[00:00] hallyn, yes. a few days, pretty predictably. [00:01] hallyn, is there a workaround? bug filed? === Wandrewvious is now known as WALoeIII [00:12] osmosis: (sorry, just trying to find where i have the fix) [00:12] osmosis: see the end of https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/588293 [00:12] Launchpad bug 588293 in qemu-kvm "Memory leak" [Medium,Fix committed] [00:12] (a package with the fix is in my ppa) [00:13] kirkland: that's for https://code.launchpad.net/~serge-hallyn/ubuntu/lucid/qemu-kvm/memleak-fix which also needs sponsoring for lucid-proposed [00:20] hallyn, thanks. good to know the status. [00:21] What's a good command line utility to edit wifi settings in ubuntu server === jjohansen is now known as jjohansen[afk] === dendro-afk is now known as dendrobates [00:53] anyone using vmbuilder and --tmpfs=- [00:53] it doesn't work for lucid === dendrobates is now known as dendro-afk [00:55] smoser: I should be able to look at the pyyaml/python-defaults question within the next serveral hours. === oubiwann is now known as oubiwann-away [01:35] ScottK, thats great. [01:35] please let me know if you need anything. [01:40] anyone using vmbuilder and --tmpfs=- w/ lucid? [01:47] RoAkSoAx: there isnt afaik you might want to go to slangasek open week session [02:20] zul: ok :) [02:20] thanks [02:54] kirkland, byobu-by-default doesn't work if the user has a .bash_profile ? [03:16] anyone using 64-bit ubuntu server 10.04? [03:17] I have 9 and was wondering if it is worth upgrading. [03:20] Never upgrade until you know it'll fix an important bug or add an important feature. [03:20] Because you can be damn sure that *any* software upgrade will introduce new and annoying bugs. [03:20] true [03:24] ok, so for a home server, is version 10 better than version 9 if all I'm doing is web/media/file hosting? [03:25] I'm an amateur and was looking for a more professional opinion. [03:31] Dravekx it depends on the software you are running on your old server.....my company has used ubuntu for years and can say we upgrade right away but rarely see performance when upgrading [03:32] because usually you just see little jumps in the version...for example 8.04 uses apache 2.2.8 and ubuntu 10.04 uses 2.2.14 [03:33] I'm sure there are a ton bug fixes between the two but unless you are focused on fixing a specific bug with a new version I would not upgrade. Though the sooner you start testing the never version the quicker you can move to the longer supported operating system. [03:34] thesheff17, awesome! thanks :) [03:35] the key is never to be on a version of ubuntu when the support runs out...because if there is a security flaw in the old version chances are it will be only patched in the new version and the old will never get fixed. [03:37] 12:25 I'm an amateur and was looking for a more professional opinion. [03:37] Dravekx: no. [03:37] Dravekx: for that, pretty much anything will do [03:37] Er, in terms of functionality; thesheff17 is right about security support. [03:38] We still have people running that kind of setup on Fedora Core 1, because they haven't cared enough to upgrade. [03:38] oh [03:39] since the server is home based, security is priority one. [03:40] I only asked because when v9 was released, there were opinions of frustration about bugs and issues. A lot of people turned back to v8. This is the reason for my question. [03:40] Dravekx: then you probably want to look for references to "hardening", paying attention to the publication date on any articles you find. [03:40] what software specifically? [03:40] thesheff17: he said "all I'm doing is web/media/file hosting" [03:41] apache/mysql/samba << that's about it. pretty simple. [03:41] MySQL won't talk to anything on its own, so you presumably also have PHP and some shitty PHP apps. [03:41] hehe [03:42] well I can say I really like ubuntu server vs just fedora [03:42] does fedora even have a server version? [03:42] thesheff17: I hate Ubuntu, but at least it's closer to Debian. [03:42] why do you hate ubuntu? [03:42] currently, it has LAMP installed which has run fine from the beginning. [03:42] I think ubuntu is very good for the lamp stack [03:43] cool :) [03:43] thesheff17: because it's largely controlled by a single group, and they make concessions to business/corporate needs that Debian wouldn't. And they tend to focus on wanky GUI desktops and sometimes break things for other use cases. [03:44] Of course, it's BECAUSE Ubuntu has e.g. fixed release dates, that I can use it in a corporate