hggdh> indeed, but per the doc it means whitespace
[20:11] <_ruben> shanezilla: stating your actual problem(s)/question(s) might yield more (useful) responses
[20:11] <hggdh> blanks/tabs
[20:11] <smoser> one or more white space
[20:12] <smoser> it shrinks empty fields
[20:12] <smoser> which you do not want
[20:12] <hggdh> oh. I see the risk now
[20:12] <hggdh> will correct. Still, the results stand since I was looking at -describe-instances
[20:12] <_ruben> amstan: the "proper" command would be to use sudo -i
[20:13] <amstan> _ruben: still alex@
[20:13] <amstan> only sudo bash works
[20:14] <_ruben> amstan: root's default shell got busted? (in /etc/passwd)
[20:14] <amstan> yeah.. why is it /bin/false?
[20:14] <_ruben> because you changed it?
[20:14] <amstan> maybe i did, and it only took effect now
[20:15] <smoser> hggdh, i would agree that it certainly seems like (from lines 7 and 9) that this instance did not get an IP addresss
[20:15] <amstan> _ruben: thanks for the suggestion
[20:15] <_ruben> amstan: no problem
[20:17] <hggdh> smoser: all of the failures are similar. What I do not know is what is cause, and what is consequence
[20:17] <hggdh> or even if they are independent
[20:20] <hggdh> smoser: the field separator is only tabs?
[20:21] <shanezilla> so I was wondering when configuring the postfix instillation with dpkg under the section for the localhost section, do I put in the ip address of my local machine the server is on?
[20:22] <shanezilla> the example is mail.example.com, localhost.localdomain, localhost
[20:22] <shanezilla> I know the mail.example.com name but should the localhost.localdomain be the default?
[20:23] <shanezilla> and as well the localhost be the default 127.0.0.0 ?
[20:29] <JackTO> hi, new to linux, where do i set the ip add of the server,  need to change it... what file should i edit
[20:29] <JackTO> address*
[20:36] <tasaif[work]> JackTO: millions of articles online
[20:38] <amstan> JackTO: /etc/network/interfaces, careful though, there's a syntax required for that you either have to read the manual
[20:38] <amstan> or follow some guide on the internets
[20:40] <bogeyd6> JackTO, amstan is correct, there are many good guides on the intarwebs
[20:40] <bogeyd6> my personal fav is http://www.ubuntugeek.com/ubuntu-networking-configuration-using-command-line.html
[20:42] <hggdh> smoser: corrected and pushed. Thank you
[20:43] <JackTO> thanks guys
[20:45] <JackTO> i messed up my MySQL listening IPaddress, anyonknow where the file is to change the ip restriction and port listining number for mysql
[20:46] <mathiaz> JackTO: /etc/mysql/my.cnf
[20:47] <JackTO> thanks
[20:49] <tyska> hello guys
[20:49] <mathiaz> bug 595117,
[20:49] <mathiaz> hallyn: ^^?
[20:50] <mathiaz> hallyn: IIRC upstream qemu is using LP for their bugs
[20:50] <mathiaz> hallyn: so I would just open a task against the upstream project
[20:50] <tyska> i cant authenticate on CUPS through windows to use a printer installed on CUPS, i get this error message on the logs: windows-ext client-error-bad-request. Someone can help me?
[20:51] <mathiaz> hallyn: and state the patch should pushed in upstream first
[20:51] <Mateo_> Hey :)
[20:51] <mathiaz> hallyn: https://wiki.ubuntu.com/ReviewersTeam/ReviewGuide
[20:51] <mathiaz> hallyn: ^^ outlines a workflow to handle patches
[20:51] <hallyn> mathiaz: thx
[20:52] <Mateo_> I have another question please, i have created a user and i would like him to have the same rights that the root user
[20:52] <mathiaz> hallyn: I haven't looked at the bug specifically but it seems that the patch should be pushed upstream as well
[20:55] <hallyn> mathiaz: but isn't "open a task" waht the guy just did?
[20:55] <mathiaz> hallyn: in *Ubuntu*
[20:56] <mathiaz> hallyn: there could also be a bug in the *upstream* qemu project
[20:56] <hallyn> i thought it was on LP
[20:56] <hallyn> still looking
[20:56] <hallyn> thx
[20:56] <mathiaz> hallyn: for example, bug https://bugs.launchpad.net/qemu/+bug/584143
[20:57] <mathiaz> hallyn: you can see there is a bug in both the *upstream* project and in Ubuntu
[20:58] <mathiaz> hallyn: doing so will help in tracking down when the upstream bug will be closed
[20:58] <hallyn> so the way i read that is launchpad.net/qemu is upstream, launchpad.net/qemu-kvm is ubuntu...
[20:58] <mathiaz> hallyn: and then you can decide what is the next step in Ubuntu (eg integrate the patch in ubuntu)
[20:59] <mathiaz> hallyn: yes - launchpad.net/ubuntu/+source/qemu-kvm is the qemu-kvm package in the Ubuntu project
[21:00] <mathiaz> hallyn: https://bugs.launchpad.net/qemu/+bug/584143
[21:00] <mathiaz> hallyn: https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/584143
[21:01] <mathiaz> hallyn: ^^ these two urls represent the same bug
[21:01] <mathiaz> hallyn: one is the upstream project, the other in ubuntu
[21:01] <mathiaz> hallyn: that way if the bug is closed in upstream, we'll know it in ubuntu as well
[21:01] <hallyn> mathiaz: spurious up arrow?
