[01:03] <bofh80> hi, can any one tell me what the "Basic Ubuntu Server" installs when i do a minimal install it is in the Software Select / tasks list with tomcat LAMP server etc
[01:07] <twb> bofh80: that list is the "tasksel" list.  If you dig out its data files, it should be clear.
[01:13] <twb> bofh80: AFAICT it only installs ubuntu-serverguide
[01:18] <bofh80> thanks twb
[01:37] <smoser> kirkland, how long does it take to build euca ?
[01:37] <kirkland> smoser: 3 minutes or so
[01:38] <kirkland> smoser: euca2ools even less
[01:38] <smoser> oh. wow.
[01:38] <smoser> i expected much longer.
[01:38] <smoser> because i think i have a fix for ramdisk issue (once you turn off the default ramdisk)
[02:20] <MTecknology> is 212 deg F really hot for a laptop?
[02:20] <MTecknology> :P
[02:24] <twb> MTecknology: what is that in a useful scale?
[02:24] <MTecknology> twb: 100 C
[02:24] <smoser> kirkland, bug 526805 , bug 525989, bug 525994 now have accurate comments/state (afaik) as to what we've found today
[02:25] <twb> MTecknology: if that's the CPU temperature, then yes, that's probably too hot.
[02:25] <smoser> now i'm looking into fixing 526805, which is the root issue.  i *thought* i had a fix, but i must have messed up somehow.
[02:25] <kirkland> smoser: cool, did you add the bit about the default kernel/ramdisk in UI?
[02:25] <smoser> yeah.
[02:25] <smoser> i guess you can't change that in euca_conf
[02:25] <MTecknology> twb:
[02:26] <MTecknology> twb: I'm thinking of killing the build and passing it onto launchpad instead
[02:27] <twb> MTecknology: I don't think launchpad will be able to fix hardware faults in your laptop
[02:28] <MTecknology> twb: building a deb package
[02:31] <smoser> kirkland, i really have no idea how to solve this.
[02:31] <smoser> for alpha3, realistically, the only thing i can think of is using ramdisk.
[02:31] <kirkland> smoser: okay, let's do that for A3
[02:31] <kirkland> smoser: and we need to talk to eucalyptus upstream for a solution
[02:32] <kirkland> smoser: did you try zero'ing out those two fields in the GUI, and then registering an image without a ramdisk?
[02:38] <smoser> yeah, thats where bug 526805 comes from
[02:39] <smoser> you do, then get a ramdiskless boot, but the lucid image hangs on boot
[03:05] <Elad> how do you view all the updates without installing them?
[03:06] <twb> aptitude -sy full-upgrade
[03:13] <Nonpython> I installed postfix by the guide and it does not respond to telnet, what's wrong?
[03:24] <d4n1> hey, what is a good ftp server i can use on my server?
[03:27] <d4n1> http://packages.ubuntu.com/hardy/ftp-server, which is the best, or which do u recommend?
[03:34] <qman__> d4n1, I recommend not using an FTP server, use SFTP instead if at all possible
[03:35] <qman__> but, if you absolutely must have FTP, I would prefer VSFTPD
[03:35] <d4n1> ok
[03:36] <d4n1> im "trying" to install vsftpd
[03:36] <d4n1> and what would be a good SFTP? does it still run on port 22?
[03:36] <qman__> SFTP is built into openssh-server
[03:36] <d4n1> i have openssh-server
[03:37] <qman__> if you need jails, it can be configured in the sshd-config
[03:37] <qman__> if not, it's already all set up
[03:37] <qman__> just add system users
[03:37] <d4n1> will it run like an ftp server on port 22?
[03:37] <qman__> no, it runs an SFTP server, on port 22
[03:38] <d4n1> ok
[03:38] <d4n1> hmm..
[03:38] <qman__> you need a client that can do SFTP, such as sftp, gftp (for linux) winSCP or filezilla (for windows)
[03:38] <d4n1> ok, sftp, ill install it, thank u
[03:38] <qman__> 'sftp' is built into the basic package set
[03:39] <qman__> it's a command line utility
[03:39] <d4n1> ok, why don't we pm
[03:41] <twb> PuTTY also has an SFTP client, IIRC.
[03:42] <twb> d4n1: at the protocol level, SFTP and FTP are completely different.  It's not just FTP-over-SSH.
[03:43] <d4n1> ok, then if i need an ftp server, id better go for vsftp twb??
[03:43] <twb> SFTP is better than vsftp, simply because SFTP isn't FTP.
[03:43] <twb> But if you needed FTP, vsftp is what I'd recommend, because it has a security focus.
[03:43] <d4n1> ok
[03:43] <qman__> using FTP is like using telnet
[03:43] <qman__> it's archaic, insecure, and limited
[03:43] <qman__> SFTP is better in every way
[03:44] <d4n1> ok
[03:45] <twb> http://mywiki.wooledge.org/FtpMustDie
[03:46] <qman__> nice link
[03:46] <d4n1> ok, then forget about ftp
[03:49] <qman__> SFTP operates entirely on one port, with encryption, securely
[03:50] <d4n1> ok, ill take a look at my sshd_conf file
[03:51] <qman__> you don't need to do anything special to set up SFTP, it's on by default
[03:51] <qman__> you only need to make changes if you want certain users to have SFTP-only access, or to use jails
[03:51] <d4n1> really, then i must have something else wrong
[03:51] <d4n1> thank u qman__ and twb
[04:03] <maxagaz> My dhcp server doesn't start at boot, it's checked in sysv-rc-conf (2,3,4,5), there's nothing in /var/log/syslog until I start it manually, what else can I check ?
[04:05] <bobsomebody> hello
[04:08] <cef> maxagaz: check the /etc/default/ directory
[04:11] <twb> maxagaz: which dhcpd implementation?
[04:15] <bogeyd6> maxagaz, we need to know your dhcpd package, is it bind9?
[04:15] <maxagaz> twb, isc-dhcpd-V3.0.5 with a patch for ldap
[04:15] <maxagaz> it works fine on other servers
[04:16] <bogeyd6> did you manually check the directories for the script and make sure its executable?
[04:16] <maxagaz> bogeyd6, it works when I run it manually
[04:16] <bogeyd6> maxagaz, is it in say /etc/rc5.d/
[04:17] <maxagaz> bogeyd6, I have /etc/rc5.d/S20dhcp-server
[04:17] <bogeyd6> maxagaz, also you can try to re-run sudo update-rc.d name_of_the_script defaults
[04:18] <bogeyd6> maxagaz, ls -l /etc/rc5.d/S20dhcp*
[04:18] <maxagaz> update-rc.d dhcp-server defaults
[04:18] <maxagaz>  System startup links for /etc/init.d/dhcp-server already exist.
[04:18] <maxagaz> bogeyd6, lrwxrwxrwx 1 root root 21 2010-02-22 11:20 /etc/rc5.d/S20dhcp-server -> ../init.d/dhcp-server
[04:18] <bogeyd6> then it sounds like you got a bonafide bug.
[04:19] <bogeyd6> !bug | maxagaz
[04:21] <bogeyd6> maxagaz, also there is this: http://ubuntuforums.org/showthread.php?t=545300
[04:21] <bogeyd6> pay particular attention to the last sentence of the last post
[04:23] <cef> err, ubuntu uses runlevel 2 by default, not 5
[04:24] <bogeyd6> cef, just an example, but running the defaults show its still set, same dif
[04:24] <twb> It also doesn't distinguish between runlevels 2 through 5 by default
[04:25] <cef> it says links exist, it doesn't tell you WHICH ones exist
[04:25] <maxagaz> what I don't undertand is that I have nothing in the logs
[04:26] <maxagaz> not even a message saying that it tried to run the dhcp-server
[04:26] <bogeyd6> maxagaz, its probably not even trying to run the script
[04:26] <cef> update-rc.d will not verify that system startup links are correct, and by default will not change them\
[04:26] <maxagaz> bogeyd6, exactly
[04:26] <cef> maxagaz: 'ls -l /etc/init.d/dhcp-server' ?
[04:26] <bogeyd6> cef, you are wasting time bro.
[04:27] <bogeyd6> cef, update-rc.d updates the System V style init script links /etc/rcrunlevel.d/NNname whose target is the script /etc/init.d/name
[04:27] <maxagaz> cef, -rwxr-xr-x 1 root root 2887 2009-11-12 14:53 /etc/init.d/dhcp-server
[04:28] <cef> maxagaz: you built from source, correct? can you pastebin that file?