environment, but I'm not allowed to use Debian. [03:44] true twb...I was coming from a red hat world [03:44] so switching to ubuntu was a dream come true [03:44] Right. [03:44] ug sorry to hear you can't use debian. [03:45] But a lot of those benefits are because Ubuntu's leveraging Debian's history of good Q/A [03:45] by far [03:45] if it wasn't for debian....ubuntu would never been around [03:45] I came from a Microsoft world. I was told Ubuntu was the "preschool way" to learn. lol. [03:45] the day I found the apt-get...used to be gentoo guy...I was so happy [03:45] so far, I love it. [03:45] Debian wins from a combination of dpkg/apt and debian-policy/lintian. [03:46] The dpkg/apt framework wouldn't be any good if they packages it installed were of RH quality [03:46] haha so true [03:47] Dravekx...ubuntu is geared twards people that haven't used linux...but it really does everything any other linux does....people are just afraid to make that jump to linux [03:47] i tried installing debian, but it kept failing. I was told, due to the server hardware being specifically based for Windows Server Only. But Ubuntu works perfect and without any problems. [03:47] I assumed I didnt know what i was doing. LOL. [03:47] Dravekx: Debian is certainly optimized for people who know what they're doing. [03:47] the fact you have heard of ubuntu is better than most...can't tell you the number of companies I interview for and say they are tired of redhat but don't know what to use. [03:48] hardware is always hit or miss on linux...one day hopefully all hardware is supported by every linux os. [03:49] "One day" we should have a better kernel than Linux. [03:49] hehe i'm not sure it will be in our life time...but I will keep my fingers crossed. [03:50] twb have you used vmbuilder I swear it worked w/ 9.04 but 10.04 won't build an image using the tmpfs which builds it ram [03:51] it is even in the damn man pages [03:51] I haven't, no. [03:51] but tells me it isn't valid [03:52] cool. files transfer complete. I'm going to blow the system off and go from scratch. thanks for the opinions, guys :) [03:52] Unfortunately I'm still stuck on the in-house analogues which I wrote five to eight years ago, because there wasn't anything like vmbuilder at the time. [03:52] Dravekx: why don't you just upgrade in-place? [03:52] took about 1 day for me to build an Eucalyptus cloud with ubuntu...and then the controller machine it was running took a crap the next day. [03:53] !upgrade >Dravekx [03:53] Dravekx, please see my private message [03:53] I'm look for a quick and dirty solution to build my company testing env [03:53] automatically [03:53] smoser: hrm, really? [03:53] smoser: that's a regression, if so [03:54] I guess I could always go back to virt-clone [03:54] twb, there's a lot of "left over" garbage. It has 4TB of information ont he tail end with 1.2TB are old files needing to be cleared. [03:54] but virt-clone breaks the network adapter and I have to manually fix it everytime [03:54] it's easier to do a fresh install. [03:55] Dravekx: if you say so [03:55] OCD doesnt help. [03:56] :S [04:00] how can ubuntu put this in the documentation if it doesn't: [04:00] As you can easily imagine, writing to RAM is a LOT faster than writing to disk. If you have some free memory, letting vmbuilder perform its operation in a RAMdisk will help a lot and the option --tmpfs will help you do just that: [04:06] thesheff17: how much free ram do you have? [04:07] I have 8GB on this system and tells me 6950 of it is unused [04:07] From first principles, I'd expect a desktop build to consume anything from 2GB to 6GB during the build process [04:07] OK. [04:08] AFAIK you haven't pastebinned the exact error output yet. [04:11] yea my exact command is: sudo vmbuilder kvm ubuntu --suite=lucid --flavour=virtual --arch=amd64 --mirror=http://192.168.1.4/ubuntu -o --libvirt=qemu:///system --tmpfs=- --ip=192.168.1.98 --mem=512 --hostname=vm1 --bridge=br0 --user=ubuntuadmin --name=ubuntuadmin --pass=password --addpkg=vim --addpkg=ssh --addpkg=ntp --addpkg=ntpdate --addpkg=xvfb [04:11] and I get the vmbuilder: error: no such option: --tmpfs [04:13] Oh. I had the impression that it was failing during the build. [04:13] or sorry no...it is failing prior to that...I'm running it without it right now...let me see what the output is. [04:15] maybe because I need this vmbuilder.partition file [04:18] hey guys - anybody with LVS experience around? [04:19] LVS? [04:19] linux virtual server [04:19] keepalived/ldirectord etc... [04:20] I