[00:00] <patdk-lap> dunno, ubuntu stopped doing xen a long time ago
[00:00] <Grenade> nonsense 12.04 supports xen !
[00:17] <patdk-lap> as a guest sure
[00:18] <patdk-lap> as a host, not as much really
[00:19] <Grenade> so i should instead install XCP
[00:19] <patdk-lap> heh?
[00:20] <Grenade> i mean 12.04 is domain 0
[00:20] <Grenade> right ?
[00:20] <patdk-lap> that would be the host
[00:21] <Grenade> which is 12.04
[00:21] <Grenade> because no other entry is shown in xm list
[00:22] <patdk-lap> I didn't say it's impossible
[00:22] <patdk-lap> it's just not on the *supported* list
[00:22] <Grenade> hah
[00:23] <Grenade> if you format to ubuntu 12.04 desktop. then install xen
[00:23] <Grenade> and xm list shows domain-0
[00:24] <Grenade> http://blog.xen.org/index.php/2012/05/06/xcp-in-ubuntu-server-12-04-lts-%E2%80%9Capt-get-install-xcp-xapi%E2%80%9C/
[00:36] <Grenade> patdk-lap
[00:36] <Grenade> ew man
[00:36] <Grenade> how to install an xml-rpc server ?
[00:57] <ClientAlive> is there any way to install vmware workstation 8 from a dvd - but from the command line?
[00:57] <ClientAlive> my host is ubuntu server 12.04 - very light weight system (cli only)
[05:59] <SpamapS> weird.. python-novaclient in precise can't talk to hpcloud basically because it says 'accept: application/json'
[06:24] <koolhead11> SpamapS: isn`t HP cloud uses Diablo
[06:24] <koolhead11> and precise comes with Essex by default?
[06:24] <SpamapS> koolhead11: yeah, something like that
[06:24] <koolhead11> SpamapS: so am assuming its the API in the releases which has resulted that
[06:25] <koolhead11> am assuming though :)
[06:46] <wrapids> What are your recommendations for ubuntu server 32v64?
[06:53] <ttx> Daviey: yes. Client libraries get their own (sequential) versioning, and it will be unrelated to server versioning
[07:18] <Daviey> ttx: Yeah, i just read the thread.  It sounds sucky IMO :)
[07:18] <Daviey> Not quite sure how i missed the thread originally
[07:25] <ttx> Daviey: what sounds sucky ? the epoch ? The fact that it's no longer connected to a particular release ?
[07:29] <pndemc_> I will be using "sudo apt-get install ubuntu-desktop" to install a gui on my server, but then I need to install a VNC program, correct?
[07:52] <smb> Daviey, Sometimes I really think the server installer takes things a bit too far when it comes to partitioning... I understand that you want to have all LVs presented. But if those LVs contain virtual disk images, I really don't care about partitions on them. Even less I want all  swap partitions in all of these VM images being marked for use... Ok, this is with Precise but again suffering from it, I thought I may spread a bit of the pain by w
[07:52] <smb> hining at you... :-P
[08:12] <Daviey> smb: I feel your pain, and i've captured it to /dev/null for later reference.  I hope this helps.
[08:13] <smb> Daviey, Thank you master. Too kind of you. ;)
[08:13] <Daviey> smb: I listen to everyones concern, note them in that file.. then look in it later, to determine what to do.
[08:13] <Daviey> currently, my list seems to be empty.
[08:14] <Daviey> This is a good thing, we've achieved everything asked of us.
[08:14] <smb> Daviey, That sounds like a great approach to handle kernel bugs assigned to me...
[08:15] <Daviey> ohnoes
[08:17] <pndemc_> I'm connecting to ubuntu server edition via VNC, however it's still command line only, how do I get a gui?
[08:17] <pndemc_> I installed gnome with sudo apt-get install ubuntu-desktop
[08:37] <ejat> how do i remote the landscape client via landscape ?
[11:23] <ea1het> good morning
[11:23] <ea1het> good afternoon for others
[11:23] <ea1het> i have a design issue about filesystems
[11:23] <ea1het> or rather a decission issue about filesystems
[11:25] <ea1het> i have a virtual machine of about 400 GB running under KVM/Qemu in Ubuntu 10.24 LTS
[11:25] <ea1het> and i need to decide either XFS or Ext4
[11:25] <ea1het> (maybe other options)
[11:25] <ea1het> any inputs from you guys will be much more than appreciated
[11:25] <ea1het> TIA !!!
[11:25] <ea1het> (listening / reading :)
[12:26] <Psi-Jack> ea1het: Still around?
[12:26] <ea1het> yes
[12:27] <ea1het> hi Psi-Jack
[12:27] <Psi-Jack> ea1het: What's the goal of this particular virtual machine?