[21:01] <hallyn> or are you getting repeated responses from me?
[21:02] <hallyn> mathiaz: i've marked it as also affecting qemu, thx.  the thing still bothering me is what state to elave it in :)  i can't "confirm" it really, but i'm not allowed to mark it 'triaged'
[21:02] <mathiaz> hallyn: ok - so in ubuntu you're not allowed to mark it triaged because you're not part of the bug-control team
[21:03] <mathiaz> hallyn: usually triaged comes after confirmed
[21:03] <bogeyd6> that bug isnt confirmed yet?
[21:03] <mathiaz> hallyn: so if you can't confirm the bug you shouldn't set to triage
[21:03] <bogeyd6> there i just confirmed it for lucid
[21:04] <mathiaz> hallyn: https://wiki.ubuntu.com/Bugs/Status
[21:04] <mathiaz> hallyn: ^^ see the wiki page above for an overview of the meaning of Bug Statuses in Ubuntu.
[21:04] <hallyn> bogeyd6: cool, thanks
[21:04] <hallyn> mathiaz: yes, i've seen it, and nothing seemed to apply
[21:05] <hallyn> there is nothing that says "noone has confirmed, it's up to upstream, I've passed it along to them"
[21:05] <hallyn> but now it's been confirmed.  i'm happy  :)
[21:05] <mathiaz> hallyn: ok - in that case I would open a task in the upstream project
[21:06] <mathiaz> hallyn: set the status to incomplete
[21:06] <hallyn> hm
[21:06] <mathiaz> hallyn: and leaving a comment outlining what information is missing (like can someone else confirm the bug?)
[21:06] <hallyn> again, by 'open task' you mean set "also affects"?
[21:06] <mathiaz> hallyn: yes
[21:07] <hallyn> is verbal confirmation on irc sufficient?
[21:07] <mathiaz> hallyn: for example, bug 595117 has two tasks now: one in the upstream project, one in Ubuntu
[21:07] <mathiaz> hallyn: if you can trust the reporter on IRC
[21:08]  * hallyn trusts noone
[21:08] <mathiaz> hallyn: you could ask the reporter on IRC to add a comment to LP
[21:08] <mathiaz> hallyn: :)
[21:08] <bogeyd6> ala james bond style
[21:08] <hallyn> bogeyd6: would you mind commenting in LP?
[21:10] <mathiaz> hallyn: if I don't know what status to set, I usually set it to Incomplete (in Ubuntu) and leave a note about the peace of information/next step to take in order to move the bug to another state
[21:11] <mathiaz> hallyn: so setting the bug to Incomplete in Ubuntu and leaving a note ""noone has confirmed, it's up to  upstream, I've passed it along to them"
[21:11] <mathiaz> hallyn: is good enough IMO
[21:12] <hallyn> mathiaz: ok, thanks very much.
[21:12] <mathiaz> hallyn: you're welcome!
[21:25] <JackTO> i'm trying to join the #mysql channel, and my nick is registed, but it says I cannot send to channel..  anyone know why?
[21:36] <_ruben> JackTO: you dont seem to be identified by nickserv
[21:36] <brandon_>  guys how can i prevent certin comands from running in sudo? like rm -rf
[21:42] <brandon_> anyone?
[21:42] <hallyn> brandon_: well you could deny rm using /etc/sudoers, but can't select on arguments that i know of
[21:43] <brandon_> but i can disable rm alltogether right?
[21:43] <hallyn> should be able to yes
[21:43] <hallyn> course then the user can probably cp /bin/rm /tmp/notrm and use that
[21:43] <hallyn> so blacklisting woudl be tough to get right
[21:43] <brandon_> h,,
[21:43] <brandon_> hmm
[21:44] <hggdh> and... you might break usage
[21:44] <hallyn> if you can whitelist, that'd be good
[21:44] <hallyn> (i.e. if all they need to do is be able to ifconfig)
[21:44] <hallyn> brandon_: i'd go ask on #ubuntu-security
[21:44] <hallyn> but anyway 'man 5 sudoers' is full of info
[21:45] <hggdh> hallyn: go ping pedro_ for you application to -control... it is in his hands now ;-)
[21:45] <brandon_> well here is the scenario, we have a server running for students to mess with and understand root there is 26 students we just dont want an ass to bring it down since there are 25 other students using it
[21:46]  * hallyn looks around for pedro_
[21:46] <hallyn> hggdh: though i assumed it woudl take awhile for something liek that to happen, i.e. i shoudl "prove myself" or something :)
[21:47] <hggdh> hallyn: indeed. But the bugs you provide as example are good enough (apart from my comment on Importance)
[21:47] <hggdh> and you got two +1, which is enough
[21:48] <hallyn> sweet
[21:48] <hggdh> pedro_ is in #ubuntu-bugs
[21:48] <hggdh> or #distro
[21:48] <hallyn> brandon_: maybe use schroot?