have a pair of servers running varnishd, and a keepalive service on one of them to balance the connections between them [04:20] chrismsnz [04:20] very little [04:20] I have used keepalived [04:21] however keepalive/ipvs is heavily favouring the local server (by over 10x) for connections - even though they're the same weight and using "weighted least connections" algorithm [04:21] thesheff17: the documentation is absolutely terrible unfortunately :\ [04:21] yea tell me about it [04:22] we considered moving to ldirectord - but that seems just as bad D; [04:22] I think that is why I abandoned it along time ago. [04:22] I use pound [04:22] but it isn't active active [04:22] pound with heartbeat? [04:23] we use the failover capabilities, too [04:23] yea...I haven't gotten that far with pound [04:23] I wish I have [04:23] I have used pound with EC2 as the load balancer [04:23] and it will just know if one of the two die [04:23] not that great but works [04:24] well, we're already using varnish as reverse proxy/cache/backend load balancer [04:24] which, is friggen awesome [04:24] we just need something to balance/failover the two varnish servers we run [04:25] and the lvs low-layer balancing is ideal... unfortunately we have this going on :\ [04:26] anybody bored and want to help me setup a web/file/media server with 2 drives? lol [04:26] can't use just use hearbeat w/ varnish? [04:27] Dravekx ask away...not sure what config you are looking for [04:27] k :) [04:29] thesheff17: we can, although we'd like to use both servers if we can - to handle traffic spikes [04:29] one of the varnish servers is enough to handle load 90% of the time, but we get busy periods so it's good to spread the load [04:29] can't you just do round robin DNS [04:29] against the two varnish? [04:30] yep, could do, it's just if one failed the other would have to pick up it's IP [04:30] that's not a terrible idea, so will look into it [04:31] which should I use: automatic updates, none, or landscape? before i was manually updating. [04:31] yea I bet some DNS services will know if the varnish dies and just redirect all traffic to the varnish that is still up [04:31] I always do manual updates....I would hate for something to break and not know it is related to updates. [04:31] not without a tiny ttl :\ [04:31] good point :) [04:32] I guess what you want is heartbeat as well on each machine [04:32] and it should pick up the other IP if one fails [04:32] yeah [04:33] we're even thinking of adding a third one [04:33] which might end up being omfglame [04:33] haha [04:34] luckily I make our data center buy F5 for production...but our dev env I have fought with pound varnish and all the others for a long time [04:34] http://www.howtoforge.com/high-availability-load-balancer-haproxy-heartbeat-debian-etch [04:34] a good how to with heartbeat w/ apache [04:35] should be the same with varnish [04:35] I would love to hear if you get it to work...I'm always looking for a free solutions to our F5 [04:36] they cost a small fortune for support. [04:37] hey don't get me wrong - when lb was running on a different server it ran amazingly [04:37] before, I was using SSH to transfer files around from win7 to my server via lan/wan. Is samba a better choice or should I stick with ssh? [04:37] never, ever had a problem with keepalived/lvs [04:37] smoser: I just uploaded a fixed python-defaults. Once that's built and published, a no-change rebuild for pyyaml should fix it. [04:37] and even now, it's balancing connections, it's just being odd about it [04:38] hehe chrismsnz...try to just offset the weirdness :) [04:38] by overloading the under utilized server :) [04:38] wonder why it won't do round robin [04:39] i use it to balance http, https, ftp connections and a couple other things on a crappy supermicro p4 with 600+ concurrent connections [04:39] wow [04:39] Dravekx [04:39] ssh is great...samba is faster [04:39] yeah i might try messing with the weights [04:39] from experience [04:39] "all current models of the BIG-IP appliance have specialized hardware for SSL handshakes" [04:40] Cool beans. [04:40] (http://en.wikipedia.org/wiki/F5_Networks) [04:40] k :) [04:40] yea that is nice about the F5 [04:40] that we don't have to load https ever [04:40] pound also though supports an SSL certificate [04:41] I'm not in that space, so I'm not familiar with such stuff [04:41] which is another problem I have found with other reverse proxy don't always [04:41] Closest I get is VIA AES in hardware [04:41] thesheff17, samba is for local