[04:28] <maxagaz> cef, yes, I built it from source
[04:28] <maxagaz> cef, it used to work until last week
[04:28] <cef> (as in /etc/init.d/dhcp-server )
[04:29] <maxagaz> cef, paste which file ?
[04:29] <maxagaz> /etc/init.d/dhcp-server ?
[04:29] <cef> yes
[04:31] <cef> bogeyd6: read the man page for update-rc.d "bro"..  If  any  files  /etc/rcrunlevel.d/[SK]??name already exist then update-rc.d does nothing.  The program was written this way so  that  it  will never  change an existing configuration, which may have been customized by the system administrator.
[04:33] <maxagaz> cef, http://pastebin.com/Y3yg20ZJ
[04:36] <cef> maxagaz: when you start it after boot, what command are you running?
[04:36] <maxagaz> cef, /etc/init.d/dhcp-server start
[04:37] <cef> maxagaz: as root/via sudo I'm assuming?
[04:38] <maxagaz> cef, yes
[04:38] <maxagaz> cef, and it works
[04:39] <cef> maxagaz: yeah.. ok my guess is that it depends on something that isn't running yet. so pastebin the output of 'ls -1 /etc/rc2.d/' pls
[04:40] <cef> btw: on Hardy, dhcp3-server starts at S40
[04:44] <cef> yup.. starting at S20 instead of later.. looks like it needs hal to be started first (well, it makes the most sense)...
[04:47] <cef> so, try 'mv /etc/rc2.d/S20dhcp-server /etc/rc2.d/S40dhcp-server' and see if that helps
[04:49] <cef> that'll just change that one script to start later in the default runlevel (2). to do that with update-rc.d, you could use 'update-rc.d dhcp-server multiuser 40', but that wont work if any of the /etc/rc?.d/ links exist
[04:52] <cef> maxagaz: if it still doesn't work, you could always add an 'echo "starting" > /root/dhcp-test' into that script to see if it's even calling it, getting to certain points, etc.
[04:56] <Omr> Is there a way to check whats running in my server?
[04:57] <smoser> good night all
[04:57] <Omr> night
[04:57] <zroysch> Omr: yes
[04:58] <Omr> Can you give me a hint what to google for?
[04:58] <zroysch> ps aux
[04:58] <zroysch> top
[04:59] <Omr> thanks
[05:05] <Omr> ok is there a way to see what one is writing to the hard drive 24/7?
[05:12] <cef> top will show you if something is touching the disk a lot
[05:13] <cef> actually no it won't, cos it doesn't show anything at all relating to disk access. :(
[05:14] <Omr> yeah ive run most commands i found but no disk info.  Something wont stop using my disk and it keeps shutting down due to overheating
[05:15] <Omr> i could turn off the safty in bios but then my disk would probably die.  The safty is set at 75 degrees celcius
[05:15] <cef> the pc shuts down?
[05:16] <Omr> Yeah bios turns it off
[05:16] <cef> ok, that's the cpu overheating then, not so much the drive
[05:16] <Omr> it was the servers first night on last night it was off in the morning and its turned off 2 times tonight
[05:16] <Omr> No the bios says the drive is to hot
[05:16] <cef> ahh server bios? ok
[05:17] <Omr> Its a laptop FMV Biblo the bios moniters the cpu gpu and hdd temp
[05:19] <Omr> when the laptop is turned back on a screen informs me it shut down due to sensor 3 over temperature
[05:19] <cef> lsof will tell you everything that is open, but that'll be a big list
[05:20] <Omr> i ran sudo apt-get remove apache2 but apache2 is still running with 7 processes using 4% cpu and 32 - 40 mb ram?
[05:21] <Omr> but i doubt apache would use the hdd anyway
[05:21] <Omr> but thats my top on ps -aux | less
[05:22] <cef> just  'killall apache2' (or whatever the process name is)
[05:22] <Omr> thanks
[05:23] <cef> could simply be all the logging, especially if it's a public webserver and a spider has found it and is indexing the thing.
[05:29] <Omr> Its not public and isn't exposed to the internet atm
[05:30] <Omr> but i just realised im an idiot
[05:30] <Omr> ever since i installed it a program keeps posting notes about reloading config files and errors and other random crap id assume thats probably my culprit
[05:30] <Omr> i just gotta wait for another message
[05:39] <maxagaz> cef, it worked! :)
[05:40] <maxagaz> cef, I mean changing the priority to S40
[05:41] <maxagaz> cef, why did you think that hal should be mounted first ?
[07:07] <cef> maxagaz: I have absolutely no idea.. I just looked at the starting priorities on the dhcp3 server here that runs the standard package. could be hal is doing something to the devices that dhcp3 needs, or it could be a completely different service
[07:16] <cef> maxagaz: the only services that match between the machine I have on hand and yours in /etc/rc2.d that lies between S20 and S40 are hal, bluetooth, pulseaudio and gdm. of those, hal is about the only one that makes any sense.
[07:19] <cef> ok, off to dinner and home. cyas
[07:48] <soren> ttx: 23:14 < soren> Am I supposed to close the server-papercut bug task when I fix something or do I leave it open until the meeting so that everyone can join in and rejoice?
[07:49] <ttx> no, just close it :)
[07:49] <ttx> soren: ^
[07:50] <soren> ttx: Will do. Thanks.
[07:52] <soren> ttx: http://surl.dk/77e/ <--- It's going pretty well this week. All those fix committed once are since around noon Monday, I think.
[07:52] <ttx> soren: cool
[07:52] <ttx> soren: do you need FFe for anything ?
[07:53] <soren> ttx: Not right now.
[07:53] <soren> ttx: I /may/ need one later, but at the moment, it's all bugfixes.
[07:58] <Sergiu24>  hi. Where can i find mysql-server 4.0.27.deb i386?  help pls!
[07:58] <Sergiu24> any1?
[07:59] <soren> Sergiu24: why do you need that specific one?
[07:59] <Sergiu24> i have a program that only runs on mysql 4
[08:02] <soren> Ubuntu has, as far as I know, /never/ shipped MySQL 4.0.
[08:02] <soren> Dapper (which is 4 years old) shipped 4.1
[08:04] <soren> Sorry, my mistake. Warty had 4.0.20.
[08:06] <soren> ..and hoary 4.0.23.
[08:06] <Sergiu24> good
[08:07] <Sergiu24> where can i download that :)
[08:07] <soren> But those are full of security holes and all sorts of other problems.
[08:07] <Sergiu24> i need the .deb package
[08:07] <soren> Seriously, that stuff is /ancient/.
[08:07] <Sergiu24> i know... f8ued up
[08:07] <soren> You should at least see if you can use 4.1.
[08:07] <soren> Dapper has 4.1 and is still in support.
[08:07] <Sergiu24> i spent 40k $ on a damn program
[08:07] <Sergiu24> and it's made on mysql 4 platform
[08:08] <soren> How did you come up with "4.0.27.deb"?
[08:08] <Sergiu24> this version was installed before
[08:08] <Sergiu24> hardware problems and now i can't find that version for reinstall
[08:09] <Sergiu24> if i update to mysql 5 there is a problem with the connectors
[08:09] <Sergiu24> and program crushes
[08:12] <soren> Try 4.1 from Dapper.
[08:21] <soren> Sergiu24: Oh, Dapper actually had 4.0 as well. My bad. Apparantly, the naming scheme changed with 4.1, so I missed it when I first looked.
[08:23] <soren> Sergiu24: So, you can at least run an OS that is supported, with an unsupported MySQL.
[08:29] <Sergiu24> yea
[08:29] <Sergiu24> will try thanks
[08:34] <ivoks> this pacemaker stuff is trully awesome.
[08:48] <Steel__> hi @ all
[08:54] <jiboumans> morning folks
[08:58] <_ruben> ivoks: care to elaborate? looking into pacemaker & co is still down there on my todo list
[08:59] <ivoks> _ruben: clustered lvm, clustered filesystems - both work
[09:00] <_ruben> nice
[09:00] <ivoks> _ruben: drbd management, services have worked before
[09:01] <_ruben> ivoks: tried drbd mc (i think that's what it called)? or all configuration done "by hand" ?
