[00:00] <yann2> hello! one small question: can VM in kvm & libvirt run as a low privileged user?
[00:40] <PryMaL> does anyone happen to know the earliest version of ubu server kernel that supports IPV6?
[01:04] <deadsmith> Is noefi required to boot on an Apple XServe?
[01:05] <twb> AFAIK GRUB 2 supports EFI.
[01:06] <twb> I've never dealt with such a system myself, but I *assumed* it all Just WorkedTM
[01:09] <deadsmith> The discovery from earlier today was that I had to build from the bzr source for grub to boot the EFI....
[01:09] <deadsmith> And, it is booting....
[01:09] <deadsmith> but hangs when loading the kernel.
[01:09] <deadsmith> I think it's video weirdness, but hard to be sure...
[01:10] <deadsmith> is the error "finding suitable mode.  Booting however" familiar to anyone?
[01:12] <twb> Sorry, no.
[01:37] <triode3> hello all, got a mdadm question. built server, 4 drives, / and swap on raid1 on sda1,sdb1 sda2,sdb2 home on sda3,sdb3. Parts sda4,sdb4,sdc1,sdd1 exist. when machine boots it auto-magically mounts the extras as /dev/md0, while root is /dev/md0. Ugh. Can't find solution.
[01:44] <thesheff17> triode3: I have never seen that happen before...do you see anything in /etc/fstab
[01:44] <twb> triode3: pastebin the contents of /proc/mdstat
[01:45] <triode3> http://pastebin.com/SvrSAedY
[01:45] <triode3> I changed fstab from the default (which listed three devices) to one which still lists three (there should only be three)...
[01:45] <twb> triode3: I'm not sure what happened there
[01:45] <pmatulis> ew, partitionable raid devices
[01:46] <triode3> The issue is that I want to build the final raid array (sda4, sdb4, sdc1, sdd1) but it will not let me as it thinks they are bsing used (sda4, sdb4).
[01:46] <twb> Oh
[01:46] <twb> Just mdadm --stop /dev/md0, then.
[01:47] <triode3> e.g md_d0 should not exist.
[01:47] <twb> triode3: md_d0 exists because both two-disk arrays want to be called md0
[01:47] <triode3> mdadm --stop /dev/md_d0 stops it, but it still will not let me create one with mdadm --create blah blah
[01:47] <twb> triode3: you need to stop one pair, add each of its partitions to the other array, then fix mdadm.conf, then update your ramdisk
[01:48] <triode3> twb: thats fine, but /dev/md0 is my root. :O will stopping that kill the server :OO
[01:48] <twb> triode3: then boot from the other pair
[01:49] <triode3> hrm I only have the OS on /dev/sda /dev/sdb (raid1). I was going to do raid5 on sda4,sdb4,sdc1,sdd1 for space.
[01:49] <twb> triode3: oh, you want a *raid5* of the four disks
[01:49] <triode3> As for fixing mdadm.conf, I _thought_ this was ok: http://pastebin.com/GCu2sdiA
[01:49] <triode3> twb yes
[01:50] <twb> triode3: I think easiest is to reinstall from scratch
[01:50] <triode3> on sda1,sdb1 /home on sda2,sdb2 swap on sda3,sdb3 /space on sda4,sdb4,sdc1,sdd1
[01:51] <triode3> twb: one more monkey wrench, this is IDE, and I only have two controllers. Starting from scratch means disconnecting one, which is why I am in this pickle... when I first brought up ubuntu server, it automagically created this strange md_d0 array
[01:53] <triode3> twb: I think the other problem is this: http://pastebin.com/hVYPiysE it is still there after removal. After a reboot too (well, obviously).
[01:56] <twb> triode3: I'm not really interested in helping you in-place migrate a 2-way RAID1 to a 4-way RAID5.
[01:59] <triode3> twb: I am not trying to migrate. I am trying to figure out why it is auto-creating my fourth raid (when I didn't ask it to) with the wrong drives (when I did not tell it to). I am assuming it is one of those "mdadm developers think users are dumber than they appear" ideologies... I was just wondering if anyone else had seen that and overcome it. I _should not_ have four raid arrays, and I am not wanting to migrate, just t
[02:04] <twb> triode3: edit mdadm.conf, then
[02:04] <twb> triode3: if you used dd, it may be that the UUIDs on the new disks match the old ones, so when it does an mdadm --assemble --scan, it finds two complete copies of md0 and assembles both
[02:05] <twb> Writing zeroes over the metadata blocks and/or telling mdadm.conf to assemble specific devices should address that -- don't forget to rebuild the ramdisk after editing mdadm.conf
[02:08] <triode3> twb: hrm, I will try that.
[02:08] <triode3> twb: thanks.
[02:12] <twb> It still won't help when you try to turn a RAID1 into a RAID5
[02:16] <triode3> twb: hrm, but it reports that it thinks its raid5: ARRAY /dev/md0 level=raid5 num-devices=3 UUID=a6d72d16:694f7d01:3ccdc9df:08582ff7 it is of course that it is listing /dev/md0 twice. Thats f00ked up.
[02:16] <twb> The mdstat you sent me said the level was raid1.
[02:16] <triode3> twb: hrm looking at that again, I think I am screwed. It lists /dev/md0 twice, as a raid1 and a raid5. I think I will just have to somehow get a cdrom/usb in there and rebuild the whole thing
[02:17] <twb> Just editing mdadm.conf to look for a raid5 will not MAKE it a raid5.
[02:17] <triode3> twb: yes, but the mdadm --examine --scan reports two: http://pastebin.com/hVYPiysE
[02:18] <twb> That might just be returning what's in mdadm.conf -- I don't know
[02:18] <triode3> twb: yes. again, I think this thing is foo-bar. I think I will have to rebuild. mdadm.conf only lists three arrays. :O
[02:18] <twb> 12:50 <twb> triode3: I think easiest is to reinstall from scratch
[02:18] <triode3> twb: yes. at this point, I agree... although I will have to finagle some kind of medium.