lan/wan files and printers, correct? [04:41] Dravekx: Samba provides Windows-flavoured file and printer sharing. [04:41] yes [04:42] ...and Windows-flavoured authentication. [04:42] yea the F5 we have is in a shared env...I hate it [04:43] Dravekx samba uses the smb protocol which I have found very quick when transfering between linux/windows [04:43] instead of ssh [04:43] I would not use samba really though on a WAN [04:44] for WAN transfers I use rsync [04:44] rsync is SSH + only transfers what has changed [04:44] FSVO SMB = CIFS [04:45] You *can* run rsync unencrypted, but I wouldn't do it over an untrusted network, except for stuff like the Ubuntu install .iso [04:46] thesheff17, I have 2 computers on the LAN, and 3 laptops on the WAN that need to access/transfer files. I was using openssh with WinSCP for transferring everything. it was really slow. [04:46] its all on secure intranet. [04:46] so the 3 laptops are in different location? [04:46] thesheff17, yes [04:47] thesheff17, diff location inside the house. [04:47] Dravekx: so you have 100baseT between them? [04:47] why do consider the laptops WAN? [04:47] they should all be on a private network [04:47] 192.168.1.x usually [04:47] uhm. Fiber Optics on the LAN. [04:48] how do the laptop connecrt to the network? [04:48] WAN: Wireless Access Network ( via router) [04:48] so a wireless B/G/N connection [04:48] thesheff17, yes :) [04:49] hmm....all your machines should really be in the same subnet [04:49] also just using wireless will be a bottleneck [04:50] Hello, I filed a merge request for mutt (LP #588736) a while ago [04:50] Launchpad bug 588736 in mutt "Candidate release mutt 1.5.20-9ubuntu1" [Undecided,Confirmed] https://launchpad.net/bugs/588736 [04:51] thesheff17, the router allows each connection to have an intranet IP of 192.168.1.x (x being 1-10) [04:51] thesheff17: funnily enough, changing the balancing algorithim from lc (least connections) to rr (round robin) seems to have fixed it [04:51] _b [04:52] thesheff17, the server is on passthrough, so it uses the main IP. (not sure the lingo is correct) [04:52] Dravekx when you say WAN usually it refers to a public IP...it my mind...so they are all 192.168.1.x [04:52] yes [04:53] thesheff17, the WAN is encrypted, so it's not exactly public. :) [04:53] how does least connections work? weird that is any different than round robin [04:53] since round robin should balance all anyway [04:53] Dravekx...I would try samba [04:53] with the laptops [04:54] and make sure there is no other traffic on the 2 LAN machines and wireless router [04:54] k :) [04:55] Dravekx you can also use sync toy and robocopy...which is rsync for windows....which will only tranfer stuff that has changed. [04:55] *transfer [04:55] thesheff17, ty for all this! :) I will look those up. [04:55] lc will send the next connection to the host with the least connections [04:56] useful when you have connections/requests of different lengths [04:56] rr will do fine though [04:56] sync toy is a GUI based front end require crap loads of .net...robocopy is command line driven and is in some windows server zip utility package. [04:56] chrismsnz: ah very true... [04:58] yea maybe one of the things you are doing are making the packets huge causing lc to favor the one machine [04:58] Er, cwRsync is rsync for Windows [04:58] well cygwin is cygwin [04:59] Granted. [04:59] cwRsync is basically just the bare minimum of cygwin necessary for rsync --rsh=ssh === clintb is now known as SpamapS [04:59] ah nice [04:59] I have used robocopy w/ a samba share for years with great results [05:01] but I really hate the dos prompt [05:01] I wish I could never touch a windows server again [05:01] lol [05:01] We have interns for that [05:02] I'm trying to steer clear of Windows Server. i have been trying for the past 2 years. lol [05:02] yea I wish I had an intern [05:02] hopefully soon [05:03] the worst is logging on having the server tell you it is time to reboot [05:03] because of updates [05:03] in the middle of the day [05:03] ...with the annoying "every 10min reminder" [05:04] omg...yea don't remind me [05:04] I ran a huge .NET app where people connected through RDP to run the application....everyday that stupid window would pop up [05:06] thesheff17, configuration time... this is the part i get lost, dazed, and dangerous. although google seems to work at times. [05:06] http://www.dravekx.com/ [05:07] so far so good [05:07] apache working [05:07] :D [05:08] yea for configuration...I always backup the template...or