[09:01] <ivoks> i have a setup where i have drbd shared storage, on top of which is LVM, and on one of LVs is GFS2
[09:02] <ivoks> pacemaker makes sure services are started in proper order, including distributed lock management, drbd master/master promotion, clvm and filesystem mounting
[09:03] <ivoks> after everything is up, apache and vsftpd are started (but before services are started, additional IPs are up on interfaces, so that apache and vsftpd bind to those)
[09:03] <ivoks> when one node fails, the other one takes over it's ip and service
[09:03] <_ruben> im still using the classic heartbeat v1 stuff for my HA setups .. replicated storage is something i'll have to look into in the not too distant future
[09:06] <ivoks> even this works:
[09:06] <ivoks> http://wiki.samba.org/index.php/Ping_pong
[09:07] <ivoks> and since samba in lucid will have support for ctdb, we are looking at the awesome additions for server in ubuntu
[09:08] <ivoks> only thing left is to convince everybody that MIRs should be accepted :D
[09:08] <ivoks> good news is that pacemaker is building as we speak, meaning rest of the MIRs could be filled in couple of hours
[09:18] <lenios> does anybody know a way to check /boot partition integrity once the system (on an encrypted FS) is started?
[09:33] <incorrect> I was thinking about installing asterisk, when i google for ubuntu asterisk i just see people compiling from source, is there any reason not to use the package?
[09:34] <ivoks> afaik, there's an intention to make asterisk shine in lucid
[09:35] <ivoks> i'm don't know what's the status with it
[09:39] <persia> There's been good progress based on the VOIP team uploads, but based on the UDS discussion, I'm not convinced karmic was that shiny
[09:40] <Daviey> persia: shiny?
[09:40] <rc55> Could someone list the best root directories to backup for a full system backup? I'm doing /home at the moment, but assuming I'll need /var and /opt, any others?
[09:40] <persia> Daviey: All rough edges buffed until there everything is even enough to provide significant reflection.
[09:40] <persia> s/there everything/everything there/
[09:43] <lenios> rc55, you might need /etc if you changed anything there
[09:43] <andol> rc55: You probably want to backup /etc as well. It will take basically no storage space in comparraions, and there's always some fancy configuration in there you might want to peak at aftewards.
[09:43] <Daviey> persia: heh. the individual packages are actually pretty shiny - a couple of issues include it being an RC release that upstream were a little unable about (they wanted a stable), the inteion being to track the release through the cycle - but for no obvious reason there was a largely delay in them releasing stable than ever before by the seems of it.  Also it's a different version from debian making the support slightly harder.
[09:44] <Daviey> persia: Also, Lucid is a "fun" release - as none of upstreams releases will be supported for the length of lucid.
[09:44] <Daviey> (by upstream)
[09:44] <rc55> lenios / andol: Thank you both, that's very useful. :)
[09:46] <Daviey> persia: We did make really good progress in lucid, with quite a few changes.  One thing that made karmic difficult is -server seed not installing the headers, so dkms does when installed.  However, lack of slippy depends means dkms usually installs the wrong headers for server.  Which means many people may have installation problems, but that is really something that was near impossible for us to resolve.
[09:50] <Pupeno> How do I set the domain (for resolv.conf) in /etc/network/interfaces?
[09:52] <persia> Daviey: Do we have headers installed in lucid to make this work?
[09:55]  * Daviey checks the seed
[09:55] <ghostlines> hi all
[09:56] <ghostlines> how do i get convirt to load my kvm virtual machines?
[09:56] <ghostlines> it's not loading my xml config files
[09:56] <ghostlines> any help is appreciated
[09:57] <soren> What are these xml files?
[09:57] <incorrect> ghostlines, did you define them?
[09:57] <ghostlines> the config files that kvm uses for the virtual machines
[09:57] <soren> kvm does not use config files for virtual machines. libvirt does.
[09:58] <soren> and convirt does not - unless something changed recently - use libvirt.
[09:58] <incorrect> ghostlines, in virsh define /etc/libvirt/qemu/something.xml ?
[09:58] <ghostlines> ohh my mistake
[09:59] <ghostlines> i have vm's running already and i use virt-manager to run them
[09:59] <ghostlines> *to manage them
[09:59] <ghostlines> but i wonder if convirt can be used to manage them also
[09:59] <ghostlines> and if so how would i go about adding existing vm's to convirt to be able to manage them
[10:00] <ghostlines> i saw an option to load config files so i chose the libvirt config files but to no avail
[10:03] <Daviey> persia: hmm, it seems to be in server-ship ok.. But if someone wants to install it on desktop (perhaps for testing), it will still crapout.  It's really a problem with dkms installing the first dependency linux-header rather than the headers of the current linux kernel.
[10:04] <persia> Daviey: Ah, this likely affects users of -preempt as well.  Maybe dkms could be extended to check if the available headers match the *running* kernel, and if not, attempt to install the right headers package?
[10:06] <Daviey> persia: is it debian policy compliant to apt-get a package in postinst?
[10:06] <soren> Daviey: no
[10:07]  * Daviey wouldn't have thought so.
[10:07] <soren> Daviey: Why would you want to?
[10:07] <persia> soren: Because dkms fails if there exists more than one flavour of the kernel per architecture right now.
[10:08]  * soren facepalms
[10:09] <persia> so, e.g., amd64, which has three flavours, just plain doesn't work for anything that needs modules except if the default kernel flavour (which is the boring one) is installed.
[10:09] <persia> powerpc has issues on PS3s
[10:09] <persia> armel is just broken for dkms
[10:09] <persia> Other architectures should be fine.
[10:09] <Daviey> this has been a long term issue, that has been avoided by luck IMO.  I heard the best solution was to wait for "slippy depends", so it's satisified at package install time, except i've not heard anything new on this.
[10:10] <acalvo> anyone using squid with windows machines and got running windows update thru it?
[10:10] <Daviey> acalvo: I assume it's not working for you?
[10:10] <persia> "slippy depends" isn't likely to happen any time soon: it requires additional semantics in a host of packages.
[10:11] <acalvo> Daviey: yes, it's not working here
[10:12] <Daviey> persia: I'm not sure i can think of a clean solution, other than suggesting to dkms a cleaner error message, with a suggestion of how to fix it.
[10:12] <Daviey> which is still rubbish.
[10:15] <persia> Daviey: Or make dkms depend on all headers (which installs extra stuff but avoids the pain)
[10:15] <Daviey> i suspect that would upset some people
[10:16] <Daviey> ISTR an upset about -headers being on a server anyway.
[10:38] <seet> I just bootet the eu-west-1 ami and the server logs looks fine. Everything is startet and my default security group allows all connections, but i get "Connection timed out
[10:38] <seet> "
[10:39] <seet> when trying to ssh to the instance
[10:39] <seet> is the ssh on another port that default?
[10:39] <seet> than
[10:41] <seet> ami-2fc2e95b
[10:52] <seet> Never mind. The security groups seems to be isolated to regions as well
[11:13] <soren> Woo!
[11:13]  * soren just fixed bug 518742
[11:13] <soren> That one's been annoying me for bloody ages.
[11:14] <soren> \sh: ^^
[11:19] <Roxyhart0> hi there, i need to add one aleas for my server which is the same dns server, how i can do that, i try with files.rx.com CNAME ns but doesn't work
[11:22] <Roxyhart0> where i should add the files with the new "configuration"?
[11:31] <Roxyhart0> somebody know use bind?
[11:31] <Ymer[Draggo]> Hello every, if I could pick ur brains about apparmor that'd be awesome. What we would like to do is allow all HTML on a webserver  and block all other (PHP, javascript etc), though since me and my friend are totally green when it comes to apparmor, we could really use some advice on how to do this. Thanks alot in advance.
[11:32] <franjpr> can i upgrade ubuntu server 9.04 to 9.10 with the alternate cd
[11:33] <bogeyd6> cef, how did maxagaz with the update-rc.d problem work out?
[11:34] <maxagaz> bogeyd6, I change S20 into S40
[11:34] <maxagaz> changed
[11:34] <bogeyd6> nice :)
[11:35] <maxagaz> bogeyd6, but I don't really understand what was happening
[11:35] <bogeyd6> you had two s20's?
[11:35] <maxagaz> bogeyd6, what do you mean ?