[02:19] <triode3> twb: thanks for the input. Sorry for the confustion.
[03:27] <e_t_> I've got a server with three network interfaces, all with static IPs. Which ever connection I set a gateway on WILL respond to pings, but the other two WILL NOT. How can I get ping responses from all three interfaces?
[03:46] <twb> All three interfaces are on the same network?
[03:49] <e_t_> No. One is 10.145.144.7, two is 192.168.254.254, three is 172.17.1.1.
[03:51] <e_t_> But the network (which I don't control) is set up so that all address *should* be reachable. Tcpdump shows that pings are recieved, but they aren't answered, except for the interface with the gateway set.
[03:52] <Delerium_> ICMP blocked by some kind of network device?!
[03:52] <e_t_> Nope.
[03:53] <CheetoBandito> sounds like its received but dumped before apps can see it
[03:53] <CheetoBandito> I'm having weird issues with networking in 10.04 concerning multicast too
[03:56] <e_t_> Shouldn't ICMP be answered directly by the kernel's TCP/IP stack?
[03:57] <twb> e_t_: you probably need a gateway for each interface, with a lower metric for non-preferred the routes
[03:58] <twb> e_t_: without a full network diagram it's difficult to be sure
[03:58] <twb> e_t_: if you're pinging from within the same network, then it should work already
[03:59] <e_t_> I'm pinging from a host on a different subnet, but as I said, the pings are received, but the computer doesn't respond.
[03:59] <e_t_> Can I set metrics in /etc/network/interfaces?
[04:00] <twb> post-up ip route add ...
[04:00] <twb> post-down ip route flush dev ...
[04:00] <twb> Or some variation thereof
[04:01] <twb> e_t_: if your three-iface host doesn't have appropriate routing rules to get back to your box (well, at least for the first hop), there's no way the return ping will reach you
[04:04] <e_t_> OK. I'd read this: http://kindlund.wordpress.com/2007/11/19/configuring-multiple-default-routes-in-linux/ about setting rules, but I wanted to make sure I wasn't missing something simple.
[04:05] <twb> e_t_: it wouldn't surprise me if your netadm blocks all ICMP traffic
[04:08] <e_t_> I can set a gateway on any one interface at a time and get ping responses from it. Besides, my network admin (my teacher at college) certified to me that ICMP was not blocked.
[04:09] <twb> Is the machine in question intended to be a router, or is it simply multihomed?
[04:12] <CheetoBandito> I'm having an interesting issue in 10.04... If I request packets from a specific group in one program, and a different group in another program... I get packets from both groups in both programs.
[04:13] <e_t_> twb: It's really just supposed to be an end-point on the 192 and 172 networks. Traffic never needs to cross from one network to another (routing). On the 10 network, it's a host.
[04:13] <CheetoBandito> If I compile and run the same scenarion on Windows, Solaris, and other distros I get the intended behavior, that the packets each program sees are only the ones they requested.
[04:14] <twb> e_t_: and it should be able to reach all distal networks via any interface?
[04:15] <e_t_> twb: that is my understanding.
[04:16] <twb> My guess is your route table should look like this, then: http://paste.debian.net/103781/
[04:16] <twb> Er, 10/8 should be the third line, not that it matters
[04:18] <twb> oh, and I forgot "via 1.2.3.4" fields for the 0/0's
[04:39] <e_t_> twb: With multiple gateway set in /etc/network/interfaces, this is my routing table: http://paste.debian.net/103782
[04:39] <twb> e_t_: please pastebin "ip r" formate.
[04:40] <e_t_> twb: http://paste.debian.net/103783
[04:41] <twb> Yeah, OK, though you have no 0/0 (i.e. default) for eth2.
[04:42] <twb> That ought to be enough for ICMP to work, although for connection-oriented protocols like TCP I think you'll want to use non-equal metrics (so all traffic goes over one default route)
[04:43] <twb> I'm actually kinda surprised ICMP didn't work across the internal wossname without those routes...
[04:43] <twb> Do you have a firewall running on this box?
[04:44] <e_t_> No. I even ran iptables -F to be sure. Security is not a concern for this server.
[04:50] <skorv> my current setup is ubuntu server 10.04 +bind9 +dhcp3-server
[04:51] <skorv> in your opnion what may crash networking
[04:52] <e_t_> skorv: What does "crash networking" mean?
[04:53] <skorv> one minute everythin
[04:53] <skorv> sorry
[04:53] <skorv> one minute everything is nice the next i cat even get an ip
[04:54] <skorv> i go to the server and do sudo /ets/init.d/networking restart and it goes back to being ok... but its temporary
[04:55] <e_t_> Which machine can't get an IP?
[04:55] <skorv> everyone
[04:55] <skorv> server setup = 3 NIC 1 wan 2 lan
[04:56] <skorv> it serves 2 lans
[04:57] <skorv> i tried wit static and dynamic ip on interfaces on wan side. lans have static ip
[04:58] <e_t_> The server has static IPs and it gives out DHCP addresses to lan clients?
[04:58] <skorv> correct
[04:59] <e_t_> But after a minute, the lan clients no longer receive DHCP addresses?
[04:59] <skorv> it basically has the same effect as ifdown -a
[04:59] <skorv> yes
[04:59] <skorv> all freezes
[05:00] <skorv> i restart network only and it goes back to normal
[05:00] <skorv> clients can get ip and internet works
[05:01] <e_t_> Can you run Wireshark on the lan clients, to see what the network traffic looks like?
