[00:11] New bug: #890917 in bacula (main) "bacula won't uninstall" [Undecided,New] https://launchpad.net/bugs/890917 [00:14] anybody tinker with 11.10 in a Meru wireless network yet? [00:38] Hi [00:38] Is there a way to implement chroot home directory using sftp in openssh server 4.7 ? [00:39] I am on Ubuntu Server 8.04 [00:42] Try http://shapeshed.com/journal/chroot_sftp_users_on_ubuntu_intrepid/ [00:43] jandrusk: the version mentioned is 4.9 [00:44] openssh-server version is 4.7 in 8.04 LTS [00:50] correct, that functionality was not available in that version [00:50] you must use a traditional jail, I've used jailkit in the past [00:51] caveats being said jail applies to the user in all contexts, not just SFTP, and it requires more effort to set up [00:51] my suggestion is to upgrade [00:52] hardy's only got about a year and a half left on it for support as it is [00:53] hi [00:53] anybody tinker with 11.10 in a Meru wireless network yet? [00:56] +1 for internal-sftp + chroot [01:38] Why not just upgrade to the latest? [01:39] LTS has its merits to be sure [01:39] but I would upgrade to lucid [01:42] lucid is lts anyway === jason is now known as Guest66271 [01:51] hmm, are server upgrades from 6.06 to 10.04 (or from 8.04 to 12.04) supported? [01:51] they were, and may still be [01:52] 6.06 to 8.04 may be tricky but was a supported option [01:52] 8.04 to 10.04 is supported and 10.04 to 12.04 will be supported [01:53] if you mean skipping an upgrade, then the answer is no [01:53] they must be done incrementally [02:02] !upgrade [02:02] For upgrading, see the instructions at https://help.ubuntu.com/community/UpgradeNotes - see also http://www.ubuntu.com/desktop/get-ubuntu/upgrade [02:02] kklimonda: there is copious documentation. If you have read it and still have problems, ask away [02:02] (As qman says, you need to upgrade to each LTS in turn.) [02:09] I don't really have problems :) I was just wondering if skipping LTS is supported (Kubuntu team did support upgrades that skipped releases, I was wondering if maybe server team does that too) [02:11] AFAIK no === Guest66271 is now known as jasef [03:25] I'm trying to get a computer to boot via PXE to my Orchestra Server, but I'm not running DD-WRT and I don't exactly understand the role of the "Static Lease" [03:26] a Static lease is where a router gives a machine the same IP everytime via DHCP [03:29] jMyles: do you have access to the dnsmasq.conf ? [03:31] twb: I'm using Zentyal, which I think uses dhcp3 instead of dnsmasq [03:31] OK that has an icky format but can be done [03:31] In dnsmasq you jsut say "go use /etc/ethers" [03:45] also known as a DHCP reservation, in dhcp3 it's defined in a host block [03:46] like so: http://pastebin.com/11i0wgW2 [03:46] I call it "fixed DHCP" to avoid confusion with static (non-DHCP) allocations [03:50] I call it reservations because that's the microsoft term, and microsoft DHCP was the first time I'd ever heard of or used it [03:51] and yes, "static DHCP" is very prone to confusion [03:52] and, unlike most microsoft terms, that one is unambiguous and technically accurate [03:54] I don't know anything about microsoft [03:55] It's fixed like a nail, not like a dog [04:23] OK, got around the Zentyal issue. Now, when we boot via PXE, we are given a list of profiles, but not the 'system' with the image (iso) that we have specified. [04:25] WHich one is zentyal again [04:25] Oh it's one of the webmin/cpanel things [04:26] turnkey ebox distro [04:29] It seems that I havent had any updates to my server in a few months. Any idea how to track down if 1) I'm up to date 2) if something is blocking updates? [04:30] cgkades: check if sources.list looks sane, run apt-get update, run apt-get dist-upgrade --dry-run [04:30] We can't seem to make orchestra work - if we pick one of the prefab profiles it goes through the kickstart and then hangs. We don't have an option to boot an image. [04:30] jMyles: dunno about orchestra, sorry. All my stuff is in-house work that predates it. [04:31] boutil: jMyles using 11.10 ? [04:31] err [04:31] sorry [04:31] jMyles: using 11.10 ? [04:31] cgkades: also check your release (lsb_release -a) and see if it has been end-of-lifed, in which case there are no updates, but you should upgrade to a new release [04:32] yep 11.10 [04:32] jMyles: hangs on the first boot or still during the installer? [04:32] twb: looks like i'm on 9.10 i thought i was on 10.04 LTS [04:32] SpamapS: It goes through what looks like the initrd, past DHCP config, and then hangs on the pinkish screen [04:33] SpamapS: why does orchestra use ks instead of preseeding? [04:33] twb: thanks for the help [04:34] twb: it doesn't [04:34] twb: but cobbler was built around the word kickstart.. so sometimes kickstart is used as the word. But it is definitely built on pre-seeds. [04:34] Ah, cobbler comes from the FC world? [04:35] Yeah, Cobbler just moved off Fedorahosted onto Github today actually [04:37] jMyles: have to run, but if its hanging during the install, you should see the logs on the cobbler server. If its hanging on first boot, try booting without 'quiet' and with '--verbose' on the grub cmdline [04:37] * SpamapS disappears [04:37] RTFL FTW :P [04:38] SpamapS: Where and when during the PXE boot am I supposed to see the images or systems I create in the cobbler gui? [04:40] SpamapS: BTW, I am also getting a python AttributeError when I try to create an image: 'NoneType' object has no attribute 'os_version [04:46] jMyles: dunno about orchestra, but normally PXE boot will load the PXE ROM off the local h/w, that will do a DHCP request, it will then do TFTP to fetch $filename (usually pxelinux.0) from the TFTP server (defaults same as DHCP server). [04:46] twb: That part seems to be working in this case. [04:47] Once pxelinux.0 takes over it usually tries to fetch a MAC- or IP-specific .cfg file, falling back to pxelinux.cfg/default. THAT then usually tells it to load some more config files, a menu.c32 or similar, etc, allowing the user to pick a boot option [04:47] Once they do, pxelinux.0 THEN will pull down (still over TFTP) a kernel and ramdisk, and these are loaded. [04:48] They then will follow their coding and boot: options to find and pull down everything else, usually preseed.txt or so coming down either via TFTP (very