[11:35] <bogeyd6> i dont even know why chaning it to s40 would fix it
[11:36] <Silent> Can anyone maybe point me to a guide on how to setup my squid proxy reports so that it mails the reports to me
[11:37] <bogeyd6> Silent, you are looking for SARG?
[11:38] <bogeyd6> Silent, http://www.squid-cache.org/Scripts/
[11:38] <Silent> Ye tho i cant get my sarg setup so that it mails the generated reports
[11:38] <Silent> ah thnx a million
[11:39] <Roxyhart0> hi  sombody know how and where i need to add a entry myhost.rx.com which is the same ip from the dns?
[12:55] <\sh> soren: wooot
[12:56] <\sh> soren: you're rockstar :)
[13:02] <soren> \sh: :)
[13:02]  * soren goes to lunch
[13:03] <mrbrdo> any advice on how I could limit global upload speed at my ubuntu server (basically a NAT for my home LAN)?
[13:03] <mrbrdo> because when i max out my upload my download suffers (DSL line)
[13:12] <skwashd> hi all
[13:13] <skwashd> i am currently planning some lamp projects which will be deployed on lucid and i noticed that there doesn't seem to be a decision on including php5.3 in lucid
[13:13] <skwashd> have a missed something where a decision has been made on this?
[13:23] <zul> skwashd: im hoping to get a FFE post alpha-3
[13:24] <skwashd> zul: excellent!
[13:25] <zul> skwashd: but if you want to play with it https://edge.launchpad.net/~zulcss/+archive/php5.3-lucid
[13:25] <alkisg> Is there any command line tool similar to gdebi, that does dependency resolution while installing some .deb files from a local dir?
[13:25] <soren> alkisg: yes. gdebi.
[13:25] <skwashd> zul: no ... just wanting to plan for building our own debs or not for a couple of years :)
[13:26] <skwashd> thanks for the info
[13:26] <alkisg> soren: well... command line == one that doesn't depend on gtk or X installed :-/
[13:27] <alkisg> I.e. one that could be used in a minimal environment...
[13:27] <soren> alkisg: Yes. gdebi.
[13:27] <soren> alkisg: The package is gdebi-core.
[13:27] <soren> alkisg: The binary is gdebi.
[13:27] <alkisg> soren: thanks, let me look into it...
[13:27] <skwashd> alkisg: $ gdebi
[13:27] <skwashd> Usage: gdebi [options] filename
[13:27] <skwashd> For a graphical version run gdebi-gtk
[13:27] <skwashd> [...]
[13:28] <alkisg> I was looking at the gdebi package dependencies :) Thanks a lot, guys!
[13:28] <mrbrdo> is there any good QoS guides for a home ubuntu router?
[13:30] <skwashd> zul: thanks again ... cya
[13:36] <ttx> smoser: ping
[13:38] <Pupeno> Any ideas what is putting this on motd:
[13:38] <Pupeno> 0 packages can be updated.
[13:38] <Pupeno> 0 updates are security updates.
[13:38] <Pupeno> ?
[13:38] <zul> ttx: when you get a chance today can you have a look at the "New" eucalyptus bugs on http://qa.ubuntu.com/reports/ubuntu-server-team/dailynewbugs.ubuntu-server.2010-02-23.html please?
[13:38] <ttx> Pupeno: /etc/update-motd.d/90-updates-available ?
[13:39] <ttx> zul: ok
[13:40] <ttx> zul: I've to do my triageday sometime today anyway
[13:40] <zul> ttx: people still do that? ;)
[13:40] <Pupeno> ttx: thanks.
[13:40] <ttx> zul: they should.
[13:40] <ttx> zul: some people might get angry if they don't.
[13:41] <zul> ttx: heh
[13:44] <Pupeno> is there an ubuntu-server package like there's an ubuntu-desktop one?
[13:45] <pmatulis> Pupeno: no, just install with the Server ISO
[13:48] <Pupeno> pmatulis: I like maintaining my dependencies clean ;)
[13:49] <kirkland> ttx: morning
[13:50] <ttx> kirkland: morning
[13:50] <ttx> kirkland: I think things are mostly in order now, I lowered the level to defcon3
[13:51] <pmatulis> Pupeno: what do you mean?  just install from the image
[13:51]  * kirkland hugs ttx 
[13:51] <kirkland> ttx: smoser's fix worked, you committed and uploaded i see
[13:51] <ttx> kirkland: yes.
[13:51] <kirkland> ttx: talk about a deep, buried issue
[13:51] <kirkland> ttx: it took some mad debugging to get to the bottom of this
[13:51] <ttx> pushed to ISO and a3 candidate
[13:51] <Pupeno> pmatulis: I keep a very short list of packaged marked as non-auto-installed, and the rest as auto. On a desktop, ubuntu-desktop covers me for the basic packages.
[13:51] <ttx> validated on amd64 a few minutes ago
[13:52] <kirkland> ttx: i'm syncing now
[13:53] <kirkland> ttx: i'll start my local tests during the IRC meeting; do the remote ones in the lab thereafter
[13:53] <ttx> kirkland: so it's still buggy (scary boot messages booting a ramdiskless image with a wrong "default ramdisk")
[13:53] <ttx> kirkland: but it works well enough for a3
[13:53] <kirkland> ttx: right
[13:53] <ttx> I targeted the "default ramdisk" bug to beta1
[13:54] <ttx> we'll have to discuss it with euca guys, I fail to see any point in that "feature"
[13:54] <smoser> ttx, thanks for testing and integrating.
[13:54] <ttx> smoser: feel free to push a3 cloud image candidates to the tracker
[13:54] <ttx> the ones without ramdisk, please
[13:54] <smoser> I just built 20100224.2 this morning.
[13:54] <smoser> so i think we'll go with that.
[13:55] <smoser> i will re-install the uec data cetner cloud here soon.
[13:56] <pmatulis> Pupeno: i still don't know what your question/issue is.
[13:56] <Pupeno> pmatulis: nevermind.
[13:56] <ttx> smoser: 20100224.2 is with ramdisk or without ramdisk ?
[13:57] <smoser>  20100224 without .  20100224.1 with,  20100224.2 without.
[13:57] <ttx> smoser: ok, push 20100224.2 then. The sooner the better
[13:57] <smoser> other than not being in the tracker (which i think i have to request someone from -release to do) its all set
[14:17] <ivoks> zul: go canada! :D
[14:17] <zul> screw the ruskies
[14:17] <ivoks> zul: at least you have healthcare :D
[14:17] <ivoks> hahahahahaha
[14:19] <zul> soren: you missed the reference
[14:19] <soren> I didn't :)
[14:19] <zul> soren: http://www.youtube.com/watch?v=CMNry4PE93Y
[14:20] <soren> zul: Seen it.
[14:20] <zul> k
[14:20] <zul> just making sure
[14:56] <EvanCC> hey there..I am trying to install ubuntu-server 9.10 in a virtual machine and I keep getting  "No Common Cdrom found"...it also doesnt seem to detect anything on boot. (at least no sda etc shown in dmesg) thoughts?
[14:56] <SirStan> Cacti just installed 8.7b . is there a way to get 8.7e on 8.04lts
[14:58] <zul> SirStan: file a bug and ask for a backport
[15:00] <ivoks> EvanCC: that's problem with your virtual machine
[15:00] <EvanCC> ivoks hmm okay I have it set to just use the normal cdrom
[15:01] <EvanCC> ill play around with it
[15:03] <EvanCC> centos has no problems btw...
[15:03] <ivoks> oh, so system boots?
[15:04] <ivoks> EvanCC: it boots from CD or it doesn't?
[15:04] <EvanCC> ivoks boots
[15:04] <EvanCC> fine
[15:04] <ivoks> oh
[15:04] <EvanCC> then goes into the install menu
[15:04] <EvanCC> asks for keyboard etc
[15:04] <ivoks> i was under impression that virt bios can't find cd
[15:04] <EvanCC> nope finds it fine
[15:04] <EvanCC> works fine until the installer tries to load the cdrom
[15:05] <ivoks> oh, ok
[15:05] <ivoks> and lspci shows what ide/sata chip?
[15:08] <EvanCC> ivoks: one second let me check
[15:09] <SirStan> Are the us ubuntu package mirrors down?
[15:09] <EvanCC> hmm it shows nothing thats a nice bug
[15:09] <SirStan> I cant seem to connect ot any of the 'us.archive.ubuntu.org' servers.