[05:02] <skorv> its 5 am and i'm in bed...been working on it sine 10 am yesterday
[05:02] <skorv> *since
[05:03] <skorv> i'll do it in a couple of hours
[05:03] <skorv> can a packet "crash" networking?
[05:04] <skorv> or trafic?
[05:05] <e_t_> Misconfiguration might cause any number of network-disabling effects.
[05:05] <skorv> after a random ime?
[05:05] <skorv> * time
[05:07] <skorv> its the 2nd time it happens... i even thought it was a bed install the last time
[05:08] <skorv> so i redid everything by hand istead of copying the config files i already have
[05:09] <skorv> i'm portuguese... excuse my english
[05:09] <skorv> i fallowed tutorials on ubuntu forums for dhcp and bind
[05:12] <skorv> i'm a windows server expert... 1st time serious with linux server (my 1st had desktop installes)
[05:12] <skorv> *installed
[05:13] <skorv> was looking forward to make a good sbs alternative
[05:14] <skorv> i have no training... self taught on almost everything
[05:15] <skorv> this nre server was to be my "crown jewel"
[05:15] <skorv> so far its crap (due to me most likely)
[05:16] <e_t_> I've always had good luck with dnsmasq, rather than bind.
[05:18] <skorv> well... fresh install tomorrow i nothing else works
[05:20] <skorv> i'lll give dnsmasq a chance as long as it supports multiole networks
[05:20] <skorv> *multiple
[05:21] <e_t_> It does. It can also do DHCP.
[08:33] <vraa_> anyone here know details about canonical support?
[11:29] <cjwatson> kirkland_: thanks, reviewing ssh-import-id now
[11:36] <RoyK> isn't that ssh-copy-id?
[11:39] <cjwatson> RoyK: no
[11:39] <cjwatson> RoyK: two different tools
[11:40] <RoyK> can't find the import one on lucid...
[12:27] <pmatulis> RoyK: it's part of openssh-server i think
[12:39] <kiall> Hiya .. I've just come across a strange situation .. i have a hostname in /etc/hosts - root can ping it fine .. but, other users cant resolve the host? Its like /etc/hosts is only being honored for the root user
[12:46] <pmatulis> kiall: try strace, it's probably a permissions thing
[12:47] <pmatulis> 'strace -e open ping localhost'
[12:47] <kiall> pmatulis, it was - I just discovered the issue .. /etc/nsswitch.conf was 664 rather than 644..
[12:47] <pmatulis> kiall: ok
[12:47] <kiall> sometimes a "smart" cloud-init script turns out to not be so smart ;)
[13:31] <webstyler> Can someone recommend CrashPlan online backup?
[13:36] <kim0> webstyler: I can :)
[13:36] <webstyler> thx
[13:36] <kim0> webstyler: been using it for a year+ .. works great! Linux/Mac/Windows
[13:37] <webstyler> yeah. thats nice
[13:37] <webstyler> need linux supprt
[13:37] <webstyler> i am considering backing up my server/nas off-site
[13:37] <webstyler> and the unlimited plan seems to be a good solution
[13:37] <webstyler> how is the restore/web interface?
[13:39] <raphink> there's also the great S4 API: http://www.supersimplestorageservice.com/
[13:39] <raphink> :-)
[13:40] <webstyler> ok thx
[13:40] <raphink> don't try it in production though ;-)
[13:41] <webstyler> ok good to know
[13:43] <xampart> http://www.youtube.com/watch?v=yGyRVymDBhw
[13:43] <xampart> wrongdow
[13:43] <pmatulis> xampart: thanks for that
[13:44] <xampart> pmatulis: np =)
[13:45] <pmatulis> xampart: he lost a testicle in 1999?
[13:48] <xampart> pmatulis: i suggest you look the other videos from him. on that information i would presume, no.
[13:50] <pmatulis> xampart: i'm not much into contemporary music.  is this a real music video?
[13:54] <hggdh> smoser: I am in now
[13:56] <CERNUNN0S> Greetings, i'm new to the whole Ubuntu server admin and have inherited a number of servers unfortunately it appears one has been compromised. In so far that I am receiving reports of spam and helo senders, could anyone point me in the direction of a good set of guides to remedy these teething problems. The server is running Lucid Lynx LTS
[13:56] <CERNUNN0S> Thank you for your time.
[13:59] <patdk-wk> CERNUNN0S, running a website?
[14:00] <patdk-wk> probably one of your cgi's was hacked into
[14:00] <patdk-wk> hopefully the kernel/libc are updated also
[14:00] <hggdh> CERNUNN0S: first action, once confirmed it has been compromised, is to take it off the network
[14:02] <CERNUNN0S> Yes running virtualhosts, and hosting two email accounts and various aliases for email forwarding, how would I go about assessing the compromised systems?