new) or HTTP/FTP/NFS (traditional) [04:48] Oh, during that last pull down there will be a second DHCP request, which a clever DHCP server might serve something different for [05:06] gpxe is nice, downloads and configs over http [05:06] but I just can't get the damned thing to fit in under a 64k rom [05:06] atleast not with iscsi support, haven't attempted it without iscsi [05:06] Eh, TFTP is a better protocol for the job [05:07] that depends [05:07] tftp has no authenication or protection [05:07] I admit the setup is a pain if you don't already have a TFTPd and you DO have an HTTPd [05:07] and I dunno any tftp that allow scripts to change the returned files [05:08] but my goal was to add iscsi netboot, to systems that didn't have iscsi nic [05:08] You mean like CGI instead of static HTTP? [05:08] yep [05:08] IMO if you're getting fancy like that (or auth) then OK, http [05:08] But I think it's a bit silly to be serving secrets during the boot process... [05:08] I think some things also get odd, and do http range requests for boot also, not sure though [05:09] static serving, put the smarts inside the pxelinux.0 (or equivalent) chainloader, no secrets so no auth needed [05:09] not sure of the idea there is [05:09] unless it's just to keep unautherized people off that network [05:09] fuse curlftpfs is pretty cute for mounting the sq over HTTP (using range requests), but I think I'll stick to nfs/nbd/aoe :-) [05:09] secrets burned into rom, with https [05:10] patdk-lap: if you want to keep unauthorized people off the net then bloody well lock it down properly (physical layer, etc), not in the bootloader :-/ [05:10] twb, hard to do that in a classroom [05:11] Oh is this one of those stupid things where lecturers want to make sure people who are studying other classes shouldn't get access to their teaching materials [05:11] doubt that [05:11] more of a thing of ignoring random things that shouldn't be on the network I guess [05:11] just seen people doing it [05:11] If students have physical access, they can bring in their own gear unless you lock it down in the switch with MAC whitelisting &c [05:11] never really played with why, myself [05:12] I mean, what's the attack vector that they're trying to close [05:12] no clue [05:12] IME it's more incompetence or resources limiting things [05:12] my university had the problem every time the linux 101 course hit DHCP they'd bring the whole LAN down [05:13] qman__: certainly IME teaching staff are clueless about tech :-/ [05:13] because they were either too dumb or too cheap to put a router in that room [05:13] qman__: haha, and STP [05:13] qman__: I have several stupid customers that have STP-less SXs, and regularly they will plug the SX into itself [05:13] no more net [05:13] yeah [05:14] I really should bother with stp more [05:14] patdk-lap: like, at all? ;-P [05:14] my switchs will heal themselfs [05:14] but they bring things up and down for a few min before it settles down [05:14] you probably paid extra for name-brand switches [05:16] And not a name like "Hong's Lucky Dragon genuine IOS-compatibule" [05:16] all my cisco switchs died a quick death [05:16] So what, procurve? [05:17] never liked hp [05:17] atleast switchs [05:17] normally go with the netgear l3 switchs [05:18] only have had the poe models of those ever fail me [05:18] netgear's just cisco these days [05:18] no [05:18] that is linksys [05:18] Oh, sorry [05:18] Brain fart [05:18] and I would never call linksys switchs cisco [05:19] if it doesn't run ios it's not cisco [05:19] I think all catos is dead these days [05:22] I do use cisco switchs in the hp blade systems [05:22] only cause I don't trust the hp switchs at all [05:23] man, long past sleep time === scar[w]_ is now known as scar[w] [06:41] hey guys, im trying to set up some special hard drive mirroring stuff. I am trying to get hardive to mirror with an external hardrive, and idea how to set up this type of RAID 1? [06:47] T3CHKOMMIE: don't do that [06:48] twb, its an experiment for an OS I am testing. [06:48] T3CHKOMMIE: USB HDD enclosures are flaky enough that you'll be dumped into degraded mode regularly [06:48] degraded mode = won't boot without a human holding its hand [06:48] its not a boot drive. ive got that on another exter device. its just a simple partition im trying to key "synced" with the orignal on the pc. [06:48] ASIDE from that caveat, you should be able to simply tell mdadm to use /dev/sda and /dev/sdb, where they are the internal and external drivers respectively. [06:49] If you're just trying to keep *files* synced, use rsync, not RAID [06:49] is rsyn instantanious? [06:49] (Come to think of it, the slow speed of USB2 would kill you're I/O to the main disk as well, even with write-behind enabled.) [06:50] T3CHKOMMIE: no, rsync is a batch process [06:50] !rsnapshot [06:50] Stupid bot [06:50] ya im trying to get instantanious syncing. :( [06:50] T3CHKOMMIE: you aren't gonna get what you want [06:50] T3CHKOMMIE: not without some crippling tradeoffs wrt. speed and robustness, anyway [06:50] well then, looks like im back to programming this shit by hand. [06:50] http://rsnapshot.org/ <-- I recommend this [06:51] T3CHKOMMIE: if you're programming, then is this for a database or something? [06:53] its a data structure I am designing. and I am trying to get it off the ground. I am using a combination of a few hard drives, jump drives and NILFS to get a web server set up that can go down, come back online with only a few seconds of down time and almost no data loss. [06:53] so far i got it all working. i am just trying to get my internal media to be 100% cloned to my external media with in a minute of write. [06:54] i was thinking if i did a softRAID ( i know its clunky and risky) i would have exactly two bit for bit copies one one internal drive and one external that i could then move to hardware that is working and have no down time for my site. [06:55] I am almost 100% possitive im going to meet my goal. I just need to figure out how to keep my /dev/sdb1 and /dev/sdc1 identical... RAID1 seemed to be the obvious choise. [06:56] That sounds like EC2's job [06:57] EC2? [07:01] "der cloud" [07:02] Rapid provisioning, high availability, etc. [07:03] ya, its just something for my small site... i cant afford cloud space :( [07:21] Hello, I upgraded a xen vps jaunty server to karmic using alternate CD method, and now the server opens the maintenance shell opens, and it doesn't continue to boot. Any ideas what to do? [07:48] Sensiva: no reason given, just right into the maintenance shell? [07:49] SpamapS http://pastebin.com/fRfxE43N [07:51] Sensiva: ahh, looks like / needs an fsck.. [07:51] I did [07:51] Sensiva: it turns up clean then? [07:51] now aptitude tells that upstart is broken (unmet dependencies) [07:51] SpamapS yes [07:52] Not sure I ever really trust aptitude. ;) [07:52] still, so after an fsck, you were able to boot.. but then something else was broken? [07:54] I got the maintenance shell after upgrade reboot, tried fsck then reboot. no luck === jason is now known as Guest35720 [07:59] Sensiva: sorry to give you any false hope.. Its about time I went to bed. :p [07:59] :D [07:59] Sensiva: maybe try mounting the root fs next tho [08:00] it's mounted read only [08:00] g #macosx === Guest35720 is now known as jasef === jason is now known as Guest78728 [09:16] hi all [09:23] hola Daviey [09:40] Hello, I have installed phpmyadmin, mysql-server, apache and php5 on my server, how can I know what is the data base host address? it has to be the same ip address of the server right? [09:41] By default, it's configured to listen to all interfaces if I'm not mistaken. So any address the server has, it'll respond on port 3306 [09:41] But you can't log on remotely, unless you first configure users/hosts [09:43] ersi: I am trying to install wordpress remotly from another computer on the same LAN and this is the IP of the server and I connect to phpmyadmin using it 192.168.10.35 but when I type it as data base host it doesn't work! username and password are correct. [09:44] Yeah, for the same reason I stated above :) [09:44] you need to command line hax this boat up the creek first [09:45] ersi: How can I do that please? [09:46] Syria: That's a part of administrating a MySQL database. I'd recommend looking at: http://dev.mysql.com/doc/refman/5.1/en/adding-users.html [09:46] specifically the GRANT statement [09:47] ersi: Thank you. [09:48] ersi: I have done this on a my VPS before and I didn't face such probelms. May I know why did this happen please? [09:49] you probably followed a guide which told you do issue some GRANT statement :) [09:50] Or you had PHPMyAdmin on localhost, and logged on as root [09:50] as in on the VPS [09:51] I know nothing about coding and that grant thing scared me a lot. http://dev.mysql.com/doc/refman/5.1/en/grant.html [09:52] ersi: On my VPS I was installing wordpress from my GUI browser firefox. [09:57] hi there [09:57] i'm still having some issues with preseed. i just installed a server using this network preseed : http://pastie.org/2871469 and after installation the host is using dhcp. what's wrong? [10:00] ersi: I have created a new user and the problem is solved.. thank you :) [10:04] Syria: Awesome :-) === jason is now known as Guest2138 [10:17] morning o/ === Guest2138 is now known as jasef [10:51] New bug: #891071 in puppet (main) "puppet lucid backport doesn't support upstart" [Undecided,New] https://launchpad.net/bugs/891071 [10:55] for the life of me, i can't understand how to preseed a server using static ip settings. it's always using dhcp in the end [10:55] New bug: #891085 in postfix (main) "package postfix 2.8.5-2~build0.11.04 failed to install/upgrade: подпроцесс установлен сценарий post-installation возвратил код ошибки 75" [Undecided,New] https://launchpad.net/bugs/891085 [11:20] htaccess mod_rewrite is kicking my trash today. I have a fresh Ubuntu LAMP server. How do I get rewrite to work? [11:21] jehoshua02: 1. update the config to first enable the module then add the rewrite and 2. restart the server [11:22] xranby: https://help.ubuntu.com/community/ApacheMySQLPHP#Edit_Apache_Configuration [11:22] ^??? [11:22] jehoshua02: Error: "???" is not a valid command. [11:22] This one right? /etc/apache2/apache2.conf [11:22] I'm unfamiliar with the layout of apache2 config files. [11:23] thanks. [11:24] jehoshua02: http://httpd.apache.org/docs/current/mod/mod_rewrite.html [11:25] jehoshua02: can you paste what your rewrite rules look like? [11:25] I need to enable the mod first. [11:25] If I have problems with the rules I'll let you know. [11:26] jehoshua02: to enable the mod add RewriteEngine on [11:26] xranby: sudo a2enmod rewrite [11:28] hmm... [11:28] looks like that just creates a symlink in /etc/apache2/mods-enabled/, which is what I tried already. [11:28] k, so there must be something else. [11:31] jehoshua02: yes you need to update the configuration files [11:31] xranby: I think it's AllowOverride in the site default [11:32] jehoshua02: than you can create a .htaccess file with the rest of the configuration in the webservers www folder [11:32] this file needs to contains three lines [11:32] 1. RewriteEngine On [11:32] 2. RewriteBase /xyz [11:32] 3. RewriteRule ^oldstuff\.html$ newstuff.html [11:34] jehoshua02: have you restarted the apache server using apache2ctl restart ? [11:34] yep [11:35] hmm... still not working. [11:35] jehoshua02: double check that AllowOverride All [11:36] jehoshua02: can you paste what your rewrite rules looks like? [11:37] 1 Options +FollowSymLinks [11:37] 2 RewriteEngine On [11:37] 3 RewriteBase / [11:37] 4 RewriteCond %{REQUEST_FILENAME} !-f [11:37] 5 RewriteCond %{REQUEST_FILENAME} !-d [11:37] 6 RewriteRule ^(.