[12:28] <ea1het> it is a web server, which serves a web service for machine translation (apache, application, web services and mysql). How much used, from 0 to 9.... 7
[12:28] <ea1het> current load.... 3% of machine resources
[12:29] <ea1het> 1024Mb RAM, 500Gb HDD (virtio) 2 cores
[12:29] <Psi-Jack> Well, hmmm. Usually it's not a great idea to run database /and/ webserver on the same machine, virtual or not.
[12:29] <ea1het> forced by the provider.
[12:29] <ea1het> i don't usually do it
[12:29] <Psi-Jack> Fire the provider, then. :)
[12:29] <ea1het> i separate everything in different tiers
[12:29] <ea1het> yesssssssssss
[12:29] <ea1het>  i would like..... but i can't :)
[12:30] <Psi-Jack> Sure you can. :)
[12:30] <ea1het> but wouldn't have that provider anymore
[12:30] <Psi-Jack> But, regardless of that. What I tend to do is this. For my database servers, I use XFS for it's data storage volumes, because it handles large files that databases will accumulate much better than any other filesystem. I use ext3 or ext4 (preferring 4 more), for most everything else.
[12:31] <Psi-Jack> Using XFS does mean you need to understand it's specific tools for filesystem recovery and maintenance to keep it nice and optimal, though, which means definitely insuring you have xfsdump installed and available.
[12:33] <ea1het> Psi-Jack: to be honest... i don't understand, neither i don't know, anything about XFS. That's one of the reason of my first question.
[12:33] <ea1het> if the community recomendation is XFS i will start over....
[12:33] <ea1het> come indications would be much more than appreciated :)
[12:34] <Psi-Jack> I specifically said XFS for database storage. :) Keep that in mind.
[12:34] <Psi-Jack> And of course, not to mix Webserver AND Database server on the same machine, because especially when using XFS, more so will you NOT want to do that.
[12:35] <Psi-Jack> XFS uses RAM as a cache/buffer more so than ext3/4 does.
[12:40]  * patdk-wk just uses ext4 for database, and at home I just use zfs
[12:41] <Psi-Jack> Heh, at home, I use ZFS for my storage servers which house my kvm qcow2 disks. :)
[12:42] <Psi-Jack> Looking into adding in CrashPlan+ to my setup so I can have better off-site backups. :)
[12:48] <patdk-wk> I've been slowly moving stuff over to my home oi server
[12:48] <patdk-wk> dunno that mythtv would be happy at all though, on it
[12:52] <zul> hallyn: libvirt is in rc mode for 0.9.13 ill upload 0.9.13 when its out
[12:56] <hallyn> when out of rc you mean?
[12:56] <hallyn> yeah i noticed the rc uploaded in debian
[12:57] <hallyn> next week i'm goign to take some time for some extra libvirt+kvm testing in q (do some harsh workloads)
[12:57] <hallyn> uh, going to try to *find* the time, i should say
[12:57] <hallyn> maybe toss in openvswitch for fun
[12:58] <hallyn> oh yeah, openvswitch dkms needs a fixin'
[13:09] <ea1het> Guys.... performance comparison between ZFS and Ext4 ??
[13:09] <ea1het> Psi-Jack:
[13:09] <ea1het> patdk-wk:
[13:10] <Jeeves_> zfs -> rules
[13:14] <ea1het> :D only that?
[13:17] <ea1het> Jeeves_: is ZFS available during install at the partition phase ?
[13:21] <tarvid> 940030
[13:23] <tarvid> today's major irritant is empty logs after logrotate https://bugs.launchpad.net/ubuntu/+source/rsyslog/+bug/940030
[13:24] <tarvid> running without logs is like driving without brakes
[13:27] <LordOfTime> tarvid: does it do that in quantal as well?
[13:27] <tarvid> haven't tried
[13:27] <LordOfTime> mind trying?
[13:27]  * LordOfTime wants to make sure its valid in Quantal before setting an importance
[13:28] <LordOfTime> or rather, recommending an importance :P
[13:28] <tarvid> this is a live server
[13:28] <LordOfTime> i meant whenever you could :P:
[13:29]  * LordOfTime would test, but his linux test system has no internet atm
[13:29] <tarvid> what are the chances or surviving on quantal?
[13:29] <tarvid> hard to simulate the load without doing it on a live server
[13:30] <LordOfTime> i wouldnt use it outside of a test environment
[13:30] <LordOfTime> heck, even my test system uses VMs for devrelease testing
[13:31] <LordOfTime> tarvid: if yuo dont test, then perhaps someone else could
[13:31] <LordOfTime> heck, i could even email the bugsquad see if they could test
[13:32] <Jeeves_> ea1het: In Linux? No
[13:32] <tarvid> I can install on VirtualBox on my laptop at the office. On EVDO here and can't stand the traffic
[13:32] <ea1het> Jeeves_: thanks!