[14:04] <hggdh> ugh
[14:04] <gobbe> not quite easy job
[14:05] <gobbe> you could run checksums of files and compare them with system thats known to be good
[14:05] <CERNUNN0S> Main problem is the set-up I have been left with is near incomprehensible, very basic and largely incompletely and completely without documentation. Would be great if there was different environments for test, developing. But alas no it's all as one in a monstrous multi-master-fail
[14:06] <gobbe> i wouldn even start to find out. i would install everything from sratch
[14:07] <hggdh> CERNUNN0S: is the system up-to-date with security fixes? kernel, apache, and libc6 comes to mind as possible venues
[14:07] <kiall> gobbe, yup - you either know what you're doing .. or not.. and really when the server is compromised and offline .. its not the time to learn when you dont even understand the setup :)
[14:08] <gobbe> yeah
[14:08] <gobbe> just wasting time
[14:08] <gobbe> because you will end up installing them from scratcg anyway
[14:09] <hggdh> actually, no, not wasting time. If you do not know how you got compromised, there is no guarantee it will not happen again
[14:09] <gobbe> gd.....its hard to irc with mobile phone
[14:10] <hggdh> the safe bet is it *will* happen again
[14:10] <CERNUNN0S> I don't think my predecessor believed in keeping things up to date so I would think it would not be running the latest software. So the answer is i'm wasting my time. Well i've already started creating a new server cluster completely separate from these. I was hoping the duck tape would hold on this set up long enough for me to get on with the new setup
[14:10] <gobbe> hggdh, well....finding it out could be impossible
[14:10] <hggdh> gobbe: I agree. But it should be tried. Part of the cost ;-)
[14:10] <gobbe> CERNUNN0S: install proper tools to prevent this in future also
[14:11] <gobbe> hggdh, yes. if you have atleast somekind of knowledge and tools
[14:11] <gobbe> in this case i would say that finding out could be near impossible
[14:11] <CERNUNN0S> The new set up includes tools and complete logs unlike current garbage,
[14:11] <gobbe> good
[14:12] <Fidelix> Hey guys, i'm having a permission problem in my server. The script runs as www-data but it cant read files uploaded through FTP by the user "rs"
[14:12] <hggdh> CERNUNN0S: try to figure out who is sending out the HELOs (apache?) -- if cgi, disable the suckers
[14:12] <CERNUNN0S> Thought as much but thought i'd ask the guys and girls who deal with this on a daily basis
[14:13] <gobbe> CERNUNN0S: i would bet that apache/php could be good place to start out
[14:13] <smoser> hggdh, ok. i'm in too.
[14:14] <smoser> zul, could you maybe do a sponsor for me ? bug 686124
[14:15] <zul> smoser: sure gimme a sec..
[14:15] <CERNUNN0S> Thanks all I shall have a check and see if there is anything I can gleen from the apache/php files
[14:16] <hggdh> smoser: I am thìnking of you, pedro, and myself together, OK with that?
[14:17] <smoser> yeah.
[14:17] <raphink> Fidelix, what are the rights for the uploaded file?
[14:19] <Fidelix> raphink, just a sec.
[14:19] <Error404NotFound> anyone has installed NagiosXI on Ubuntu successfully? I have been doing a lot of google but can only find its manual for centos/rhel
[14:20] <Fidelix> raphink, -rw------- 1 rs       users       696320 2011-01-05 12:20
[14:22] <raphink> Fidelix, and are you surprised that www-data can't read it?
[14:22] <Fidelix> raphink, not really. But can you give me a light on how to change the default permission of uploaded files in vsftpd ?
[14:22] <raphink> sure
[14:23] <raphink> give me a minute
[14:24] <raphink> Fidelix, you could set local_umask in vsftpd.conf for example
[14:24] <raphink> there's more fine umask settings, too
[14:25] <Fidelix> raphink, i'll try to set local_umask to 777
[14:25] <raphink> hmmm
[14:26] <Fidelix> raphink, -rw------- 1 rs
[14:26] <raphink> I would say more like 022
[14:26] <Fidelix> It did not work
[14:26] <Fidelix> oh, ok
[14:26] <genii-around> umask doesn't have same values chmod does, it uses the complements of 777 etc
[14:26] <raphink> a umask of 022 will give you files are that 644
[14:27] <raphink> another option would be to use the chown_username setting in vsftpd.conf
[14:27] <Fidelix> raphink, -rw-r--r-- 1 rs
[14:27] <raphink> so files that are uploaded are given to www-data
[14:27] <raphink> Fidelix, yes, that's 644, and it should allow www-data to read it
[14:28] <genii-around> raphink: I'm pretty sure 022 will give you perms of 755
[14:28] <Fidelix> raphink, i already tried chown_username
[14:28] <Fidelix> Files are simply not chowned
[14:28] <raphink> genii-around, 022 gives 755 to dirs, 644 to files
[14:28] <raphink> Fidelix, did you set chown_uploads to yes?
[14:28] <Fidelix> yes i did
[14:28] <raphink> hmmm
[14:29] <Fidelix> But my problem is already solved. I thank you for your time!
[14:29] <raphink> I use chown_username on my server and it works fine
[14:29] <raphink> no problem Fidelix
[14:29] <raphink> ;-)
[14:31] <raphink> genii-around, the complement is 777 for dirs, 666 for files
[14:33] <zul> smoser: done
[14:33] <smoser> gratzie
[14:36] <macno> Hi, I have to setup a bridge interface on maverik , and I need both ipv4 and ipv6
[14:37] <macno> this is my config http://paste.ubuntu.com/550654/ but I can't reach host using the ipv6
[14:38] <Error404NotFound> macno, not directly related to your question but why are you using IPv6? just playing and see how it works? or you are actually deploying it in a network?
[14:39] <macno> Error404NotFound, we use it
[14:40] <macno> Error404NotFound, I'm connected to freenode using ipv6
[14:40] <Error404NotFound> macno, hmmm, i see... i am just using gogoc to experiment it :)
[14:40] <Error404NotFound> macno, so am i, but using different nick :P
[14:44] <gobbe> macno: can you run ping6 for example?
[14:45] <macno> gobbe, no, but I think I found the problem right now
[14:45] <macno> 1 sec
[14:47] <macno> gobbe, yep, the hwaddress row was wrong
[14:47] <gobbe> ok :)
[14:47] <gobbe> sometimes when asking help, you realize by yourself what was wrong =)
[14:47] <macno> it must be under the eth0 section
[14:47] <gobbe> usually happens to me =)
[14:48] <macno> not under the br0 one
[14:48] <gobbe> yes
[14:49] <zul> hggdh: is the iso still failing?