whatever is there [05:08] and then start editing slowly [05:14] etckeeper is your friend [05:14] hehe I'm lazy [05:14] Well, current versions have an automated daily commit [05:14] oh wow [05:14] nice [05:15] I thought I broke my web site today messing with the default...it was actually just a problem with cc gateway [05:15] On 8.04 you can approximate it by simply adding an /etc/crontab entry @daily root etckeeper commit "Daily autocommit." [05:15] etckeeper would have helped today [05:17] I hear more and more people using puppet [05:17] but haven't used it yet myself [05:17] Puppet's a pain in the arse [05:17] good you just saying that...I don't want to learn it ;) [05:17] Unless you have a rack full of gear, don't bother [05:18] I do...but I get by with scp/ssh fine [05:18] this is also really nice: http://www.capify.org/index.php/Capistrano [05:19] Puppet's primarily useful when you have a large number of proximal, wildly heterogeneous hosts in heterogeneous roles. [05:19] yea very few machines are the same in my env [05:19] It's no use to me because most machines I babysit are nominally under political/administrative control of my customers. [05:20] I see [05:20] haha prob have to just permission to do sudo [05:26] well i got vmbuilder to run w/ out that tmpfs [05:26] it creates the virtual machine...but then I go to start it and it says: error: Failed to start domain vm1 [05:26] I love when I find post on the ubuntu forums with the same problem with no answers :-/ [05:27] though without tmpfs it is taking 10+ min [05:27] very terrible results :( [05:29] twb: huh? Puppet is very useful when you have 100 identical machines.. did you mean homogeneous? [05:30] in fact [05:30] thats sort of the point of puppet. ;) [05:31] if they all have different roles.. then its still very useful, but its more useful in keeping machines of the same role consistent. [05:31] SpamapS: if you have 100 identical machines, then you can use a simpler solution like an SOE [05:31] ...unless they have different roles, right [05:31] twb: SOE? [05:31] standard operating environment [05:32] oh like a disk image? [05:32] those are a real bitch to update [05:32] Yeah, basically. [05:32] yea I have heard people using puppet to manage both apt-get and yum [05:32] Depends how you deploy it [05:32] I've used systemimager+puppet in the past to great ends. [05:32] e.g. you could just have /usr shared read-only by the SAN [05:33] That'd be the super-traditional way [05:33] SAN's are so old fashioned. ;) [05:34] my poor company doesn't have a san :( [05:35] puppet environments seem to thrive with a few classes of machines that are all 90% the same.. like, fast web servers, slow web servers, medium web servers, they're 90% the same but have slightly different configs and builds of apache. [05:35] thesheff17: well even if you could afford one, you'd be poor again after you bought the SAN. ;) [05:35] haha so true [05:36] because either a) it would be ridiculously expensive bankrupting you, or b) you'd spend too little, and it would fail, bringing on your own demise and subsequent poverty. ;) [05:36] hence the newer model of limiting exposure through independent, eventually consistent servers. [05:37] hopefully though...new capital soon which means new hardware :) [05:37] spend it on people [05:37] if you saw my hardware you would laugh [05:37] 1 rock star developer will save you 50 servers every 6 months. ;) [05:37] true [05:38] we did have an old IBM san that crashed running oracle [05:38] it was the worst week ever [05:38] SpamapS: aren't rockstar devs drones, just like at EA? [05:39] usually good developers don't work for drones like EA [05:40] http://en.wikipedia.org/wiki/Rockstar_Games [05:40] twb: [05:40] even if they enter as great devs. [05:40] they leave as shells [05:41] You mean like Dr. zsh? [05:41] (http://cs-www.cs.yale.edu/homes/shao/) [05:46] ug now I'm getting AttributeError: 'Libvirt' object has no attribute 'vm' w/ the vmbuilder [05:46] still without tmpfs [05:46] giving up for tonight...will be back on in the morning...talk to everyone later [05:48] thesheff17, do you have a little time? this is where i get confused. [05:48] ok [05:53] anyone good with symlinks? or is that a different channel? [06:02] hi [06:02] hi [06:06] I need some help with symlinks. if someone can help, can you please pm? [06:06] qwert: hi [06:06] is there a way to know what causes high wa on Ubuntu Server ? [06:11] I mean high iowait [06:12] Dravekx: what do you want to do? [06:13] kaushal: your programs are