*) index.php/$1 [11:38] if you have the 1-6 numbers in the config file then remove the numbers [11:38] xranby: the numbers are only part of vim configuration. [11:46] xranby: any other ideas? [11:46] jehoshua02: pasye any error mesasge you recive [11:46] and check your access log [11:46] 403 forbidden [11:47] can the webserver access the folder? [11:47] /var/www/ [11:48] yes. [11:48] inside that folder theres the .htaccess, index.php, and media/ [11:48] index.php and media/ are symlinks. [11:53] jehoshua02: tro figure out why you get a 403 forbidden.. it are totally unrelated to the rewrite rules [11:54] xranby: I know. Never said they were. [11:55] ah, I mentioned htaccess at the start. sorry. [12:27] Good morning [12:28] * spoonbow slaps spoonbow around a bit with a large trout [12:29] hi! i've got a cron that properly uses sendmail to send mails but when i use a php webform to send mails they go nowhere, is php using also sendmail for delivery? === gustav- is now known as beerbro [12:54] #debian +l 1177 [12:54] Mode #debian +l 1177 by debhelper [12:56] hi there [12:58] I am trying to rdesktop via a ubuntu server but i'm not getting a connection, I'm connecting to a VM via headless mode, the only thing i've changed on the host was the hostname, would this cause the rdesktop to error? I am still trying to connect via the IP [13:10] Guest9461: Just curious, is this a VirtualBox server? I have one setup. I had to make sure Guest Additions were properly installed to get rdp and other features working. [13:11] it's a ubuntu server and virtualbox is installed onto it [13:11] I then ssh into the server and launch VBoxHeadless [13:12] and then rdesktop into the server [13:12] Guest9461: Cool. I have the same thing. I use PHPVirtualBox installed on my laptop to connect to the server though. [13:12] oh? [13:12] PHPVirtualBox has a flash component that does the rdp. [13:12] nice? and what are the benefits to that? [13:13] Web-based management of VBox VMs. [13:13] Instead of having to remember all the command lines for Vboxmanage [13:13] eitherway, I just changed the hostname of the server, and the damn rdesktop command is no longer working, and I can't find out why [13:13] Oh, that's nice... [13:13] Guest9461: It's possible. [13:14] I just read it all up and setup the one box as it's a free (kinda) way of using windows on my server [13:14] hmmmm yea, wish I never did... silly move really... [13:14] Here's the steps I followed to setup everything, including PHPVirtualBox: http://www.grokensteins.org/2011/11/ubuntu-1004-lts-virtualbox-412.html [13:15] Guest9461: But about your host name problem... [13:15] * jehoshua02 thinking [13:15] Guest9461: Can you connect to the server via IP instead of hostname? [13:15] nope, always used the IP [13:16] i can ssh in fine via the terminal [13:16] and even run the virtual machine via headless mode [13:16] Guest9461: then I don't see why changing the host name would break anything. [13:16] VBoxHeadless -s NAMEOFMACHINE [13:16] but rdesktop isn't connecting into it... [13:17] it's not a firewall, as i've checked and also checked on different connections [13:17] Guest9461: you were able to rdp before correct? [13:17] i've just done a sudo apt-get dist-upgrade and am updating virtualbox to see if that makes a difference [13:17] yes always [13:17] Guest9461: Have you installed the extras? You need to have those to get VRDP [13:17] The regular package doesn't have the extras with them [13:17] VRDP == extra [13:18] Guest9461: yep, maybe try reinstalling Guest Additions. [13:18] yes the extras were always installed as I was rdesktop'ing into it fine earlier [13:18] okay brb [13:18] but I know the extras are installed [13:18] Have you upgraded virtualbox? In that case, reinstall the Extpack [13:18] Guest9461: ah, yes, right, there's guest additions, and there's the extras. Perhaps you'll need to reinstall one/both. [13:19] I had this trouble when going from 3.X to 4.X [13:19] as the VBoxHeadless -s NAME command is bringing up "listening on 3389" [13:19] Yeah, but it doesn't bind to it right? If you netstat [13:19] so it's listening, just not connecting.... [13:19] I was quite confused when that happened as well [13:20] it's odd as I was in it earlier today, and I just decided to change the server hostname, and that's it, so the only thing I could have thought of [13:20] Hm. [13:20] unless the crappy XP vdi broke? [13:20] it's an XP vm [13:20] unlikely, you should atleast get a blackscreen RDP in that case I think [13:21] I just get denied [13:21] I'd try #vbox on freenode, those guys kick ass at debugging VBox [13:21] i've tried rdesktop via linux 2 machines and also via windows 7 on another machine [13:21] but I'd still recommend doing an extpack clean and reinstall [13:21] okay np i'll reinstall right now see if I can sort it and then if not i'll head there [13:21] thanks [13:23] also I did learn more jehoshua02 introduced me (just) to PHPvirtualbox [13:23] Yeah, I'm checking it out as well - it seems sweet. It's grown a lot more since last time I checked it out [13:24] Guest9461, ersi: it's figgin sweet. Incredible work. [13:24] Guest9461: I run PHPVirtualBox in WAMP on my laptop. [13:24] i guess it's all getting more popular, I was thinking at the time I learned this I was one of the few people in the world crazy enough to do it all this way... guess it's clearly not the case lol [13:25] Guest9461: Actually, I think that us crazy people just know how to find eachother. [13:25] IRC!!!!! [13:25] lol :) [13:26] well... xchat for me ;) lol [13:27] It's still using the IRC protocol ;) [13:27] yea I knew i'd get that as soon as i hit enter... i was always memorised from an early age by the MIRC logo and the smiley lol [13:28] yeah, I figured :) [13:37] I hate smileys :) [13:54] good morning [13:56] !ask | zul [13:56] zul: Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience [13:58] elz89: Did i miss where zul asked to ask a question? [13:58] Daviey: Yes, you and me both, buddy. [13:58] Daviey: I thought he may have said 'good morning'to check if anyone was around? [13:59] zul: g'day! :) [13:59] elz89: ah, no - This is zul's office, he walked in and said Hi. He is a respected member of the ubuntu server team :) [14:00] i dunno about respected :) [14:00] hah [14:00] Ahh sorry I am relatively new, zul, my appologies. [14:01] And I have just come from spending some time watching over #ubuntu and trying to help people. You know how heated it gets over there. lol. [14:01] Yeah, hehe [14:02] elz89: Welcome!