[13:33] <Psi-Jack> ea1het: No, ZFS is not available during a normal install.
[13:33] <ea1het> using packages...right?
[13:33] <Psi-Jack> ea1het: You /might/ be able to install ZFS during such, similarly to how you do an LVM install on a Desktop install, but I dunno,
[13:34] <Psi-Jack> There's an Ubuntu ZFS PPA for ZFS.
[13:34] <Psi-Jack> ZFS though you really need at /least/ 4GB RAM just to use it minimally optimally.
[13:34] <Psi-Jack> And that's just as a file server.
[13:36] <Jeeves_> Nonsense
[13:36] <Jeeves_> I run a backup-server with freebsd and zfs with 1TB storage and 2GB memory
[13:37] <Jeeves_> with compression enabled
[13:37] <Jeeves_> works fine
[13:37] <Psi-Jack> Sure, you can, but it will not be optimal.
[13:37] <Psi-Jack> Enable dedup, and it will slow down a /lot/ on just 2GB RAM.
[13:37] <Jeeves_> Depends on what you're trying to achieve.
[13:39] <Psi-Jack> I'm primarily using my ZFS on two custom Ubuntu 12.04 setups on DELL PowerEdge 830's, each with 4GB DDR2 ECC memory, using ZFS for just my /tank which I allocate into /home, /srv/nfs, and /srv/nfs/vmdata (so I can specifically quota vmdata as needed), along with any other subvols I need. For example, I have a /srv/nfs/mail for my Zimbra mailstore, and /srv/nfs/cweb for one of my clustered webservers storage.
[13:40] <Psi-Jack> Then I have 4 Proxmox VE 2.1 hypervisors running which uses Kerberos5 NFS4 mounts to access the storage servers for their qcow2 disks. :)
[13:48] <patdk-wk> I don't see how anyone can use zfs with <32gigs ram, ideally you want >100gigs
[13:48] <Jeeves_> https://twitter.com/Jeeves_Mark/status/218339888796413955
[13:48] <Psi-Jack> LOL
[13:49] <LordOfTime> lol
[14:43] <bobg> I need to support lucid server on a new Supermicro X9 motherboard but the installer does not recognize the NICs (i350 - igb.ko) nor the SATA drives (?.ko).   I think that I need to rebuild the installer's initrd.gz with updated drivers built for its kernel (2.6.32.21). Does that sound right?
[14:45] <Psi-Jack> bobg: You're using the Server installer disc, correct?
[14:47] <genii-around> There seems to be quite a few sub-models of the X9, with different chipstes
[14:48] <genii-around> *chipsets
[14:48] <Psi-Jack> Yeah, there is. hehe
[14:53] <bobg> Psi-Jack, I am using the pxe boot installer -- but I have also tried the 10.04.4 server install disk
[14:54] <bobg> its a X9DRW-iF  (http://www.supermicro.com/products/motherboard/Xeon/C600/X9DRW-iF.cfm)
[14:54] <Psi-Jack> bobg: Then yeah, what you're saying sounds about right, however unsupported. Keep in mind though too, post-installation procedures WILL have to be done as well to get SATA back up and running to actually boot it, as well, before you reboot preferably, obviously. :)
[14:55] <bobg> oh, right  -- this sucks:(
[14:55] <Psi-Jack> You could get it to work, but maintaining it will continue to break, especially with each kernel update.
[14:56] <smoser> anyone have any improvements / features / ... that they worked on and are in alpha2 ?
[14:56] <Psi-Jack> Main part being about the SATA part primarily.
[14:56] <bobg> i run a local deb repo. I guess I need to rebuild the current kernel and put that in my repo. and then with every new kernel, do the same :(
[14:57] <Psi-Jack> Correct. Unless there's a PPA out there for that Sata driver which uses DKMS proper.
[14:57] <genii-around> dkms would be the way to go
[14:58] <bobg> Psi-Jack, oh, that is a good thought!  I will check.  Not sure what dkms is -- I will look that up
[14:58] <Psi-Jack> DKMS is amazing. :)
[14:58] <bobg> boy, now I am excited to find out what it is:)
[14:59] <Psi-Jack> I got two Ubuntu 12.04 servers acting as NAS storage, and ZFS using DKMS from the ZFS PPA. Everytime I upgrade it's packages, DKMS follows it along perfectly.
[14:59] <genii-around> bobg: It's what allows stuff like nvidia drivers to be re-made every kernel upgrade
[14:59] <Psi-Jack> automagically.
[14:59] <bobg> oh, cool
[15:13] <rhatherall> Hi, I'm running 12.04 LTS and every time I log in it tell's me I have security updates pending. I've done the apt-get update and upgrade but they're still there.
[15:13] <rhatherall> How do I update them?
[15:15] <c0d3br4w13r> hello everyone..