[21:48] <hallyn> or lxc
[21:48] <hggdh> or evn set a VM, and let them run under the VM
[21:48] <hallyn> giving each one his own container shoudl be about as lightweight as each their own login shell
[21:49] <brandon_> im unfamiliar with those two what would it allow me to do?
[21:49] <hallyn> heh, he said 'students' so i assume they're light on hardware
[21:49] <hallyn> brandon_: it's not completely foolproof, but it should protect your host from any serious damage from the students
[21:49] <brandon_> we have a full server rack :D
[21:49] <brandon_> and give them root access?
[21:49] <hallyn> oh then why not just give each a kvm VM?
[21:50] <hallyn> yes, root access, but limited somewhat to their own container
[21:50] <hallyn> so what they see as / is actuall /var/lxc/container1
[21:50] <hallyn> when they edit /etc/ it's actuall /var/lxc/container1/etc
[21:50] <hallyn> and, with their own virtual network interface so they won't ifdown the real eth0
[21:51] <brandon_> and they can ssh into it?
[21:51] <hallyn> brandon_: yup
[21:51] <kees> brandon_: "rm" is a system call; you can't really remove it.  perl -e 'unlink("/the/file");'  etc
[21:53] <brandon_>  and if they rm -rf it only kills there container? if so thats what ill use just point me towards a guide
[21:54] <hallyn> brandon_: well if your hardware is beefy enough to do kvm for each of them, then http://www.linux-kvm.org/page/Main_Page
[21:55] <hallyn> otherwise, lxc.sf.net (link to tutorial on the right)
[21:55] <fr1sco> i'm sure this isn't possible but, anyone know of a way to start screen in an open local session while ssh'd?
[21:55] <hallyn> kees: what's the best intro to schroot?
[21:56] <hallyn> fr1sco: 'open local session'?  don't get what you mean
[21:58] <fr1sco> hallyn: i was working physically at a svr and left the terminal session open but forgot to start screen.  now i'm away but wanted to get back into that session.
[21:58] <hallyn> oh, just screen -d
[21:58] <hallyn> well, screen -list to get the name if there are more than 1 open
[21:59] <hallyn> screen -dr somesessionanme
[21:59] <hallyn> will detach that one, and reattach you to it from your new login
[22:01] <fr1sco> sweet thx
[22:01] <hallyn> np
[22:02] <kees> hallyn: "best"?
[22:02] <kees> hallyn: I would say: https://wiki.ubuntu.com/SecurityTeam/BuildEnvironment
[22:02] <kees> hallyn: or just "man mk-sbuild"
[22:03] <hallyn> kees: yeah that's the link i used, but i was wondering if that was the best for brandon_
[22:04] <hallyn> hm, really i guess lxc might be the better way to go anyway (keep hte kids from kill-9ing each other)
[22:05] <brandon_> so lxc over kvm
[22:06] <hallyn> brandon_: that'd probably turn into a vi-vs-emacs style debate :)
[22:06] <hallyn> i think lxc would be more appropriate...  but it has more gatchas still
[22:06] <mathiaz> brandon_: which version of Ubuntu are you planning to run on?
[22:06] <hallyn> particularly depending on your distro
[22:06] <hallyn> distro/version
[22:07] <mathiaz> brandon_: does your hardware support VT extensions?
[22:07] <brandon_> well its probably going to be one server ubuntu on server fedora
[22:07] <brandon_> it may i would have to examine it
[22:08] <mathiaz> brandon_: if your hardware doesn't support VT extensions then it rules out kvm
[22:08] <hallyn> and qemu for a bunch of kdis would be far too slow
[22:08] <brandon_> what does kvm do that lxc dosent?
[22:08] <mathiaz> brandon_: are you planning to use the same Ubuntu release for all of the students?
[22:08] <hallyn> it virtualizes the hardware
[22:09] <brandon_> same ubuntu on one and same fedora on the other
[22:09] <brandon_> 2 servers
[22:09] <hallyn> how many kids did you say?
[22:09] <brandon_> 26
[22:09] <brandon_> with acces to both servers
[22:09] <brandon_> so 26 people per server running 24/7
[22:11] <hallyn> brandon_: AFAIK most ppl look to http://blog.bodhizazen.net/linux/lxc-configure-ubuntu-lucid-containers/ for setting up ubuntu containers
[22:14] <brandon_> thanks all look into it from here thanks so much for the help hallyn
[22:39] <wtf1> what is partial upgrade?
[22:51] <kirkland> smoser: ping
[22:51] <zul> smoser: it got rejected again?
[22:51] <kirkland> smoser: could you take a look at https://help.ubuntu.com/community/UEC/Provisioning/DD-WRT
[22:52] <kirkland> smoser: and update it accordingly?
[22:52] <kirkland> smoser: seems like you and I have looked at this a few times;  i decided to document it once and for all
[23:07] <hggdh> kirkland: on that page, s/Aadress/Address/
[23:08] <hggdh> kirkland: nah, forget, I can do it myself...
[23:11] <kirkland> hggdh: thanks