@ [14:03] Daviey: what is that? [14:03] a 'welcome' with a typo before i pressed enter. [14:07] Daviey: thank you, I am not going to be very valued, as I am still a pup. I have been into linux for a couple of years now. I run my own homebrew stuff, and I have just finished converting my Mum, Step Dad, Little bro and Little sisters computers to Ubuntu, (sis runs Lubuntu - crap PC), then they have a ubuntu server there backup up, media server, dhcp, dns, ntp etc... still playing around and trying to learn and servers. [14:07] Daviey, hey, last night i did the syslog-ng merge request just bc i didn't have perms to push it to the archive myself. if i prefer i can sling you a debdiff (though it'll be huge, since its an upstream merge) it tests fine. [14:07] I have no idea why I just told you that but I guess I wanted you to get I feel fior what I am doing right now. [14:08] elz89: thank you for exposing yourself. let us know if you are interested in helping the ubuntu server community [14:09] elz89: Welcome :-) I've had something similar poppin' before [14:12] pmatulis: I am very interested, been looking into some documentation I would like to upgrade a little. I don't know where to start really, so I just try to help on IRC, and just set things up and tests things out. I also do some objective-c stuff in xcode, love iOS :-P [14:12] I am trying to get everyone I can onto Ubuntu and it is working! :-) [14:13] My mum says that she is making huge savings in electric alone, now that the whole house in running this great OS. [14:13] elz89: let me know if you want to help out with the ubuntu server guide, i'm looking for technical reviewers [14:14] pmatulis: you are on ;-) [14:14] feel free to PM. [14:18] hallyn: debian -> ubuntu debdiff might be easier. [14:18] sure. or a udd merge request :) do you want the debdiff? [14:19] grr. ctrl-w in xchat [14:19] Daviey, lemme know what you want :) I can send you a tarball of the .dsc/etc like i used to too, but i *know* you hate that :) [14:19] hallyn: debdiff works for me.. I'm sure it'll be great.. I prefer UDD, if i'm likely going to ask for changes.. But this will be a perfect debdiff, right? :) [14:20] Daviey, of course :) [14:21] apart from two weirdness i had to change, it's almost a sync [14:21] ok, i'll post a debidff, biam [14:23] Daviey, http://people.canonical.com/~serge/syslog-merge-3.3.1.debdiff === Ursinha is now known as Ursinha-lunch [14:32] hallyn: assume that is because systemd isn't in main? [14:32] that package doesn't exist in precise [14:33] my testing, fwiw, consisted of forwarding syslog msgs from one canonistack instance to another [14:34] hallyn: That is Great to hear! [14:42] hallyn: you know it will build-dep wait on, libevtlog-dev ? [14:42] (it's universe) === jasef is now known as semicolon [14:45] hallyn: If you believe it to be required, i'll upload.. but can you raise and follow up on a MIR? [14:45] If it doesn't add benefit, can we drop it as a build-dep? [14:46] Daviey, thanks, no i didn't notice that. I don't know what it is or if we need it, will get back to you [14:47] Daviey, well the changelog entry was: * Build-Depend on libevtlog-dev (>= 0.2.12-5~) for correct shlibs. [14:49] grr [14:49] I'm afraid it may need to be MIRd [14:49] Daviey: I'm checking the mcollective debdiff descriptions and they're right I reckon http://paste.ubuntu.com/740266/ [14:49] Daviey, no, syslog-ng is in universe [14:50] lynxman: Did "Description: build server orchestration or parallel job execution systems" come from Ubuntu or Debian? [14:50] hallyn: you are kidding me? [14:50] nope :) i guess rsyslog is out default? [14:50] maybe we should MIR syslog-ng [14:50] hallyn: quite right, my bad - sorry. [14:51] Daviey, is that something lintian would have warned about otherwise, or not? [14:51] (cause it did come up empty) [14:51] hallyn: sadly, not. [14:51] good to know [14:51] somethign ineed to be more careful about in the future (bc it didn't occur to me) [14:51] thanks [14:52] hallyn: Good, o - Uploading, looks perfect and builds here! [14:53] Daviey, and on a related note, the libvirt i was wanting to upload now build-depends on libnuma (popular user request), but that is in universe [14:53] GAH [14:53] hallyn: have you figured out the gnome-keyring stuff? [14:53] hallyn: really popular? [14:53] zul, i opened a bug on it, yes [14:54] Daviey, not sure. but the guy who last wanted it wanted it so bad he went and fixed it through debian [14:54] hallyn: I wonder what people are using it for. [14:54] dunno, but they'll be turning into rabid anti-ubuntu pro-debian fans soon [14:54] Daviey: From debian [14:55] zul, bug 888119 [14:55] Launchpad bug 888119 in juju "fixes for new error interface" [Undecided,In progress] https://launchpad.net/bugs/888119 [14:55] Daviey: if you don't fancy it we can change it as well, but I find it an unnecessary point of delta [14:55] lynxman: aaaaahhhh... sorry, i thought that was our strings [14:55] no that's not it [14:55] Daviey: no no :) [14:55] lynxman: agreed, as close to possible as debian as possible on this. :) [14:55] heh. zul: bug 888199 [14:55] Launchpad bug 888199 in p11-kit "spurious warning about gnome-keyring-pkcs11.so" [Undecided,New] https://launchpad.net/bugs/888199 [14:55] hallyn: i would be concenered :) [14:55] Daviey: so if everyone agrees on the activemq switch I think this one is golden by now :) [14:55] zul, opinion on libnuma? [14:55] hallyn: do it [14:56] MIR it? [14:56] lynxman: lets upload that baby. [14:56] Daviey: \o/ [14:56] feh. there are already 10 MIRs to do for $(*&%(*$&% spice [14:57] zul, 'do it' meant MIR libnuma? [14:57] hallyn: yeah i dont see the problem with it, if people want it then why not [14:58] zul, ok. for this first merge though i guess i have to turn it off [14:58] (merge is otherwise ready, wiating to build in ppa and on some more testing) [14:59] * hallyn back in an hour [15:03] AHA! [15:04] ahs3, you uploaded numactl; would you have any concerns about MIRing it? [15:04] anyone here familiar with preseeding? i'm having issues with the network config. it always uses dhcp, whatver i do [15:06] hallyn: hrm, that was in a previous life :). no, i think it would be fine; the maintainer is pretty active and it seems like reasonably stable code. [15:07] ahs3, ok, cool, thx. (now, really, bakc in an hour :) [15:16] bwahhhhh internet connection === semicolon is now known as obelus === Tm_K is now known as Tm_T [15:38] rbasak: hey, if you need additional pandaboards, I suggest getting smoser and/or roaksoax to send you theirs....as the boards are insanely backordered right now [15:39] i can send one to rbasak. i haven't used mine. i dont *want* to give it up, but i think it makes sense. [15:40] great, thanks! [15:40] I'm not sure what I'll be doing with them yet, but I guess I'll need them for juju testing! === huats_ is now known as huats [15:48] How can I attach more than one client