[15:15] <ivoks> SirStan: us mirror is alias to archive.ubuntu.com, iirc
[15:15] <ivoks> SirStan: and that one is slow atm
[15:48] <maxagaz> Is it possible to reattach a terminal session ?
[15:48] <maxagaz> for example, I have a sessions on pts/0 with user foo
[15:49] <maxagaz> how can I reopen it from another terminal ?
[15:49] <jiboumans> maxagaz: you'll want to use 'screen' for that
[15:49] <jiboumans> that let's you detach/re-attach terminal sessions easily
[15:50] <maxagaz> jiboumans, even for a session not opened using screen ?
[15:50] <jiboumans> maxagaz: i don't think you can reattach to one of those =/
[15:51] <maxagaz> ok
[15:51] <maxagaz> thanks
[15:58] <RoAkSoAx> ivoks, wehat do you think on what i posted on facebook?
[15:59] <ivoks> RoAkSoAx: if i'll have time, i'll do it
[15:59] <RoAkSoAx> ivoks, no, let me... i wanna try this.. i'll do it this weekend
[15:59] <ivoks> ok
[15:59] <RoAkSoAx> aafter my exams
[16:00] <ivoks> RoAkSoAx: i'm finishing last two mirs
[16:00] <ivoks> RoAkSoAx: you might want to write one for keepalived
[16:00] <ivoks> if you haven't already
[16:01] <RoAkSoAx> ivoks, keepalived is in main
[16:01] <ivoks> RoAkSoAx: you've solved the issue you had?
[16:01] <ivoks> RoAkSoAx: oh, that's great :D
[16:02] <RoAkSoAx> ivoks, an yeah i fixed it... it was a dumb issue with default routes
[16:02] <ivoks> ok
[16:02] <kiko> quick question: is it possible to resize a parition to reduce free space in an LVM entry and give more to a raw partition? I think the answer's no but..
[16:03] <RoAkSoAx> ivoks, aight. im going back to studying then :) l8r
[16:03] <ivoks> bye
[16:09] <mario_> Hello!
[16:10] <mario_> can you set up folder based quota? (not based on uid or gid)
[16:10] <Jeeves_> mario_: Yes
[16:10] <Jeeves_> xfs can do that
[16:11] <Jeeves_> it's called project quote
[16:11] <Jeeves_> quota
[16:11] <mario_> cool!
[16:28] <ttx> mathiaz: are you running your ISOtest magic on the A3 candidate ?
[16:28] <smoser> kirkland, so how is going?
[16:28] <mathiaz> ttx: yes
[16:28] <ttx> mathiaz: let us know again what's not covered by those tests
[16:29] <mathiaz> ttx: test Install (JeOS on ESX)
[16:29] <smoser> ttx, i'm testing ec2 images.. going good so far. slower than normal due to 'apt-get update' hitting security.ubuntu even though they have nothing to get.
[16:29] <ttx> smoser: I will be running the UEC image test in a few
[16:29] <mathiaz> ttx: test Install (default + crypted LVM)
[16:29] <smoser> i'm pulling isos as best as i can and will reinstall to my local cloud.
[16:29] <mathiaz> ttx: and all the UEC related tests
[16:29] <ttx> I cover the UEC tests
[16:30] <kirkland> smoser: howdy
[16:30] <ttx> Will cover the crypted LVM tomorrow morning if nobody else do them
[16:30] <kirkland> ttx: hmm, doesn't appear node autoregistration worked in my first local installation here
[16:30] <ttx> kirkland: did on mine
[16:30] <smoser> kirkland, just wondering if you were testing UEC. i'm anxious to see if all this shook out correctly.
[16:31] <ttx> kirkland: ISo testing ? PXEboot ?
[16:31] <kirkland> ttx: pxe, from archive
[16:31] <kirkland> smoser: mathiaz has the test rig right now
[16:31] <kirkland> smoser: ttx: b/c mathiaz has the lab rig, i'm testing locally
[16:31] <kirkland> ttx: i'm burning to USB now
[16:31] <kirkland> ttx: i'll test from USB
[16:32] <ttx> kirkland: i suspect it failed to get the preseed from CC
[16:32] <ttx> kirkland: if you start the NC install a little too early, and you preseed everything else
[16:32] <hggdh> soren: do you have any writeups on the automated tests?
[16:33] <ttx> kirkland: then the authorized_keys will be the only thing missing in the end
[16:33] <smoser> kirkland, you have 20100224.2 ?
[16:33] <RoAkSoAx> ivoks, how "broken" is heartbeat right now?
[16:33] <ttx> kirkland: I had to wait like 1min+ after the CLC+CC booted to have the preseed available
[16:34] <ttx> kirkland: it gets more obvious using the ISo, since if it fails downloading the preseed you'll get all the questions
[16:34] <ivoks> RoAkSoAx: it belongs to universe
[16:34] <ivoks> RoAkSoAx: it will die
[16:34] <ivoks> RoAkSoAx: let it die peacfully
[16:34] <smoser> when did 20100224.2 pop up? i rsynced cdimages like 2 hours ago , but only have 20100224.1.
[16:34] <ivoks> RoAkSoAx: it doesn't have to pull us with it :)
[16:34] <kirkland> ttx: okay, maybe so
[16:34] <kirkland> ttx: i'll be testing from usb shortly
[16:35] <RoAkSoAx> ivoks, haha i will let it die, i just would love to see it in good shape for all of those who want it as an update path
[16:35] <RoAkSoAx> and that's the same reason why Linbit took is maintainance
[16:36] <kirkland> smoser: ce1c4ebc1997ef9a49f7564f263b3c75  lucid-server-amd64.iso
[16:36] <ttx> 20100224.1
[16:36] <ivoks> RoAkSoAx: oh, it's usable
[16:36] <kirkland> smoser: i don't see a .2
[16:36] <ivoks> RoAkSoAx: it's jut not main material
[16:36] <kirkland> smoser: ttx: is there a .2 coming?
[16:36] <ttx> there is no .2
[16:36] <smoser> shoot
[16:36] <ttx> kirkland: no
[16:36] <smoser> never mind
[16:36] <smoser> gah
[16:37] <smoser> i read http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all wrong, and saw the UEC's .2
[16:37] <RoAkSoAx> ivoks, certainly not
[16:38] <ttx> kirkland: please cover the UEC cloud image test on amd64
[16:38] <ttx> http://iso.qa.ubuntu.com/qatracker/test/3673
[16:38] <ttx> I'm on the i386 one
[16:38] <kirkland> ttx: will do
[16:39] <ttx> my rig is i386 right now :/
[16:39] <kirkland> ttx: gross
[16:39] <ttx> someone has to be gross :)
[16:39] <kirkland> :-*
[16:42] <ttx> zul: can you help smoser with some of the EC2 image validation tests ?
[16:42] <smoser> ttx, its fine. i'm not in need of help.
[16:42] <ttx> smoser: ok
[16:42] <zul> ttx: sure right after lunch
[16:43] <ttx> zul: or pick your own tests :)
[16:43] <smoser> its all automated now... at least the -multi test is. i'm going to try to test some of the cloud-config and boothooks stuff, but that will take putting some stuff together to do so.
[16:44] <ttx> smoser: is uec-images rsyncable ?
[16:44]  * ttx loses time downloading the cloud image tarball
[16:45] <smoser> ttx, yes.
[16:45] <smoser> but almost useless to try to get anything from data center
[16:45] <ttx> smoser: hah, good to know
[16:45] <smoser> i have a mirror script
[16:45] <smoser> ttx, you might actually get better time from here, if you'd like to try
[16:46] <ttx> i'm 72% complete
[16:46] <smoser> and getting 50kbps ?
[16:46] <ttx> a little more. 135.
[16:49] <ttx> smoser: thanks for uec-publish-tarball btw
[16:49] <ttx> smoser: works alright here :)
[16:49] <smoser> bug free now... well, almost.
[16:50] <smoser> i'll happily remove the symlink work around soon after alpha3
[16:50] <smoser> as right now you have to have write permission to the directory where the image is stored.
[16:50] <ttx> funny to see how euca is reusing that eri reference. running euca-describe-instances and see the duplicate eri is... eye-opening :)
[16:55] <ttx> smoser: hmm, looks like your candidate has ramdisk in the tarball
 smoser: 20100224.2 is with ramdisk or without ramdisk ?