[14:50] <hggdh> zul: fix was published today, should be on tomorrow's ISO
[14:50] <hggdh> zul: so yes, still failing ;-)
[14:51] <zul> hggdh: meh
[14:52] <hggdh> heh
[15:27] <smoser> Daviey, fyi, i'm using pad.daviey.com more and more
[15:27] <smoser> very useful until we manage to get a pad.ubuntu.com
[15:36] <gobbe> uuh
[15:36] <gobbe> quite nice tool :)
[15:36] <azizLIGHTS> where is the per-user www root by default?
[15:36] <azizLIGHTS> for apache2
[15:37] <RoyK> $HOME/public_html
[15:37] <RoyK> by default
[15:37] <RoyK> if you enable usermod (iirc)
[15:37] <thesheff17> that does look like a sweet tool...does it have a name?
[15:37] <azizLIGHTS> how do i know if i enabled usermod
[15:38] <RoyK> mod_userdir
[15:38] <azizLIGHTS> sorry what?
[15:39] <RoyK> azizLIGHTS: a2enmod userdir
[15:39] <RoyK> then apache2ctl graceful
[15:40] <azizLIGHTS> thanks that worked
[15:40] <RoyK> azizLIGHTS: see /etc/apache2/mods-enabled/userdir.conf for the drive
[15:40] <RoyK> s/drive/directory/
[15:46] <hallyn> huh.  pad.daviey.com.   yeah, nice.  we should find some ppl to test how high it scales :)
[15:46] <gobbe> hih
[15:47] <gobbe> i would like to get something like that to my office
[15:47] <gobbe> would be usefull there :D
[15:47] <hallyn> i see ,it's etherpad.  i thought Daviey had coded it up :)
[16:05] <azizLIGHTS> is it ok to have joomla, oscommerce, and zen-cart all on same system
[16:05] <azizLIGHTS> or should i make snapshots for my vm
[16:05] <azizLIGHTS> 3 snapshots , 1 for each
[16:19] <gvandeweyer> hi all. could someone give me an opinion on this: Our server is running against it's limits and i'm looking at the ubuntu private cloud option.  Will this be the way to go? We now have a single server that runs very specific tasks (home-brewed code), and it should be able to distribute the tasks to spare cpu's in some core-duo's that are now lying around
[16:19] <gvandeweyer> however, as far as I get the guides, the cloud starts regular ubuntu systems (virtual) on each node, that you can login to (ssh?)
[16:21] <gvandeweyer> would it be possible to launch these speficic programs on these nodes, and how about storage after loggin in? will each user still have their own homedir with their data as they do now?
[16:22] <gvandeweyer> and as a side node: isn't starting virtual complete OS's a bit overhead ?  should i go for some sort of cluster option, with a system running ubuntu, accepting tasks from the network monitor and running it in their native OS ?
[16:22] <compdoc> your users would use it to store files?
[16:22] <gvandeweyer> they would need their data to process them
[16:23] <gvandeweyer> and preferable store them somewhere on it to (several 100 gigs, not suitable for desktop usage)
[16:24] <gvandeweyer> I could use NFS/samba for remote access if using physical machines that get jobs from a central queue
[16:24] <compdoc> when I think of the could, I think of users connecting to a public or private server to save files, store contacts, and maybe use apps to write docs or spreadsheets. I havent found a use for it myself
[16:24] <compdoc> *cloud
[16:24] <vraa> does anyone have any experience with ubuntu canonical server support?
[16:24] <gvandeweyer> compdoc: i feel the same, but it looks easy to implement (compared to cluster design)
[16:25] <compdoc> a cluster is a group of servers that prevent interruption of service from the failure of one server. I dont think the cloud is the same kind of thing
[16:26] <compdoc> clouds can use clusters, certainly
[16:30] <hallyn> hm, i've been trying the 'test memory' option in all the livecds, but it doesn't seem to ever give me the full memtest86 i expect...  just blanks for 2 seconds and then comes back to the grub menu
[16:30] <hallyn> what am i doing wrong?
[16:32] <AndyGraybeal> should I ever be worried about cron not working?  does it ever not work?  I'm depending on it to rsync data.
[16:34] <kirkland> smoser: around?
[16:36] <azizLIGHTS> is it ok to install joomla, oscommerce, and zen-cart on same machine? or should i seperate the vm into 3 snapshots
[16:41] <hallyn> kirkland: can you try booting an ubuntu livecd in kvm and running memtest?  does it actually run memtest?
[16:41] <patdk-wk> azizLIGHTS, it depends on how paranoid you are about one of them getting hacked, and if you care if cross contamination
[16:41] <kirkland> hallyn: let me try ....
[16:41] <azizLIGHTS> im just testing right now
[16:42] <azizLIGHTS> its not live
[16:42] <kirkland> hallyn: nope, just reboots
[16:46] <hallyn> kirkland: hrmph
[16:47] <hallyn> kirkland: a rhel cd did the same thing for me.
[16:47] <hallyn> kirkland: doh, i'm an idiot
[16:47] <kirkland> hallyn: qemu problem then?
[16:48] <hallyn> yeah, the very one that i thought was responsible for jdstrand's bug
[16:48] <hallyn> (see http://www.spinics.net/lists/kvm/msg47235.html)
[16:48] <hallyn> I guess I'll file a separate bug, pull the bugfix, and then ask jdstrand if it fixes his problem too :)
[16:49] <patdk-wk> how do I make lvm forget about vg's that no longer exist?
[16:49] <hallyn> patdk-wk: not sure, but in the past I've had to dd from if=/dev/zero into the first bit of the old vg bc the tools were finding non-deleted header info, iirc
[16:50] <patdk-wk> hallyn, no, the drive doesn't exist anymore
[16:50] <patdk-wk> just lvm keeps attempting to locate it
[16:50] <hallyn> then i guess you prolly can't write zeros into it :)
[16:50] <patdk-wk> :)
[16:51] <azizLIGHTS> o crap
[16:52] <azizLIGHTS> i was saving vmware snapshot and ubuntu says BUG: soft lockup - CPU#0 stuck for 112s
[16:52] <hallyn> patdk-wk: so, you had a vg spanning multiple partitions, one ofthe partitions is gone, and lvm keeps trying to find the partition that's gone?