[15:15] <rhatherall> hello
[15:16] <Myrtti> rhatherall: have you tried apt-get dist-upgrade?
[15:16] <c0d3br4w13r> can i ask a network related question here? need some help..
[15:16] <rhatherall> Is that safe to do? I mean isn't the point of being on LTS that it;s a stable base?
[15:16] <rhatherall> or am I missing the point of what it's doing?
[15:17] <genii-around> rhatherall: dist-upgrade won't change it from being 12.04
[15:17] <rhatherall> I thought dist-upgrade was for going from 10.04 to 12.04 etc.
[15:17] <rbasak> Yeah it's safe, provided you don't edit sources.list first.
[15:17] <Myrtti> rhatherall: there's a whole different command for that
[15:17] <rbasak> dist-upgrade can work out if dependency changes require extra packages to be installed. I don't see why it would do this on a security update though
[15:18] <rbasak> rhatherall: I suggest you pastebin what apt-get upgrade and apt-get dist-upgrade say they'll do
[15:19] <rhatherall> This is upgrade:
[15:19] <rhatherall> The following packages have been kept back:
[15:19] <rhatherall>   linux-headers-server linux-image-server linux-server
[15:19] <rhatherall> 0 upgraded, 0 newly installed, 0 to remove and 3 not upgraded.
[15:20] <rhatherall> I'll go and try the dist-upgrade and see if that sorts it. Thanks.
[15:20] <rbasak> Looks like it probably will
[15:36] <bobg> when I was looking at the new features of 12.04 server, I was surprised to see that xen is officially supported -- I thought ubuntu was dropping xen in favor of kvm.  Anyone know about that?
[15:38] <bobg> maybe I am thinking of RH/centos dropping xen -- not ubuntu
[15:39] <RoyK> !kvm
[15:39] <RoyK> !xen
[15:39] <RoyK> so, yes, xen is supported, but kvm is preferred
[15:41]  * RoyK prefers kvm for its simplicity
[15:44] <bobg> can a xen VM disk image be run in kvm?   or is there a process to migrate it?
[15:46] <RoyK> I guess that depends what format it is
[15:47] <RoyK> libvirt/kvm can use a few formats, and conversion should be possible
[15:58] <smoser> roaksoax, ping.
[15:59] <smoser> bobg, "disk image" is not the issue of "can something run on xen and kvm"
[16:00] <smoser> its much more of a 'v2v' issue. disk image conversion is generally possible between most formats.
[16:00] <smoser> the ubuntu cloud images should function properly on xen or kvm.
[16:01] <roaksoax> smoser: pong
[16:02] <axisys> is there any undelete option availble? there are times when user delete a file or dir of their account by mistake and like to retrieve it.. with solaris we have daily zfs snapshot which provides a copy of file/dir back from past 7 days (we keep 7 days snapshot) ..
[16:03] <bobg> smoser, so its a matter of whether the installed kernel type (-generic, -server, etc...) supports the target hypervisor environment. right?  In 12.04, -server might run in all places?
[16:03] <axisys> using lvm .. so I can think of lvm snapshot.. but that will require more space
[16:03] <axisys> unlike zfs snapshot .. which is just hard link
[16:03] <smoser> bobg, thats a lot of it, yeah.
[16:04] <axisys> may be some undelete feature inheritent to ext4 filesystem ?
[16:04] <smoser> roaksoax, ping was in regard to maas for 12.04.1
[16:04] <smoser> are we expecting an update there?
[16:04] <smoser> and if we are, i guess i should try to get one that includes a fix for the oauth stuff.
[16:04] <smoser> oauth time/clock issue
[16:05] <bobg> are zfs snapshots really more space efficient than lvm snapshots? lvm snapshots are just deltas of changes
[16:05] <roaksoax> smoser: AFAIK, we are not updating anything until cobbler removal, are we?
[16:05] <roaksoax> Daviey: ^^
[16:06] <smoser> roaksoax, thats fine... just wanting to see where we were with that for 12.04.1. and make sure we're working in that direction.
[16:06] <RoyK> axisys: no such thing as undelete
[16:06] <Daviey> roaksoax: Well that was the plan.. but we need to start getting ready.. So if you could do what you can, that would help
[16:07] <RoyK> axisys: btrfs supports snapshotting, though, but it isn't really stable yet
[16:07] <RoyK> !btrfs
[16:07] <roaksoax> Daviey: yeah I'm currently working on getting the latest trunk packaged
[16:07] <roaksoax> smoser: ^^
[16:07] <RoyK> axisys: there's some support for snapshotting with lvm too, but I wouldn't recommend it - it's dead slow when you have many snapshots
[16:07] <smoser> roaksoax, ok. so i guess if i want oauth fix in i need to get that towards trunk.