to a 'screen' ? I run irssi on a server, and I want to be able to mimic both sessions through my tv, and my netbook in kitchin, as I move around my flat. Sad I know, but just always wanted to know how it could be done? [15:48] elz89: screen -x [15:50] andol: so I need to relaunch this because at the moment I do 'screen -S irssi && irssi' [15:51] So I need 'screen -x -S irssi && irssi' ? [15:51] then I can ssh from tv and netbook, and 'screen -r irssi' on both clients? [15:51] infact I should just 'man screen' and have a better look shouldn't I? lol [15:56] andol: will they mimic eachother, or should I pi** off and have a play? [15:56] elz89: screen -x replaces screen -r [15:57] elz89: Of course, if you don't need to open at different places at the same time you can just do a combined detached and open; screen -dr === cloakable_ is now known as cloakable [16:03] andol: I would like to attach to ssh'd clients to the same 'screen' [16:08] elz89: the answer you seek should be in the man page. [16:14] greppy: yeah sorry having a lazy moment, I guess it was because its really important. I was feeling lonely :-P [16:17] "Attach to a not detached screen session." should say "Attach to a currently attached screen session." maybe? lol [16:21] does anyone argee with me that if it was written that way, it would make more sense? [16:23] New bug: #891204 in libvirt (main) "Essential configuration in /etc/libvirt/qemu.conf commented out" [Undecided,New] https://launchpad.net/bugs/891204 [16:25] At the bottom of 'man screen' it says "A weird imagination is most useful to gain full advantage of all the features." === onre is now known as Guest13624 [16:28] Daviey, syslog-ng, the .xz suffix is the only problem, right? is that an ubuntu archive-specific thing? [16:31] New bug: #891227 in nova (main) "nova.conf force_dhcp_release doesn't take any parameter" [Undecided,New] https://launchpad.net/bugs/891227 === z30 is now known as KaZeR_W_ === KaZeR_W_ is now known as KaZeR_W__ [16:34] Daviey, (separately) filed bug 891232 [16:34] Launchpad bug 891232 in numactl "[MIR] numactl" [Undecided,New] https://launchpad.net/bugs/891232 === KaZeR_W__ is now known as KaZeR_W [16:36] New bug: #891229 in nova (main) "nova-manage_flagfile_location.patch identation error" [Undecided,New] https://launchpad.net/bugs/891229 [16:38] SpamapS: could VoltDB packages be put in universe? I think they're GPL3 [16:40] hallyn: *sigh* i assumed Launchpad now had that support. [16:41] https://wiki.ubuntu.com/MOTU is a dead link [16:42] Daviey, well it complains about the dpkg version or something [16:43] Daviey, upload fails says INFO Require Pre-Depends: dpkg (>= 1.15.6) when using xz compression. [16:45] (it's too bad it requires further delta against debian, but...) [16:45] Daviey, do you want to add that in-line, or do you want a new debdiff? [16:46] hallyn: I think Launchpad needs a newer dpkg version before we can use the native package [16:46] I think it needs a re-pack of the orig tarball for now. [16:46] *sigh* [16:46] Daviey, are you sure a 'pre-depends' wont' work? [16:47] hallyn: No, i'm not certain.. can i suggest uploading to a PPA to see what happens? [16:47] ok [16:48] (not convinced the ppa will fail/succeed the same way though) [16:48] can i put pre-depends on the source pkg? [16:49] no wait that doesn't even make sense. Pre-depends is about before the binary package is installed [16:51] yowza, crashed xchat. i think it's time to switch [16:54] Daviey, is there any downside to just repacking? [16:54] i had been working with the result of 'bzr bd -S -- -sa -vxyz' and it had jsut created the .orig.tar.gz for me [17:09] would it be possible to set up a vpn server remotely? [17:14] m_3: haven't looked at VoltDB [17:16] m_3: does look like that would work though [17:16] SpamapS: helping them own the volt charms (from hadoopworld)... they were asking about packaging too [17:20] hallyn: no downside, just more work. [17:20] I can toss you the built package ify ou prefer. (otherwise, just take my bzr tree and bzr bd -S it) [17:21] New bug: #891252 in euca2ools "--config option is not honored" [Undecided,New] https://launchpad.net/bugs/891252 [17:23] hallyn: I'm going for a lunch shortly... but will pick it up after? [17:23] * honored [17:23] Daviey, ok [17:28] m_3: yeah, the best path to the distro is really to setup a PPA and then ask for a review of the packages inside it. :) [17:28] ah, gotcha [17:29] SpamapS: perhaps the "VoltDB Charmers" team ppa? [17:30] Daviey: you can either grab the bzr tree and build, or grab http://people.canonical.com/~serge/syslog-ng-merge-3.3.1.tar.gz [17:32] m_3: heh that would be cool [17:41] andol: greppy: quick question, any way to get round the fact I have different physical sized screens when using the 'screen -x' method? === RudyValencia- is now known as RudyValencia [18:02] zul: btw, thanks for uploading 2.2.2 with my patches! :-D [18:02] SpamapS: no problem all in a days work [18:02] Probably need to SRU some of those [18:02] SpamapS: i thought they made it into security [18:03] zul: oh maybe they did [18:03] not as yet, afaik [18:04] rbasak's branch is based on SpamapS.. SpamapS would you might verifying you are happy, and we can ask for sponsorship. [18:04] s/might/mind [18:04] launchpad says no [18:04] i dont think SpamapS needs sponsorship anymore :) === Ursinha-lunch is now known as Ursinha [18:08] zul: I think he does to -security :) [18:08] Daviey: ah the security team got their own ways of doing things [18:08] yas. [18:26] hello: I'm getting this error. W: GPG error: http://downloads-distro.mongodb.org dist Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 9ECBEC467F0CEB10 [18:26] I'm trying to install mongodb and from their page http://www.mongodb.org/display/DOCS/Ubuntu+and+Debian+packages#UbuntuandDebianpackages-gpgkey it says to do sudo apt-key adv --keyserver keyserver.ubuntu.com --recv 7F0CEB10 -- but that fails [18:27] so I get that it needs the public key... but any ideas how i can get the public key and manually install it? [18:27] or maybe even tell ubuntu to not worry about this particular repo [18:34] robo: the Key exists in the system. [18:34] When you do sudo apt-key... what do you get? [18:34] s/system/keyserver/ [18:34] Resistance, i get an error that it couldn't find it [18:34] sec, i'll pastepot [18:35] oh, that worked from my laptop but not from the server... [18:35] firewall [18:35] * Daviey ponders offering avahi publication service for cobbler [18:36] Resistance, I take it that apt-key doesn't use port 80. I get this error from the server: gpgkeys: HTTP fetch error 7: couldn't connect to host [18:36] nope it doesnt [18:36] at least not afaik [18:36] * Resistance runs some checks [18:38] okay, so i can simply grab the key then import it using apt-key from a file [18:39] so on my workstation i imported the key. Is there a way to dump that key into a file? This way I can import it on the server? [18:39] Or is there a better way to do this? [18:39] the better way is to use apt-key :/ [18:40] and yes, i read your previous statement :P [18:40] I know there's a way to set the default editor system-wide using update-alternatives, but is there a way to do this per-user? [18:45] savid, the EDITOR environment variable? [18:46] robo, yeah that's what I thought of doing, but apparently visudo doesn't respect that variable [18:46] normally it's EDITOR or VISUAL [18:47] I used to always set both [18:49] kirkland: ping [18:49] roaksoax: yo! [18:49] kirkland: yo! so I was testing the autoreloding thingy for powernapd and it works like a charm [18:49] kirkland: but that's only for the "general" config [18:50] kirkland: do you think it should also be for the monitors? [18:50] roaksoax: yeah, i think so [18:50] roaksoax: but again, this is going to be *very* expensive [18:51] kirkland: kirkland yeah, but since it is checking the time of modifucation of the config file is not that expensive anymore [18:51] Resistance, any suggestions on what i can do? [18:51] roaksoax: hmm, yeah [18:52] kirkland: though for now I'll just leave the general configuration autoreloading and look into the monitors later [18:52] robo: perhaps temporarily deactivate the firewall while you get the key? [18:52] Resistance, yeah, that's not going to happen :-) [18:52] although that restrictive of a firewall not allowing you outbound is pretty harsh [18:52] robo: then i got nothing. there's pretty much no *easy* way other than using apt-key... *shurgs* [18:52] * Resistance has a meeting he has to get to anywayws [18:52] we have proxy servers that allow port 80. Apparently apt-key doesn't use port 80 or it doesn't respect Acquire::http::Proxy [18:52] roaksoax: cool === Guest13624 is now known as onre [18:53] okay, ty for your time Resistance [18:53] I set EDITOR=/usr/bin/vi and VISUAL=$EDITOR, visudo doesn't seem to use it though. oh well. [18:53] kirkland: now, for integrating bryce's modules/services, I was thinking on 1. either writing a new binary that disables services/modules or 2. script under /etc/pm/power.d. What do you think? [18:53] savid, it's probably something with sudo, not visudo [18:53] * Resistance hates systems redesign meetingns [18:53] namely cause i have to implement the design changes >.> [18:53] [18:53] enjoy [18:55] roaksoax: cool [18:55] roaksoax: i thought we have a script that enables/disables services/modules? [18:56] kirkland: yeah bryce's script. though it needs adaptation to support pm-powersave approach if we place it on /etc/pm/power.d [18:56] roaksoax: right [18:57] roaksoax: so it should be a wrapper script in /etc/pm/power.d === med_out is now known as medberry3 [18:57] roaksoax: that looks at our config file [18:57] roaksoax: and enables/disables those modules/services [18:57] roaksoax: probably two separate scripts, one for modules, one for services? [18:58] kirkland: ok, I think that sounds better [19:02] ahs3, sorry, just to be sure i understood right, you're just saying rename netcf1 to libnetcf1 and netcf-dev to libnetcf-dev, right? (or did i also get the contents wrong? hmm, i'll take another look) [19:05] roaksoax: cool [19:05] hallyn: you've got it right [19:05] ahs3, thx [19:06] hallyn: another question occurred to me -- have you run piuparts on the package? [19:06] never heard of it [19:06] (that's a no :) [19:08] interesting [19:09] trying to think how i'm gonna remember that name. what on earth does it stand for? [19:09] heh. i need to understand it better myself, but it's a nice QA check on a deb === RoyK^ is now known as RoyK [19:09] dude that looks aweseome [19:09] package-install-update-blah-blah ... i dunno :) [19:10] * hallyn goes to run that on the syslog-ng merge [19:10] cool. it gets run on the entire archive from time to time...kinda handy [19:11] * ahs3 wonders if ubuntu does that... ? [19:13] ahs3, i'm guessing not, because we have all those install/upgrade manual tests as apart of the qa testing before release [19:13] ahs3: It's been discussed, but not as yet. We've only just got a lintian service! [19:13] hallyn: nod. i'm poking around at the QA folks to see what'll happen :) [19:14] great black bear, or hornet's nest? now we'll know. [19:14] Daviey: heh. well, get on the ball! what are you doing chatting with me :-)? [19:15] ahs3: http://yes.sir.daviey.com [19:16] lmao [19:18] Daviey: do we have the different requirements and installation scenarios for orchestra/cloud documented anywhere? i dont see it on the whiteboard/blueprint. there was quite a bit discussed at UDS === adam_g_ is now known as adam_g [19:21] adam_g: it's still being pulled together [19:23] adam_g: We need to build on what we have already done, but also cater for the demo experience aswell as deployment [19:24] There was a thought to use the liveusb model for that, but i'm not sure that is that good. Do you have thoughts? [19:26] Daviey: right. i think the demo experience within a larger infrastructure will difficult to nail down, but if we've having everything laid out and defined, we can see what overlaps [19:26] liveusb = orchestra-live? [19:26] perhaps. [19:27] I'm not entirely convinced it'll be less work doing that TBH [19:27] i think a demo is a bonus if it can be done then great but.. [19:30] hmm, any idea why do I have to edit common-password after enabling pam-krb5 to get changing password working? [19:30] hi, i have problem setting up new install with 10.04, but the part manager dont set the "bootable flag" on, any idea? [19:31] (by default pam_krb5 entry is requisite and I have to change it to [success=2 default=ignore] or sufficient) [20:00] jdstrand: hey, what's