[16:52] <patdk-wk> nope
[16:52] <azizLIGHTS> is it ok or what...
[16:52] <patdk-wk> I haved a vg on a single drive, that drive is gone (cause I was recovering it)
[16:53] <patdk-wk> but vgs throws out a crapload load of, read failed after 0 of 4096 at 0: intput/output error
[16:53] <hallyn> patdk-wk: hm, does 'vgremove' help at all?
[16:53] <patdk-wk> hallyn, nope, just says, not found
[16:53] <smoser> kirkland, here now
[16:54] <kirkland> smoser: okay, so ssh-import-id is in the archive
[16:54] <kirkland> smoser: wanna propose your merge with your getopt rework?
[16:55] <kirkland> smoser: and i'll review sponsor upload
[16:56] <smoser> ok.
[16:57] <patdk-wk> hallyn, seems to be cause I didn't realize it, and never did a vgchange -an on it before I removed it
[16:58] <hallyn> hm, you'd think that'd be fixable after the fact...
[17:00] <doko> eucalyptus guys: please change the order of the recommends in eucalyptus-walrus, bittornade should go first
[17:01] <patdk-wk> hallyn, evil, but deleting the stuff out of /dev and /dev/mapper worked
[17:01] <hallyn> some say evil, I say handy
[17:02] <patdk-wk> well, evil, cause it's just not a clean or nice way :)
[17:09] <smoser> kirkland, i just pushed to lp:ssh-import-id
[17:14] <macno> I used vmbuilder for the first time this afternoon, it's a great tool!
[17:14] <kirkland> smoser: sweet, thanks
[17:14] <smoser> where do you want to go from there ?
[17:15] <macno> I just have a couple of questions:
[17:15] <macno> 1) is it possibile (or what do I have to look at) to add also ipv6 address ?
[17:16] <robbiew> kirkland: zul: Daviey: SpamapS: JamesPage: hallyn: smoser: FYI.. RoAkSoAx will be joining us in Dallas next week ;)
[17:16] <zul> sweet!
[17:17] <kirkland> robbiew: \p/
[17:17] <kirkland> RoAkSoAx: welcome, dude!
[17:17] <RoAkSoAx> kirkland: thanks man :)
[17:18] <Daviey> robbiew, \o/
[17:18] <macno> 2) how can I set the disk type block ?
[17:22] <TREllis> RoAkSoAx: \o/
[17:23] <kirkland> hallyn: the memtest thing ... has that always been there, or is that a recent regression?
[17:24] <RoAkSoAx> kirkland: btw... got hold up with PowerNap dev... ended up not having a laptop the last couple weeks... and I guess next week I'll resolve some doubts
[17:24] <kirkland> RoAkSoAx: okay, I'll bring a wattmeter
[17:24] <kirkland> RoAkSoAx: and I'll have a few spare laptops
[17:25] <RoAkSoAx> kirkland: cool!!
[17:25] <azizLIGHTS> im getting some funny characters on putty via ssh?
[17:25] <azizLIGHTS>  ââcron
[17:26] <azizLIGHTS> instead of cron
[17:26] <azizLIGHTS> what to do
[17:26] <Datz> azizLIGHTS: try: settings->translation->UTF-8
[17:26] <azizLIGHTS> i dont have a gui
[17:26] <azizLIGHTS> are you talking about putty?
[17:26] <Datz> on putty you said
[17:27] <Datz> right click
[17:27] <Datz> then you'll need to refresh the screen
[17:30] <T3CHKOMMIE> hello all, im trying to change the default page loaded by apache2. i have /var/www/index.html and identity.html i want to keep both files but i want apache to defualt to identity.html before it looks for index.html, where do i configutre this? thanks for the help!
[17:32] <KurtKraut> T3CHKOMMIE, your question seems a bit odd for me. Could you explain me why you want to do this?
[17:32] <smoser> kirkland, did you want me to make a branch from lp:ubuntu/ssh-import-id ? and propose merge ? how are you expecting to keep the two related (lp:ssh-import-id and lp:ubuntu/ssh-import-id)
[17:32] <T3CHKOMMIE> my professor says so. honestly i would overwrite the index.html but they want me to change apache defualt settngs and i cant find it for the life of me.
[17:34] <kirkland> smoser: nah, i trust you
[17:34] <kirkland> smoser: you have commit access to that project/branch
[17:34] <smoser> kirkland, but not to the ubuntu
[17:34] <kirkland> smoser: i'll just need to sponsor/upload for you
[17:34] <kirkland> smoser: right
[17:34] <Datz> T3CHKOMMIE: probably a question for #http. But I think in apache config you can set it to default to something other than index.html
[17:34] <smoser> i already commited to "upstream"
[17:34] <kirkland> smoser: just keep the changes upstream
[17:34] <kirkland> smoser: perfect
[17:34] <kirkland> smoser: i want to do some testing, and i'll get it uploaded today
[17:35] <T3CHKOMMIE> Datz, do you know where that conf is? i have an empty https.conf file and if i ad
[17:35] <T3CHKOMMIE> "DefaultIndex identity.html"
[17:35] <T3CHKOMMIE> nothing changes.
[17:35] <KurtKraut> T3CHKOMMIE, Datz is right. You can even set the priority of default files, where you can set that identify.html should be serverd first even if there is a index.html, but this still a quite odd scenario. I can't figure out who would need such thing and for what purpose.
[17:36] <KurtKraut> T3CHKOMMIE, if your http.conf is blank Apache was installed incorrectly.