[16:08] <axisys> RoyK: yeah.. i am not planning to use btrfs yet.. specially seeing chris left oracle and not sure how much he will have while working with new company fusionio
[16:08] <roaksoax> smoser: yeah
[16:09] <axisys> bobg: yes.. it is
[16:09] <RoyK> axisys: but zfs ftw!
[16:10] <axisys> RoyK: ftw?
[16:10] <RoyK> axisys: http://en.wiktionary.org/wiki/for_the_win
[16:10] <roaksoax> Daviey: are we SRU'ing or are we doing a backport for precise?
[16:11] <axisys> RoyK: :-)
[16:11] <RoyK> axisys: setup some rather nice servers on zfs in my last job, total capacity of 400TiB or so
[16:11] <axisys> RoyK: i guess rsnapshot is to the rescue? seems like it does hard links..
[16:12] <Daviey> zul: When is cinder landing in Quantal?
[16:12] <Daviey> roaksoax: TBD
[16:12] <zul> Daviey: i need a proper tarball which ill nag jeblair/mtaylor about today
[16:12] <RoyK> axisys: I haven't used rsnapshot, only read of it - hardlinks are good, but won't stop someone overwriting data
[16:13] <axisys> RoyK: yep.. we use solaris on lot of them.. but no where close to that size.. our 7410 is I think few TBs only
[16:13] <Daviey> zul: Will it be ready next week?
[16:13] <axisys> RoyK: i think you can keep them hidden
[16:13] <zul> Daviey: it should
[16:13] <Daviey> zul: is cinder not part of folsom-2?
[16:13] <RoyK> axisys: we just got a truckload of supermicro systems and installed openindiana on them ;)
[16:13] <axisys> RoyK: not 100% percent sure
[16:13] <zul> Daviey: i havent seen nova-volume being deprecated yet
[16:14] <axisys> RoyK: sweet.. i like omniitos a lot.. newer.. based on illumos latest
[16:14] <RoyK> axisys: yeah, seen it, but openindiana is updated regularly too ;)
[16:14] <axisys> RoyK: zfs version is latest on it?
[16:14] <RoyK> iirc the current build uses illumos kernel
[16:15] <RoyK> zpool v28, which will be *the* latest
[16:15] <axisys> i wonder what was the outcome between the last meeting between Linus Torvald and Jeff Bonick
[16:16] <axisys> hate the stupid license conflict.. so againt technology growth
[16:16] <RoyK> a new versioning regime is coming, removing the versions, and adding flags for functionality
[16:16] <bobg> Psi-Jack, i have read  up on dkms a little. this is my strategy. I install the DKMS. and copy the igb driver source to the server and edit dkms.conf to know about it. Then when I install a new kernel, dkms will kick in and build the driver from source from for that kernel. Am I on the right track?
[16:16] <axisys> s/so/soooo.../
[16:16] <RoyK> axisys: what meeting was that?
[16:18] <zul> Daviey: but i have it packaged at least
[16:18] <axisys> RoyK: it was in Jeff Bonick's blog.. have to google it.. pretty sure at least a year back
[16:19] <RoyK> k
[16:20] <axisys> RoyK: https://blogs.oracle.com/bonwick/entry/casablanca
[16:20] <Daviey> zul: why not get it in Quantal asap?
[16:21] <zul> Daviey: because i need to bug mtaylor and jeblair about the tarball generation
[16:22] <axisys> RoyK: looks like oracle butchered the picks
[16:22] <axisys> s/Bonick/Bonwick/
[16:23] <axisys> RoyK: http://jhauser.dyndns.org:8080/archives/html/humlug/2008/msg00681.html
[16:23] <Daviey> zul: ok, cool
[16:25] <souliaq> Can I install a deb manually in UbuntuServer 64?
[16:25] <axisys> RoyK: got it! http://storagezilla.typepad.com/.a/6a00d834519dbf69e2010534a68d83970b-popup
[16:25] <pdtpatr1ck> Question - is it possible to have upstart give a different exit code than 0 when it is not running? seems like whether the app is running or not running.. echo $? shows 0.
[16:25] <axisys> and here http://storagezilla.typepad.com/storagezilla/2008/09/the-rise-of-the-zfs-start-up.html
[16:26] <pdtpatr1ck> whereas the prior sysv scripts would give various exit codes
[16:26] <RoyK> axisys: :)
[16:26] <patdk-wk> currently, everyone is using zfs like ssd san's
[16:26] <patdk-wk> so many people are selling them now
[16:27] <RoyK> SSD SANs must be rather fun for databases...
[16:27] <RoyK> that is, what would you use for access? infiniband?
[16:27] <axisys> so any other suggestion besides lvm snapshot or rsnapshot?
[16:27] <RoyK> backup? :)
[16:28] <axisys> RoyK: which one? backuppc or bacula or amanda or something else?