  20100224 without .  20100224.1 with,  20100224.2 without
[16:57] <smoser> ttx, you're right it does.
[16:57] <smoser> shoot.
[16:57] <ttx> smoser: shoot you ?
[16:58] <smoser> :-(
[16:58] <ttx> smoser: i'd prefer you to respin them / republish them :)
[16:58] <smoser> yeah. other option is to comapre manifests to the 20100224
[16:58] <smoser> which *is* without
[16:59] <ttx> smoser: I'm ok with that
[16:59] <smoser> < euca2ools 1.2-0ubuntu3
[16:59] <smoser> ---
[16:59] <smoser> > euca2ools 1.2-0ubuntu4
[16:59]  * ttx checks the diff
[16:59] <smoser> :-(
[16:59] <ttx> ah
[17:00] <smoser> spin takes forever
[17:00] <smoser> with data center completely shot
[17:00] <ttx> smoser: forever like... ?
[17:00] <ttx> smoser: it's the only change ?
[17:01] <smoser> < x11-common 1:7.5+1ubuntu7
[17:01] <smoser> ---
[17:01] <smoser> > x11-common 1:7.5+1ubuntu8
[17:01] <smoser> is the other
[17:01] <ttx> smoser: i don't mind shipping with euca2ools 1.2-0ubuntu3
[17:01] <ttx> smoser: you don't really need that fix, or do you ?
[17:01] <smoser> forever, like 2 hours and 55 minutes ago i started a karmic server build.
[17:02] <smoser> and its still doing the i386 portion (amd64 not yet started)
[17:02] <ttx> "order kernel/ramdisk output correctly in euca-describe-images"
[17:02] <smoser> yeah. its not absolutely needed.
[17:03] <ttx> smoser: ok, then make 20100224 the candidate ?
[17:03] <ttx> we'll respin if absolutely needed
[17:04] <smoser> yeah, i guess. it sucks. that is normally 90 minute operation.
[17:04] <ttx> smoser: do we know why it's so slow ?
[17:04] <smoser> open office.org security update
[17:05] <ttx> smoser: did you mirror that build ?
[17:06] <ttx> smoser: let me check if you can beat the DC...
[17:06] <smoser> i do have that build.
[17:07] <ttx> smoser: you'll have to make the ec2 images match the uec images, so restart your EC2 tests
[17:08] <ttx> kirkland: same for you, wait for the new UEC image/amd64 candidate
[17:08] <alkisg> Urm, could someone verify that the following gdebi syntax is correct, mainly concerning the "--option" part?
[17:08] <alkisg> sudo gdebi --non-interactive --root /opt/ltsp/i386/ --option '=--no-install-recommends' --option '=--y' google-chrome-beta_current_i386.deb
[17:08] <kirkland> ttx: eta?
[17:08] <alkisg> Those "=" at every option look wrong, but it's the only way that gdebi accepts them :-/
[17:08] <smoser> ttx, the one other option is just recreating the tarballs
[17:08] <smoser> and checksums and such
[17:08] <smoser> but thats kind of yucky as it means me touching stuff
[17:09] <ttx> smoser: that's ugly
[17:09] <smoser> so i'd rather go with the 20100224
[17:09] <smoser> yeah
[17:09] <ttx> smoser: I prefer 20100224
[17:09] <ttx> (well, I'd prefer 20100224.3)
[17:09] <ttx> but if 20100224 works, i'll take it.
[17:10] <ttx> kirkland: you should be able to download the 20100224 right now
[17:10] <kirkland> ttx: okay, my nodes are installing
[17:10] <ttx> kirkland: it just won't show on the tracker until ara updates it
[17:10] <kirkland> ttx: it did detect the CC, FWIW
[17:11] <ttx> ah :)
[17:11] <kirkland> ttx: i must not have waited long enough
[17:11] <zul> ttx: yay it works
[17:14] <ttx> zul: what works ?
[17:14] <zul> ttx: ec2 images im starting to test them now
[17:14] <ttx> zul: they are being rereshed by smoser
[17:15] <smoser> zul, dont worry about testing. i'll get them tested. script does it all.
[17:15]  * zul shakes his fist at smoser
[17:16] <zul> good ill never have to use ec2 again then ;)
[17:16]  * ttx longpauses, will run uecimage/20100224/i386 test in a few
[17:17] <ttx> zul: cover the upgrade test ?
[17:17] <zul> ttx: sure lemme go fetch the cds
[17:25] <kirkland> ttx: nc autoreg worked
[17:25] <ttx> ack
[17:26]  * ttx runs uecimage/20100224/i386
[17:26] <kirkland> ttx: can i/you update http://testcases.qa.ubuntu.com/System/UECCloudImages with the current correct image?
[17:27] <ttx> kirkland: The candidate to test is indicated on the ISO tracker, I wouldn't change it every time the tracker changes
[17:28] <ttx> hm, tracker not updated yet
[17:29] <ttx> kirkland: we should be testing http://uec-images.ubuntu.com/lucid/20100224/
[17:29] <kirkland> ttx: ack
[17:29] <smoser> ara said she'd do that after meeting
[17:29] <kirkland> ttx: downloading now
[17:29] <ivoks> ok...
[17:29] <ivoks> mirs filled
[17:29] <kirkland> 11% [[17:29] <ivoks> waiting for the verdicts :D
[17:31] <ttx> uecimage/20100224/i386 passed, confirmed bug 525989
[17:32] <ttx> will update tracker as soon as the canidate is refreshed there
[17:33]  * ttx disappears until then
[17:37] <maxagaz> is there a command line to export firefox history in a json file ?
[17:39] <kirkland> ttx: smoser: the instructions at http://testcases.qa.ubuntu.com/System/UECCloudImages still talk about ramdisk ... is that what we want there?
[17:41] <smoser> kirkland, no. those should be updated.
[17:41] <smoser> and updated to suggest uec-publish-tarball
[17:41] <kirkland> smoser: let's get ttx to ack those changes, and i'll update
[17:44] <kirkland> smoser: ttx: \o/  instance running, ssh'd in
[17:45]  * kirkland grabs lunch, will be back to test other topo's
[17:45] <smoser> ywoowhoo
[17:55] <mathiaz> smoser: UEC install up and running on the DC test rig
[17:55] <smoser> thanks.
[17:55] <mathiaz> kirkland: ISO installation in the DC test rig is working correctly
[17:55] <mathiaz> smoser: so you can log on cempedak and play with UEC
[17:55] <smoser> mathiaz, thanks.
[17:55] <mathiaz> smoser: there is only one NC for now
[17:55] <mathiaz> smoser: the others are being installed as we speak
[17:56] <jiboumans> mathiaz++
[17:56] <smoser> i'm not in terrible need now that kirkland is up and going and early results are that things are working as planned.
[17:56] <zul> ttx: samba 3.4.6 got released today there is a couple of fixes that i think we should have that Im going to cherrypick
[17:58] <mathiaz> zul: why not push 3.4.6?
[17:58] <mathiaz> zul: it seems that it's a bug-fix only release
[17:58] <zul> mathiaz: thats an option but im already in the process of doing a FFE for one thing already ;)
[18:00] <mathiaz> zul: right - I still think it's worth pulling all of 3.4.6 in lucid
[18:00] <zul> mathiaz: as do i
[18:00] <mathiaz> zul: I don't see why we should only cherrypick (ie backport patches)
[18:02] <zul> mathiaz: i agree but just in case im going to cherry pick the patches though
[18:20] <awesomeguy> please excuse me if this off topic but could u please tell me is there any way to protect files like jpg etc from copying from a website even from printscreen
[18:21] <jiboumans> awesomeguy: yeah, quite off topic and i'm not aware of any such thing being possible
[18:21] <jiboumans> awesomeguy: although google will proably be a better resource than this channel tbh
[18:25] <awesomeguy> thanks
[18:25] <netritious> I'm booting the latest hardy amd64 iso from a usb cd-rom and receive a message 'Detect and mount CD-ROM' after keyboard detection..same thing happened using unetbootin+1GB thumbdrive..am I missing something?