most likely just using the disk systems a lot. do you have 'sysstat' installed? [06:15] SpamapS: let me look at it [06:16] SpamapS: its installed [06:17] sar,iostat and mpstat [06:17] sar, iostat and mpstat - system performance tools for Linux [06:22] e_t_, :( I want to host websites from user home directories. BUT I want my home directory to be the main site. [06:22] kaushal: ok, sar -d .. does it show which disks are being used and how much? [06:23] Dravekx: maybe you want virtual hosts? [06:23] SpamapS, I think so. Im not sure. I'm reading about it now [06:23] Dravekx: do you want the webserver to use servername/~username/ to find their content, or userhostname/ ? [06:26] SpamapS, sort of I think. let me try the apache channel first. lol. I think it is more on subject. :) [06:31] :/ [06:31] zip just hard crashed my file server [06:31] not even sysrq commands would work [07:07] SpamapS: sar -d says Requested activities not available in file [07:07] is there a way to alert it via email using sar ? [07:26] i am currently working on LDAP, how do i flush the LDAP database?? ...help..... [07:29] Is it possible to use whole unpartitioned disk in ubuntu-server installer as LVM PV? [07:34] darkk^: yes, it's possible === Lanta is now known as Lantizia [07:46] joschi, how can I do that? I want to use vda as /boot and vdb for LVM, but installer does not allow me to use unpartitioned vda for /boot [07:48] darkk^: pvcreate /dev/vdb [07:48] darkk^: given that /dev/vdb is a block device [07:49] darkk^: I actually don't know if lvm works with virtio devices, but *real* devices (/dev/hd*, /dev/sd*) definitely work [07:59] joschi, I tried to "pvcreate /dev/vdb" but installer says, that PV is used. [08:01] seems, I should rather use debootstrap instead of installer :-) [08:01] darkk^: `pvdisplay` should show the device then. You could also try `dd if=/dev/zero of=PhysicalVolume bs=512 count=1` like mentioned in the man page of pvcreate [08:08] pvdisplay displays pv and I've zeroed virtoio backend devices before using them - it's just installer. Though, it works after manual creation of vg, but still does not allow to use whole vda as /boot suggesting forcing me to use vda1 - but that's not a big deal. [08:56] New bug: #600106 in bind9 (main) "apparmor blocks journal creation for dnssec" [Undecided,New] https://launchpad.net/bugs/600106 [08:59] Hi I cant connect mysql server from a remote machine. I get this "Lost connection to MySQL server at 'reading initial communication packet', system error: 111" [09:06] rahman: $ perror 111 [09:06] OS error code 111: Connection refused [09:06] rahman: have you set bind-address in your my.cnf? are you running a packet filter on the system which will block access to port 3306/tcp? [09:10] hi [09:10] can some one please guide me about the sar utility [09:10] I have been following http://www.cyberciti.biz/tips/identifying-linux-bottlenecks-sar-graphs-with-ksar.html [09:10] basically i have logrotate application which runs at 1:00 am [09:11] i need to gather reports using sar [09:11] */10 * * * * root /usr/lib64/sa/sa1 1 1 [09:11] 53 23 * * * root /usr/lib64/sa/sa2 -A [09:12] not sure i understand that [09:12] basically i need to send sar report it to a specific team [09:12] Please suggest [09:37] can someone please guide me in helping about my query ? [09:39] ttx: ping? [09:40] RoAkSoAx: pong [09:40] ttx: maybe you can help me. Is it possible to backport a new upstream that makes use of DKMS. For example, DRBD? Since it registers to the kernel and stuff [09:41] RoAkSoAx: you mean, in -backports ? or as SRU ? [09:41] ttx: what ever works [09:42] I think that change would be too big for SRU, but sounds like it would be possible to -backports it [09:42] ttx: ok. Will look into itsince I know believe the DRBD kernel module is now included into the kernel that maverick uses [09:43] and lucid will need to make use of DKMS [09:53] yay kosmic koala: bug 599910 [09:53] Launchpad bug 599910 in libvirt "Libvirt/kvm permissions/ownership issue on upgrade from Kosmic to Lucid" [Undecided,New] https://launchpad.net/bugs/599910 [10:00] ttx: btw... would it be possible to backport if the one in maverick is not using DKMS and the one in lucid is. This means lots of changes are necessary for the lucid package [10:01] because the maverick kernel includes the drbd module, so no need for dkms anymore [10:02] RoAkSoAx: I think that would still be acceptable for -backports [10:03] ttx: ok than. Thanks! === rapha_ is now known as rapha [10:43] Hello ! [10:43] Hello [10:44]