[16:28] <axisys> rsnapshot is used for backup as well.. actually
[16:28] <RoyK> I used bacula on my last large-scale setup
[16:29] <RoyK> 100ish servers, mixed environment, five SDs for different networks, some 240TiB of storage in total
[16:29] <RoyK> bacula can scale rather well ;)
[16:30] <axisys> i like amanda .. cuz company can buy support from zmanda .. they feel safe that way instead of backup admin run over by a bus
[16:30] <RoyK> you get commercial support for bacula too
[16:31] <axisys> RoyK: oh cool.. who supports it?
[16:31] <RoyK> http://www.baculasystems.com/
[16:31] <patdk-wk> how come it's always bus's running admins over
[16:31] <RoyK> or trams
[16:32]  * patdk-wk would think it's more, snowboarding, atv, jetski's, ...
[16:32] <genii-around> They're probably always running across the street to catch another one
[16:32] <RoyK> patdk-wk: I know a few sysadmins, but not many snowboarding ;)
[16:32] <patdk-wk> heh, boring people :)
[16:32] <patdk-wk> and you have the alps
[16:33] <RoyK> good for bike trips
[16:35] <zul> Daviey: ok cinder and python-cinderclient uploaded
[16:36] <Daviey> zul: thanks
[19:02] <zul> smoser: yeah so eucatest why cant we adapt it for openstack
[19:09] <smoser> zul, what all is there?
[19:09] <smoser> all i'vve read is the readme.md there.
[19:09] <zul> still looking but it looks like basics so far which is a start
[19:13] <smoser> so yeah, it looks like at least there is a reasonable amount of stuff there.
[19:25] <zul> smoser: but it seems to be replicating alot of stuff that tempest supposedly already does
[19:25] <smoser> yeah. does tempest do the ssh stuff?
[19:26] <adam_g> zul: smoser the boto wrapping in that test suite seems to work well
[19:26] <zul> in tempest?
[19:26] <adam_g> no, in that euca test thing
[19:27] <zul> ah so maybe we should help enhance tempest
[19:27] <zul> rather than inventing the wheel again
[19:28] <adam_g> http://paste.ubuntu.com/1064869/ was enough to get me an instance
[19:28] <smoser> do you need an eutester?
[19:28] <adam_g> smoser: no
[19:30] <zul> adam_g/smoser: i say we just add stuff to tempest
[19:49] <Daviey> zul/adam_g/smoser: eutester is likely going to put more effort into ec2 support than anyone in tempest..
[19:50] <Daviey> Putting significant effort into tempest for ec2 is probably a mistake.
[19:51] <antihero> Hello, I'm looking into my ubuntu server via SSH, and when I use rxvt-unicode-256color it goes absolutely mental
[19:54] <Daviey> zul: How is nova SRU looking?
[20:03] <kyle__> Oooh, it looks like 'buntu will finaly let you install onto mdadm-raid1 without a lameo /boot partition!
[20:04] <kyle__> grub has been capable of it for years, but the installers just never let you before.
[20:04] <SpamapS> kyle__: I think thats been since.. 10.04 maybe
[20:04] <kyle__> SpamapS: I thought I tried with an 11.04 server install and it didn't let me, but I could be mistaken.
[20:04] <SpamapS> I did a lot of ISO testing on 10.10+ and they all didn't need /boot
[20:04] <SpamapS> for RAID1 I mean
[20:04] <FunnyLookinHat> I keep getting this error when trying to reinstall mysql-server :  http://i.imgur.com/LNJFG.png
[20:04] <FunnyLookinHat> Any ideas?  :D
[20:05] <kyle__> SpamapS: Actually that may have been LVM now that I think of it (also someting grub has been able to boot from for ages).
[20:05] <FunnyLookinHat> I've completely removed /var/lib/mysql just to try to remove everything...
[20:05] <kyle__> FunnyLookinHat: Did you try to apt-get purge mysql-server?  It will clean out any cruft.
[20:05] <kyle__> s/will/should/
[20:05] <FunnyLookinHat> kyle__, that doesn't seem to work... doing that before didn't reset my password
[20:06] <kyle__> FunnyLookinHat: apt-get purge _should_ reset the password, as it removes all the mysql config files data files directories magic-pixies.
[20:07] <SpamapS> FunnyLookinHat: indeed, purge will remove files which keep track of the state of the package's data.
[20:07] <kyle__> FunnyLookinHat: This is from the official repos right?  And not a unix-user you created yourself to run mysql as?