[18:29] <netritious> ^^the exact error is 'No common CD-ROM drive was detected'
[18:29] <netritious> thx for clearing that up for me uvirtbot lol
[18:40] <ttx> kirkland: test tracker updated, please update test results for http://iso.qa.ubuntu.com/qatracker/test/3685
[18:42] <ttx> kirkland: ack on migrating http://testcases.qa.ubuntu.com/System/UECCloudImages and http://testcases.qa.ubuntu.com/Install/ServerEConfig with uec-publish-tarball (that's how I tested)
[18:43]  * ttx will bbl
[18:55] <gabriel_> i think this is a virtualization problem, but I'll ask here in case it is a problem with Ubuntu server: anyone know why my kvm-based ubuntu 9.04 server vm refuses to get past "Starting up..." on first boot after a successful install?
[19:54] <hggdh> soren: still there?
[20:06] <sherr> gabriel_: no idea - but perhaps remove the kernel options "quiet" and "splash" from the grub kernel line you boot (i.e. stop and edit (e) at grub)
[20:12] <bogeyd6> gosh i cannot even get a cloud working from the base install and following the guide
[20:20] <bogeyd6> Why would the 9.10 install cd installing the UEC not create a bridge?
[20:32] <bogeyd6> How can I test communications between the Cloud Controller and the Node
[20:34] <bogeyd6> I do sudo euca_conf --no-rsync --discover-nodes and nothing spits out
[20:36] <gabriel_> sherr: tried that, but i still didn't get any output after "Starting up ..."
[20:37] <gabriel_> strange, i know
[20:49] <ivoks> oh...
[20:49] <ivoks> #!include /etc/dovecot/conf.d/*.conf
[20:49] <ivoks> that makes life a lot easier
[20:51] <ivoks> hm... ddeliver doesn't support them currently.
[20:51] <bogeyd6> sudo -u eucalyptus ssh-copy-id -i ~eucalyptus/.ssh/id_rsa.pub eucalyptus@10.249.88.126     << shouldnt this ask for a password
[20:52] <bogeyd6> imma reinstall the node
[20:52] <reisi> does anyone have vendor (hp/dell/ibm) recommendations for running ubuntu-server? it'd seem that especially hp (which is the only one I've got experience on) supports rhel and debian 3.0
[20:54] <Jeeves_> reisi: Almost everything works
[20:54] <Jeeves_> hp is the vendor which whines a lot about support
[20:55] <Jeeves_> 'i will not support that server, it is not mounted in a HP-rack'
[20:55] <Jeeves_> hp--
[20:55] <reisi> heh
[20:56] <bogeyd6> HP and IBM
[20:56] <bogeyd6> Dell support and hardware is more consumer grade
[20:57] <bogeyd6> dont get me wrong its good, but you cant match it up to an IBM or HP
[20:57] <reisi> and fujitsu would seem to be so hung up on microsoft that they couldn't care less about linux admins
[20:57] <reisi> bogeyd6: i've always had that kind of feeling about dell; but never used one actually
[20:58] <Jeeves_> reisi: We've just switched to Fujitsu
[20:58] <reisi> Jeeves_: oh, and you have deployed ubuntu-servers on them?
[21:00] <Jeeves_> reisi: Yes, without much problems
[21:00] <Jeeves_> Hardy had an issue with the raid card
[21:00] <Jeeves_> but everything else worked just fine
[21:00] <Jeeves_> (I don't like raidcards anyways, mdadm++)
[21:01] <reisi> Jeeves_: same here.. but for example with the hp ml350g5 with entry level e200i raid controller, i couldn't figure out how to get it expose the disks and not just logical volumes
[21:02] <Jeeves_> reisi: it seems that they can't
[21:02] <Jeeves_> I've bugged my fujitsu sales rep about it
[21:04] <reisi> Jeeves_: that's the conclusion i made; and that's most of the reason why i'm here wondering if anyone else had better products :)
[21:04] <Jeeves_> I've never had issues with Sun
[21:05] <Jeeves_> Except for their buggy nvidia NICs, and the terrible sales-team
[21:06] <reisi> heh
[21:06] <Jeeves_> But i'm not sure if the latter is better anywhere else :)
[21:12] <reisi> hmm interesting
[21:12] <bogeyd6> if i had to build a company from scratch i would either buy IBM or HP with my reqs at the cheapest price
[21:13] <jpds> bogeyd6: HP.
[21:13] <ivoks> i'd go with supermicro
[21:13] <ivoks> with 3ware raid controlers
[21:13] <bogeyd6> fo shizzle ivoks
[21:14] <ivoks> no wonder people don't like raid controlers if they use that cciss-help-me-god things :D
[21:14] <bogeyd6> I would never use fakeraid controller
[21:14] <StrangeCharm> how can i install 9.10 (alt installer) from a usb disk? i want to use full-disk encryption - which doesn't seem supported by the livecd/installer. whenever i use unetbootin to put the alt installer on a usb drive, the image boots, but it can't find the 'cd' containing the rest of the installation materials. do i need to manually mount something, prepare the disk differently, or what?
[21:14] <bogeyd6> StrangeCharm, unetbooin
[21:14] <ivoks> 3ware is not fakeraid
[21:15] <reisi> ivoks: you're saying cciss ctrls are bad?
[21:15] <bogeyd6> StrangeCharm, unetbootin
[21:15] <ivoks> reisi: no; i'm just teasing
[21:15] <StrangeCharm> bogeyd6, yes?
[21:15] <ivoks> reisi: but i wouldn't say they are better than 3ware
[21:16] <reisi> ivoks: so what would you do prefer?
[21:16] <bogeyd6> StrangeCharm, did you put the extracted iso on the usb drive?
[21:16] <ivoks> reisi: i have around 50 3wares deployed in production
[21:16] <ivoks> maybe 5-10 cciss
[21:17] <ivoks> and couple of percs
[21:17] <StrangeCharm> bogeyd6, isn't unetbootin doing that?
[21:17] <reisi> ivoks: so i guess you also like all the tools available for 3ware controllers?
[21:17] <ivoks> reisi: yeah, tw_cli is second best thing after ssh
[21:18] <reisi> ivoks: mind you, in 8.04 you cannot even query SMART with smartctl from an SATA drive, well except for the fact that it has been turned on
[21:18] <ivoks> reisi: ?
[21:18] <ivoks> reisi: that's not true
[21:18] <reisi> ivoks: with cciss e200i controller
[21:19] <ivoks> reisi: oh, yeah...
[21:19] <bogeyd6> StrangeCharm, i dont think unetbootin puts the packages on the usb drive
[21:20] <reisi> ivoks: somehow i'm not buying the disks not having proper firmware to support smart queries, so it must be either the (old) cciss module or the fact that hp thinks that it's an enterprise feature
[21:20] <StrangeCharm> bogeyd6, it just copies the image &c? the status messages seemed to suggest that it was copying the rest of the packages over
[21:20] <ivoks> reisi: it's the controler issue
[21:20] <ivoks> reisi: but it should work with cciss
[21:20] <bogeyd6> StrangeCharm, definitely does not
[21:21] <ivoks> let me check
[21:21] <bogeyd6> StrangeCharm, you need to extract the iso to its own folder on the USB drive if you wanna go that route
[21:21] <reisi> ivoks: by controller issue you mean that it's most likely by the cheap controller limiting the queries?
[21:21] <bogeyd6> plus unetbootin is strictly for livecds
[21:21] <reisi> ivoks: it works yes, but smartctl does not get any replies other than smart is enabled from each hdd
[21:22] <StrangeCharm> bogeyd6, not for installers at all?
[21:22] <ivoks> reisi: smartctl -d cciss,0 -a /dev/cciss/c0d0
[21:22] <ivoks> reisi: >> Terminate command early due to bad response to IEC mode page
[21:22] <ivoks> hahaha
[21:23] <ivoks> hp sucks.
[21:23] <reisi> ivoks: thats interesting :)
[21:24] <reisi> have to give hp props for the cciss team email being the only email address from @hp.com ever giving me any sensible answers for technical questions
[21:25] <StrangeCharm> is there something that i can run from within an ubuntu live environment that will give me the alternative installer's options?
[21:28] <bogeyd6> Can someone give me a command to run to check communication between the UEC Controller and the Node?
[21:30] <soren> hggdh: Here again.
[21:31] <ivoks> 'night
[21:34] <soren> lool: Apologies for the bumpy VMBuilder ride. I'll be uploading 0.12.1 shortly which should fix a /lot/ of things. Both regressions from 0.11, but also long standing bugs.