the status of auth-client-config? It hasn't been updated since intrepid, nor pushed to debian and it overlaps some pam-auth-update features. [20:01] any human can help me? [20:01] kklimonda: yep. I don't do much with it. I keep meaning to fix the couple of bugs that are open against the sample profiles, but it hasn't been a priority. people should be using pam-auth-update. I haven't removed/demoted it cause people have told me they use it [20:03] 11.10 will easy up to 12.04 right? [20:21] Matrix3000: yes [20:22] !ask @ dub_ [20:22] RoyK: I am only a bot, please don't think I'm intelligent :) [20:22] !ask [20:22] Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience [20:28] Hm. !ask should maybe also say something about give us !details [20:28] !details [20:28] Please give us full details. For example: "I have a problem with ..., I'm running Ubuntu version .... When I try to do ..., I get the following output: ..., but I expected it to do ..." [20:36] RoyK, hi, i have problem setting up new install with 10.04, but the part manager dont set the "bootable flag" on, any idea? [20:37] at 16:30 hrs [20:42] RoyK: can you describe the storage config you are using? [20:43] ninjix: what storage config? the one for backups? for scientific data? for high-speed file/database access? [20:44] ;) meant for dub_ [20:44] want to know more about the disk arrangement [20:45] royk, the one for little availablility :) [20:48] if i need to install java openjdk-6-jre is recommended instead of sun-java6-jre right? [20:51] patdk-wk: heh - those 100TB boxes run well now, with better controllers and no FSCKING WD drives [20:51] all upgraded to megasas now? [20:51] Matrix3000: I find it depends on the how Ubuntu/Debian friendly the app developers are [20:51] or well, the sas2008 chipset? [20:52] ah ok [20:52] kirkland: hey do you have your slides from that power talk you did at some $conference? [20:52] I default to the Oracle/Sun when I'm hosting services or apps that cater to the entire Linux community [20:52] i think oracle is moving it from java-jre to openjdk anyways [20:52] I need to steal content for a lug meeting tonight [20:52] jcastro: yup [20:52] jcastro, wget mirror works well :) [20:53] http://openjdk.java.net/faq/ [20:53] jcastro: http://blog.dustinkirkland.com/2011/08/powernap-your-data-center-linuxcon-2011.html [20:53] ta [20:54] kirkland: didn't you have another too? [20:54] Yes. Oracle and IBM announced in October 2010 that we will collaborate in the OpenJDK Community to develop the leading open-source Java SE implementation, and make the OpenJDK Community the primary location for open-source Java SE [20:55] patdk-wk: lsi9201 [20:55] jcastro: yeah, i did linuxcon and linux plumbers this year [20:56] jcastro: and roaksoax did Texas Linux Fest [20:56] * jcastro nods [20:56] jcastro: you need the originals? [20:57] nope, I am all good [20:58] ninjix: what did you mean storage config? I have one server with 8TB on striped mirrored 1TB drives and some SSDs for read and write caching, it's blazingly fast - I have three backup target machines, two with 11 x 7-drive RAIDz2 and some spares, and one with an 8-drive RAIDz2, all 2TB hitachi deskstar [20:58] add another server for scientific data with a mix of RAIDz2 levels (I suddenly had to extend storage very quickly [20:58] nice [21:08] hi all [21:08] where can i find documentation about sudoers? including common configuration options [21:10] Resistance: man sudoers? [21:10] RoyK, grilled ;) [21:10] RoyK: without using man. i'm stuck on a windows system [21:10] Resistance: ssh into a linux system or perhaps use google? [21:11] Resistance, on windows system you can install cygwin [21:11] Resistance: you can find it on the web, google "man sudoers" :) [21:11] !lart [21:11] LART has been disabled in this bot. [21:11] hrmf [21:21] roaksoax, jamespage referred me to you. thought you could help answer this question. the qa lab control the CDUs from the Enablement "Sentry Switch". We'd like Orchestra to manage the power. Does that CDU type match any of the supported types in this wiki: https://fedorahosted.org/cobbler/wiki/PowerManagement#WhatsSupported [21:30] jdstrand: and is there a tool similar to pam-auth-update we can use to modify nsswitch.conf ? [21:35] hmm, there is update-nsswitch proposed for debian, but they yet again got stuck at discussing "the right way to do it".. [21:35] over a year ago.. sigh [21:38] yeah, there is not [21:38] (afaik) [21:44] patrickmw: let me look into it [22:34] I want to use a compression proxy (Ziproxy) and I want it to also compress SSL'd Content as well, I use it over a ssh tunnel so SSL being in the clear is not a issue for me [23:29] jamespage: could you sync libnb-parser-java (bug 890504)? It's blocking netbeans from building. [23:29] Launchpad bug 890504 in libnb-javaparser-java "Sync libnb-javaparser-java 7.0.1-1 (universe) from Debian testing (main)" [Wishlist,Triaged] https://launchpad.net/bugs/890504 [23:46] how do you guys set up your network printers? [23:46] i have ubuntu 10.04 as my server. [23:46] do you use lpd://ip-address/passthru ? [23:46] or something else? [23:47] and do you use PCL drivers or postscript drivers? [23:48] andygraybeal: I always used cups and ipp:// urls [23:48] SpamapS, okay - that's is what i would like to do too. [23:48] why not lpd:// ? [23:48] andygraybeal: as to drivers.. typically PCL is faster as it is quite a bit less verbose. [23:49] yea, fast is what i need, precision is not. [23:49] andygraybeal: because ipp is more modern? No real reason. :) [23:49] SpamapS, okay awsome. thank you. [23:55] for some reason, when i go to setup the network printer - i select 'IPP' and it says "host" and "queue" - i type in 192.168.2.180 and the 'verify' button... doesn't become enabled, and neither does the 'apply' button -- i wonder if something is wrong. [23:55] i tried just now to type ipp://192.168.2.180 - and nothign still [23:55] im going through a fresh server install setting up a raid 5. if i understand this correctly i want to partition each drive with 2 physical volume for raid partitions. one small on and one "max", then set the bootable flag to on? [23:55] for each drive? [23:56] oh i'm using the printer console in the gui, maybe i should instead be using the cups website on localhost.