[20:07] <SpamapS> FunnyLookinHat: then on install the package should create the dbs from scratch
[20:07] <FunnyLookinHat> check this out: http://pastebin.com/JsLdL8Dw
[20:07] <FunnyLookinHat> Yeah - official precise repos
[20:08] <FunnyLookinHat> it definitely seems like a package isn't installing correctly
[20:08] <FunnyLookinHat> namely mysql-server-core-5.5
[20:09] <FunnyLookinHat> This is just a desktop - so if I really have to I can re-image... but... I'm avoiding that for obvious reasons... and because I shouldn't have to.  :D
[20:09] <kyle__> FunnyLookinHat: did you do an apt-get remove or apt-get purge to remove it?
[20:09] <FunnyLookinHat> kyle__, I've tried both
[20:10] <kyle__> hum.
[20:11] <FunnyLookinHat> err
[20:11] <FunnyLookinHat> Somehow I had 4.0 / 4.1 / 5.0 all installed ?
[20:11] <FunnyLookinHat> I just did a: sudo apt-get purge mysql-server*
[20:11] <FunnyLookinHat> And NOW it seems to be working again.  :)D
[20:11] <FunnyLookinHat> wait
[20:11] <FunnyLookinHat> scratch that
[20:11] <FunnyLookinHat> it failed to set the root password still
[20:11] <FunnyLookinHat> lol
[20:11] <kyle__> :D
[20:13] <FunnyLookinHat> ha.
[20:13] <FunnyLookinHat> Man I've never seen something this crazy before.
[20:13]  * kyle__ frowns
[20:13] <kyle__> I don't know what this server is so mad at me for.  I'm _only_ having it move 1TB of data from it's drives to an NFS mount.
[20:14] <kyle__> stupid 8.67 load.
[20:15] <antihero> ubuntu server doesn't have rxvt-unicode-256color in it's terminfo database
[20:15] <antihero> Why exactly is this?
[20:16] <kyle__> antihero: it's a terminal type.   Like vt100, vt220, xterm.  It tells the system how (and if it can) render certain things.
[20:17] <kyle__> antihero: echo $TERM from your terminal.  See what it tells you.
[20:17] <antihero> yeah, but if I ssh to my server with $TERM = rxvt-256color-unicode or whatever, it goes nuts
[20:19] <kyle__> antihero: Well then try setting the term to xterm-color, vt220, or something like that.
[20:19] <antihero> why would I do that?
[20:19] <antihero> my term is rxvt-unicode-256color
[20:19] <antihero> I am asking why ubuntu does not have a terminfo entry for this
[20:20] <FunnyLookinHat> wow.
[20:20] <FunnyLookinHat> I'm going to re-image.
[20:20] <FunnyLookinHat> No idea.
[20:20] <FunnyLookinHat> hah
[20:22] <kyle__> antihero: Because it's not a common term type, IIRC it was inented for the rxvt program and nothing else.
[20:22] <kyle__> A) Decide the world is unfair and right the wrongs of ubuntu by making them put in that term type before you can work
[20:23] <kyle__> or B) set your term when you login and get what you need/want to get done done.
[20:24] <Daviey> antihero: I don't think i've ever used rxvt-unicode-256color myself.. if you want to raise a bug.. we can certainly look into it for the next release
[20:25] <Daviey> antihero: wait, don't you just need to apt-get install rxvt-unicode-256color ?
[20:26] <Daviey> hmm, that is X support
[20:27] <zul> Daviey: slowly
[20:28] <Daviey> zul: so.. what is outstanding ?
[20:28] <Daviey> zul: We are on a deadline for completion EOD Monday. :)
[20:28] <zul> Daviey:  what?!
[20:29] <Daviey> zul: is that viable?
[20:29] <Daviey> zul: What approach have you devised to track outstanding verifications ?
[20:29] <zul> nova, keystone, glance is already there and when we did the the sru run there was no pending sru for horizon
[20:29] <zul> well i was giong to get the community to do some
[20:30] <zul> Daviey: but a new nova is pending
[20:30] <Daviey> zul: so, i'd like to get nova done for EOD Monday.. it shouldn't really be tied to other components, right?
[20:30] <zul> Daviey:  umm...i away monday
[20:31] <Daviey> Ah yes.
[20:42] <BinaryMaster> Question:  I am setting up dhcpd on a server running 12.04 and I thought it was working but the Mac os X  and iOS deveices can not get an ip address I can see in the syslog that a DHCPOFFER is made to the mac address of the os X and iOS devices but the device doesn't use it, if I switch on the dhcp server on my router all works fine any ideas what to check?
[20:45] <kyle__> BinaryMaster: What else is the DHCP server saying?  And is it set to authoratative?
[20:45] <BinaryMaster> it is set to authoratative
[20:46] <kyle__> BinaryMaster: Humm.
[20:46] <kyle__> Is it saying anything about the pool?
[20:46] <RoyK> or authoritative? ;)
[20:46] <BinaryMaster> let me grep for that hang on
[20:47] <BinaryMaster> neither word is in the syslog
[20:48] <RoyK> authoritative is the state in which your dhcp server should be
[20:48] <kyle__> BinaryMaster: pastebin your config file maybe?