[21:39] <lool> soren: Did you see something which could explain losing my locales on my system?
[21:40] <soren> lool: "losing my locales"?
[21:40] <lool> soren: I suddenly only had /usr/lib/locale/en_US.utf8 and nothing else, but I had no related package update since my last reboot a couple of days ago
[21:40] <soren> lool: On the host?
[21:40] <lool> After locale-gen, things were fine again
[21:40] <lool> soren: Yes!
[21:40] <soren> Yikes.
[21:40] <lool> soren: And it's precisely the locale I pass to vmbuilder
[21:40] <soren> Uh... No idea.
[21:40] <lool> It scares me a lot, but I couldn't reproduce it
[21:41]  * soren ponders
[21:41]  * soren checks a few things
[21:41] <soren> Oh, wow.
[21:41] <soren> I know why.
[21:41] <soren> I had no idea that would happen.
[21:42] <soren> I'm /amazed/ noone has seen this before. This code is /old/.
[21:42] <soren> lool: I call "locale-gen $LANG" to check if the given lang is valid.
[21:42] <soren> During preflight, so way before there's even a basic chroot.
[21:42] <lool> wow
[21:42] <soren> ...so I have to run it on the guest.
[21:42] <lool> soren: But I just ran it again, and didn't lost my locales
[21:43] <soren> Ah.
[21:43] <soren> That I cannot explain.
[21:43] <lool> Perhaps when running as root
[21:43] <soren> Well, another mechanism for validating locales would be great.
[21:43] <lool> soren: Well that's a plausible explanation that you have
[21:43] <soren> I just tried your command line from bug 527253, and it works with current trunk.
[21:44] <lool> soren: You can actually create a vm with it?
[21:44] <lool> It still fails for me
[21:44] <lool> perhaps you have a ~/.vmbuilder or something setting domain?
[21:44] <lool> I can't pass --domain
[21:44]  * soren tries that.
[21:46] <soren> lool: No errors so far.. Exactly which branch are you using?
[21:47] <lool> soren: I'm using the lucid package right now
[21:47] <soren> 22:43:45 < soren> I just tried your command line from bug 527253, and it works with current trunk.
[21:47] <soren> "with current trunk"
[21:47] <lool> Oh ok
[21:47] <soren> lool: The lucid package is busted up.
[21:47] <lool> I thought you meant you could never reproduce, but you meant it's fixed
[21:47] <soren> Hence: 22:34:19 < soren> lool: Apologies for the bumpy VMBuilder ride. I'll be uploading 0.12.1 shortly which should fix a /lot/ of things. Both regressions from  0.11, but also long standing bugs.
[21:47] <lool> Ok
[21:47] <soren> :)
[21:48] <soren> Yes, --domain works for me now.
[21:48] <soren> I just have one more thing to merge, and I'll upload 0.12.1.
[21:50] <hggdh> soren -- on the automated test bzr, I should only worry with the ./client piece of it, or have you changed other places?
[21:50] <Tomm89> Hey.. I have a web server which is now all setup hosting some websites but.. I want to add mailboxs to each website on the server.. but i cant find any info on this anyone have any ideas?
[21:51] <Tomm89> I only have shell access... so im learning along the way :)
[21:52] <soren> hggdh: Only client/tests/kvm
[21:56] <lool> soren: I think you had the right explanation
[21:56] <lool> at least part there of
[21:56] <lool> 2010-02-24 22:55:31,469 DEBUG   : ['locale-gen', 'en_US.UTF-8']
[21:56] <lool> 2010-02-24 22:55:31,527 DEBUG   : Generating locales...
[21:56] <lool> 2010-02-24 22:55:32,955 DEBUG   :   en_US.UTF-8... done
[21:56] <lool> If I delete the locale on the host, it's created during the run -- but the other locale still stay
[21:58] <lool> soren: very suspicious: /var/lib/belocs was created tonight
[21:59] <lool> soren: I think what happened is a) I hadn't /var/lib/belocs, b) I ran vm-builder with a locale which wasn't on my system c) locale-gen detected that STATEDIR was missing and turned on purge mode
[21:59] <lool> Now I can't reproduce becasue I have /var/lib/belocs
[22:02] <Tomm89> anyone help me ?
[22:02] <hggdh> soren: thank you
[22:02] <lool> soren: Yup, reproduced now
[22:03] <lool> filing a bug
[22:07] <pteague> any reason why `pidof apache2` would say "bash: /sbin/pidof: No such file or directory"?  i was trying to figure out why i couldn't run /etc/init.d/apache restart (or reload) & that's the point at which it's failing
[22:11] <lool> pteague: pidof is supposed to be in /bin, not /sbin, but it's a symlink to killall
[22:11] <lool> pteague: So if you don't have /usr mounted, it would give such a message
[22:13] <pteague> hmm... doing `ls -l /bin/pidof /sbin/pidof` shows that they're both pointing to /sbin/killall5
[22:13] <lool> Uh right, sorry I misread
[22:14] <lool> pteague: But youd /sbin/pidof comes from where?
[22:14] <lool> *your
[22:14] <soren> pteague: Does /sbin/killall5 exist? What does "file  /sbin/killall5" say?
[22:15] <pteague> here's where the softlinks point "/bin/pidof -> ../sbin/killall5" & "/sbin/pidof -> killall5"
[22:16] <pteague> ok, that's weird... "/sbin/killall5: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.0.0, stripped"... is it supposed to be 32-bit on a 64-bit?
[22:17] <lool> No
[22:17] <lool> /sbin/killall5: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, stripped
[22:17] <pteague> `file /bin/bash` shows "/bin/bash: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, stripped" so i know it's 64-bit
[22:17] <lool> At this point I would start getting paranoid
[22:17] <pteague> hmm... any idea what deb killall5 is in?
[22:17] <lool> soren: filed lp #527381
[22:18] <lool> pteague: sysvinit-utils
[22:19] <pteague> ok, there we go
[22:24] <pteague> well, there's only 2 ways to connect to that box from outside the firewall - ssh or http (ports 22 or 80) as that's all my firewall is pointing to it...  & ssh is set up so you have to have an ssh key to get in
[22:24] <pteague> & i'm the only 1 that can use sudo on that box
[22:32] <beeman_nl> hmm i'm trying tou use preseeding to automatically install a minimal ubuntu installation
[22:32] <beeman_nl> i used kickstart before but couldn't get the pacakge selection to work
[22:33] <beeman_nl> now i use this line to select the base system, but it still installs about 450 packages: tasksel tasksel/first multiselect base
[22:33] <beeman_nl> i'd love to have less than that :)
[22:51] <beeman_nl> whatever, i'll take debian.... :(
[22:58] <mathiaz> zul: what's the reason from moving puppet templates to /etc/puppet/templates instead of /var/lib/puppet/templates?
[22:59] <soren> lool: 0.12.1-0ubuntu1 uploaded. It should be a much smoother ride.
[23:18] <mathiaz> zul: nm - I've found out why
[23:23] <pwnguin> i have an install of 8.04 i want to upgrade (several times). how do i get it off the LTS hump?
[23:25] <pwnguin> do-release-upgrade insists theres no later version than 8.04
[23:25] <pwnguin> -d tries to grab lucid
[23:26] <lifeless> well, its an LTS, so its default is to stay LTS
[23:26] <pwnguin> sure
[23:26] <pwnguin> i'd like to override that default behavior
[23:26] <lifeless> I don't know of a specific way, and checking the options doesn't make anything stand out to me
[23:26] <lifeless> I think you should mail the list and/or file a question. There may be a bug here.
[23:27] <pwnguin> which list?
[23:27] <lifeless> ubuntu-devel I would say, dru is not server specific
[23:29] <pwnguin> found it
[23:29] <pwnguin> https://help.ubuntu.com/community/IntrepidUpgrades#Network%20Upgrade%20for%20Ubuntu%20Servers%20%28Recommended%29
[23:32] <pwnguin> i thought i'd use LTS on server for less hassle, but im finding i'd rather have newer packages =/
[23:39]  * Aniya i can't believe this!!!! who the f*ck are you to do this, DIEGOPOP??? http://www.youtube.com/watch?v=cAMMiiAcSjk
[23:40] <jpds> Aniya: What?
[23:56] <hggdh> soren: what is the stepmaker? Or, better, where do I find it?