[17:36] <KurtKraut> T3CHKOMMIE, or you're looking for the wrong file in the wrong folder.
[17:36] <T3CHKOMMIE> crap....
[17:36] <T3CHKOMMIE> its the httpd.conf file in  /etc/apache2 right?
[17:36] <T3CHKOMMIE> or is there another i dont know about.
[17:36] <KurtKraut> T3CHKOMMIE, in Ubuntu, yes. In other Linux distributions it might be placed in different paths.
[17:37] <T3CHKOMMIE> dang, is there a way i can put what is supposed to be there?
[17:40] <Pici> No. /etc/apache2/httpd.conf is blank on normal Ubuntu apache2 installs.
[17:40] <cjwatson> RoyK: you won't generally find information about current development in lucid.  ssh-import-id was added in maverick
[17:40] <Pici> T3CHKOMMIE: We use a per-site configuration;. See /etc/apache2/sites-enabled/
[17:40] <T3CHKOMMIE> Pici,  thanks anything inparticular i should be looking for?
[17:41] <Datz> T3CHKOMMIE: you probably have all that info in apache2.conf.. athough I'm not finding what I was looking for either :p
[17:41] <Pici> T3CHKOMMIE: The sites you have enabled will have config files in that path
[17:42] <T3CHKOMMIE> ok
[17:43] <Pici> T3CHKOMMIE: You could even drop a single .htaccess file in /var/www/ if you wanted to just modify the behavior for that one path. Its up to you how you want to do it.
[17:44] <T3CHKOMMIE> i just want apache to load /var/www/ideneity.html when i got to localhost rather than /var/www/index.html, stupid i know but its a requiremnt for class :S
[17:44] <Datz> T3CHKOMMIE: it's in mods-available
[17:44] <Datz> not sure where though :P
[17:45] <Pici> What is in mods-available?
[17:45] <T3CHKOMMIE> mods avaiable.... ok lemme check that too. thanks Datz
[17:46] <Pici> mods-available isn't really relevant to this simple change.
[17:46] <shauno> I'd just rgrep DirectoryIndex in /etc/apache2.  Chances are all will become clear pretty quickly
[17:46] <Datz> T3CHKOMMIE: I think: dir.conf
[17:46] <Pici> T3CHKOMMIE: You probably have 000-default in /etc/apache2/sites-enabled/
[17:47] <Datz> Pici: ok..
[17:47] <T3CHKOMMIE> ok i changed /etc/apache2/sites-enabled/000-default and added DefaultIndex identity.html index.html under docuement /var/www and that seemed to work.
[17:47] <Pici> T3CHKOMMIE: Exactly :)
[17:47] <T3CHKOMMIE> shazam! thanks guys i uber appriciate it! thats taken everyone almost 4 horus in my class to figure out
[17:48] <T3CHKOMMIE> awsomness, now off to computer architecture! thanks again!
[18:24] <RoAkSoAx> smoser: where you able to test the fix for bug
[18:25] <RoAkSoAx> smoser: where you able to test the fix for bug #680138
[18:31] <azizLIGHTS> im installing 10.10 server
[18:32] <azizLIGHTS> why does vmware say when you are done and the operating system boots up, click i finished installing
[18:32] <azizLIGHTS> why
[18:32] <azizLIGHTS> what will it do
[18:33] <Datz> probably a question for #vmware
[18:33] <azizLIGHTS> sorry
[18:33] <Datz> np, (I don't know the answer) :P
[18:38] <smoser> RoAkSoAx, i suggested an alternative solution.
[18:41] <RoAkSoAx> smoser: awesome! Thanks!
[18:51] <hallyn> kirkland: i don't think 0.12.5 has that memtest problem.  do you have a maverick box to check on?
[18:55] <RoAkSoAx> smoser: Ok, I don't think that will work because that cmd will be splitted later to be used by a subprocess in the GTK, and as far as I could remember having that type of commands made subprocess to fail
[18:55] <RoAkSoAx> i'm testing it now though
[18:59] <smoser> i dont know what you're doing in the gtk. it works from the commanad line.
[19:14] <RoAkSoAx> smoser: ok. It indeed works well with the command line, but fails with the GTK, because of subprocess. Will work on it to provide a better solution
[19:15] <smoser> well, you have a general issue if you're consuming the response from that functoin in two different ways
[19:19] <smoser> RoAkSoAx, see subprocess.Popen(shell=True)
[19:19] <smoser> http://docs.python.org/library/subprocess.html
[19:20] <smoser> i'd ditch the 'split' and just use that.
[19:21] <RoAkSoAx> smoser: well the difference is that in the command line it only runs as "os.system(cmd)", while in the gtk it uses a subprocess to run cmd, specially because to obtain the progress and stuff
[19:21] <RoAkSoAx> smoser: and as far as I can remember, when I first did it, when using shell=True, I couldn't grab the progress. But will give it a try since that part has been changed
[19:22] <smoser> i just tested this, and it shows me progress in the gui
[19:22] <smoser> http://paste.ubuntu.com/550775/
[19:23] <RoAkSoAx> smoser: ok cool then!
[19:25] <RoAkSoAx> smoser: works great, thank you!!
[20:33] <deadsmith> trying to remake the ISO for amd64-server; can I replace /boot/grub/efi.img with a new grub.efi ?  Or is this not even the boot file for EFI systems?
[20:34] <deadsmith> shorter question:  does anyone know the mkisofs flags used to make the amd64-server ISO ?
[20:42] <deadsmith> n/mind... found it.
[20:46] <CheetoBandito> I'm having an interesting issue in 10.04... If I request packets from a specific group in one program, and a different group in another program... I get packets from both groups in both programs.
[20:47] <CheetoBandito> If I compile and run the same scenarion on Windows, Solaris, and other distros I get the intended behavior, that the packets each program sees are only the ones they requested.