[20:48] <BinaryMaster> will do
[20:49]  * zul steps away for a bit
[20:50] <BinaryMaster> http://pastebin.com/He3wC8YT
[20:54] <BinaryMaster> anything stand out?
[20:57] <RoyK> looks ok to me
[20:58] <RoyK> only, what is that "option dhcp-server-identifier"?
[20:58] <BinaryMaster> just found it from google and thought I would test it
[21:36] <rbasak> smoser: I've got utlemming's script working with the by-hash scheme now. I need to patch debootstrap and stick that and apt in a PPA, and speak to utlemming about running the modified mirror code regularly (to its own bucket). And finally generate test installer images with the patched debootstrap. Next step is to run regular installer tests I guess. Right now I'm doing main/restricted only (and quantal only). Do you think we'll want universe?
[21:53] <Daviey> rbasak: For testing this doesn't impact installing, main/restricted should be plentiful
[21:53] <Daviey> rbasak: confirming that none of this will hit production stuff yet, right?
[21:53] <rbasak> Daviey: correct. No plans to hit production stuff right now.
[21:53] <rbasak> Daviey: not sure what you mean about impact installing? I thought that installer failures were a primary use case?
[21:54] <Daviey> rbasak: exactly, and that should only need main/restricted.. i can't see that universe adds much value?
[21:54] <rbasak> Daviey: ah right. Yes - we're on the same page.
[21:55]  * Daviey jumps to the next page, before rbasak finishes reading that page.
[21:55] <rbasak> The next page is all about deployment to production
[21:55] <rbasak> You have some ideas on that then? :)
[21:56] <Daviey> rbasak: no, the next page is all pictures describing the narrative
[21:56] <Daviey> lots of ponies
[21:57] <rbasak> Oh
[21:57] <rbasak> One more thing I should probably do.
[21:57] <rbasak> I'll resign with an InRelease file
[21:57] <rbasak> To get rid of the Release/Release.gpg race condition too
[21:58] <Daviey> rbasak: it would be good to present this to cjwatson at an early opportunity
[21:58] <Daviey> Before validation.. he is the person most likely to find areas of issue/improvement
[21:58] <rbasak> Daviey: at what stage are you thinking? Just what I've done so far?
[21:59] <rbasak> Daviey: he was present at the UDS session and knows the general plan
[21:59] <Daviey> rbasak: well, i'd start showing him nowish
[22:00] <rbasak> OK
[22:00] <Daviey> Would be a shame to start putting it into validation, for it all to be a waste
[22:00] <rbasak> I'll gather together what I've done tomorrow. And email ubunu-devel I suppose?
[22:00] <Daviey> rbasak: I'm sure it will all be good.. but doesn't hurt for a review, right?
[22:00] <rbasak> Sure - good idea
[22:00] <Daviey> rbasak: that sounds good
[22:22] <blendedbychris> can i get these packages on precise somehow…. the lucid packages. https://launchpad.net/~mojocode/+archive/ppa/?field.series_filter=lucid
[22:22] <blendedbychris> I'd love to know how to create my own ppa and build my own packages but i'm not smart enough just yet
[22:25] <KM0201> i know this will probably be a bastard question here, but is there a good webmin alternative?
[22:25] <KM0201> i've seen ajenti, but.. i dunno
[22:25] <blendedbychris> none
[22:26] <KM0201> kinda disappointing
[22:26] <KM0201> i wish OMV had been based on Ubuntu... debian's repositories are void of anything good.
[22:26] <KM0201> and everytime you add a repo, something breaks
[22:27] <kyle__> KM0201: There are some web based ssh terminals, and even an ajax based VNC client (no java).  Would those count?
[22:27]  * kyle__ lothes webmin.
[22:27] <KM0201> not really.
[22:27] <kyle__> Heh, damn.
[22:28] <KM0201> i can manage the system by SSH, i just like having a webUI
[22:28] <kyle__> Are you looking for yourself or a client/parent?
[22:28] <KM0201> myself
[22:28] <kyle__> Ah ok.   I guess I"m just an old fogey.  I like the command line.  LIke that I can do almost everything form it.
[22:28] <KM0201> yeah, it doesn't bother me to be truthful.
[22:28] <KM0201> i honestly don't know why i'm so hung up on having a webUI
[22:28] <KM0201> cuz.. i'm pretty efficient at using it from command line
[22:29] <kyle__> If you want, you can install tigervnc with it's java applet.  It's one of the few OSS vnc servers that supports ad-hoc SSL (don't need to cary around a key with you).
[22:29]  * kyle__ nods
[22:35] <RoyK> deilig vær for sykkeltur i morra :) http://www.dagbladet.no/2012/06/28/nyheter/jordskred/ekstremver/innenriks/22339890/