[20:57] <kirkland> Daviey: around?
[20:57] <kirkland> Daviey: would you mind dropping me a few instructions as to setting up that video chat flash thing we used yesterday
[20:58] <jdstrand> hallyn: hey, so today I seem to be hitting bug #694029 pretty regularly, with 0.12.5+noroms-0ubuntu7 on natty
[20:58] <jdstrand> hallyn: how do I disable ksm?
[20:58] <jdstrand> kirkland: hey, you may actually know that one since you implemented it in Ubuntu iirc ^
[20:59] <kirkland> jdstrand: edit /etc/default/qemu-kvm
[21:00] <jdstrand> easy enough
[21:00] <jdstrand> kirkland: thanks!
[21:00] <kirkland> jdstrand: if that fixes your problems, try enabling it, but adjusting the sleep milliseconds
[21:00] <jdstrand> kirkland: this bug really stinks :(
[21:00] <kirkland> jdstrand: there have been complaints that the default sleep (20ms) is too aggressive
[21:00] <jdstrand> kirkland: will do
[21:00] <kirkland> jdstrand: i don't disagree, but i also haven't seen a solid suggestions as to what a sane default would be
[21:00] <jdstrand> kirkland: would the aggressive default cause flaky guest behavior?
[21:01] <kirkland> jdstrand: if someone told me it did, i'd be inclined to belive them
[21:01] <jdstrand> kirkland: oh, I am not suggesting anything. it is just been very annoying for me. I only just started looking at ksm
[21:01] <kirkland> jdstrand: but i've not heard that yet
[21:01] <jdstrand> s/would/could/
[21:02] <jdstrand> (and therefore have no opinions)
[21:08] <jdstrand> kirkland: ^
[21:08] <kirkland> jdstrand: okay
[21:08] <kirkland> jdstrand: i don't know of any issues with ksm causing instability
[21:08] <jdstrand> kirkland: I've only started hitting this bug lately, and have never had a problem with the default
[21:09] <jdstrand> kirkland: well, by 'lately' I mean with the natty kernel
[21:09] <jdstrand> which has probably been a month and a half (used the maverick kernel on natty for quite a while before that)
[21:11] <jdstrand> disabling ksm didn't help
[21:15] <deadsmith> still no successful EFI boot.  Does anyone know where there's a description of the files on the ISO?
[21:15] <deadsmith> or otherwise have any good ideas about what I should be reading to make a custom EFI install ISO?
[21:20] <cole> deadsmith: did you see this? http://tinyurl.com/24alt9k
[21:21] <cole> deadsmith: didn't scroll up, just saw you had trouble booting via efi
[21:22] <deadsmith> cole: no worries, let me look this over and see if it mentions something I've missed...
[21:22] <deadsmith> cole:  I have a working grub2 for EFI, I'm just trying to get it on the bleeding CD
[21:22] <cole> ah!
[21:24] <deadsmith> cole:  I was able to install the system by bootstrap-installing with rEFIt and an OSX partition (it's an Xserve)....
[21:24] <deadsmith> cole:  but I'm trying to get a reuasable install image..
[21:25] <cole> deadsmith: i assume you used rsync --one-file-system --exclude /proc ??
[21:25] <deadsmith> cole:  no... at which step?  making the ISO?  I have the amd64-server loopback mounted on a linux machne and I'm using mkisofs to remaster it...
[21:39] <jdstrand> hallyn: summary of bug #694029. qemu-kvm 0.12.5+noroms-0ubuntu7, using an ide disk and disabling ksm made no difference
[21:49] <hallyn> jdstrand: yeah, so basically any qemu on natty kernel is bad?
[21:49] <jdstrand> hallyn: bottom line, yes
[21:49] <jdstrand> I'm back to a maverick kernel
[21:49] <hallyn> I've yet to reproduce it, but wil try with your suggestion about maverick update in one, and (wahtever) in another kvm
[21:49] <jdstrand> (I have to actually get work done)
[21:50] <jdstrand> hallyn: what chipset do you have? I'm using i7 here
[21:50] <hallyn> yeah, i have core i7 laptop
[21:57] <jdstrand> hallyn: it is hard to find a simple reproducer.... :\
[21:59] <jdstrand> hallyn: not sure it would make any difference, but I am using qcow2 files, the one upgrading is a snapshot qcow2
[22:00] <jdstrand> hallyn: the one installing ubuntu desktop is qcow2, but not snapshotted yet
[22:01] <jdstrand> I have to believe it is going to be a more widespread problem when people upgrade...
[22:05] <hallyn> jdstrand: how do you install ubuntu minimal anyway?  i think i'll try apt-get install ubutnu-desktop from ubutnu server?
[22:06] <jdstrand> hallyn: I use vmbuilder
[22:07] <hallyn> jdstrand: have all of these been built using vmbuilder?
[22:07] <jdstrand> hallyn: yes
[22:07] <hallyn> hm.  potentially relevant
[22:08] <hallyn> jdstrand: thanks, i'm just gonna play a lot, i'll stop bugging you :)
[22:08] <jdstrand> hallyn: these are the packages I install after using vmbuilder: screen ubuntu-desktop vim openoffice.org
[22:09] <jdstrand> hallyn: the precise method I use to generate VMs is in https://wiki.ubuntu.com/SecurityTeam/TestingEnvironment
[22:10] <hallyn> kthx.  after an attempt with installs from iso, i'll try from vmbuilder
[22:11] <jdstrand> hallyn: sure! thanks for looking at it
[22:12] <jdstrand> it has been a real hair-puller for me
[23:35] <Hilikus> hey guys
[23:35] <Hilikus> can someone help me configure alsa for 5.1. i try the sound test and i only hear front left and right. i don't use pulse audio or gnome