[00:00] can someone please find me a guide on how to set up automount to a specified directory [00:00] i only need read only ability [00:01] no. you find a guide [00:01] oh well, i tried [00:01] time to load up google [00:01] how can i prevent ubuntu server 8.10 to turn of monitor after inactivity ? [00:02] press a key every few minutes [00:03] is that the only way ? [00:03] techsupport: Comment out DPMS line in xorg.conf [00:03] techsupport: of course it wasnt the only way... i just felt like being an idiot [00:04] Xperiment62, how often do you feel that way ? [00:05] genii, that'll require xorg installed no? [00:05] techsupport: quite often... its a bad habbit that ive picked up from linux support channels [00:05] Kamping_Kaiser: Heh, yes. Sorry, forgot was in -server for a minute :) [00:06] genii, so should i comment it out or no ? [00:07] genii, :) [00:07] techsupport: Do you have some gui installed? (Gnome/KDE/XFCE etc) ? [00:07] genii, nope [00:07] techsupport: Then no [00:08] so i cant turn that off ? [00:08] techsupport: There's probably still some way [00:10] I'd look whereever the ttys are configured. [00:11] it used to be inittab, but 8.10 doesnt have that. perhaps some upstart file? [00:11] techsupport: Perhaps -x /etc/acpi/screenblank.sh [00:11] (chmod) [00:12] that file looks like it relates to xorg, but give it a go [00:13] i dont get it , what do i do ? [00:13] -x ? [00:14] non executable [00:14] techsupport: No. Try this: sudo chmod -x /etc/acpi/screenblank.sh [00:15] genii, i dont have /etc/acpi/... [00:15] techsupport: Hm, OK. No other immediate ideas come to mind then [00:15] techsupport, tried running `apropos blank screen` ? i have a result called "setvesablank (8) - Turn VESA screen blanking on or off" that may or may not be useful to you [00:16] Kamping_Kaiser: Nice find [00:16] why does it say "The following packages have been left out" and lists linux-server/image? [00:17] genii, makes sense to see whats already available :) [00:17] hmmm [00:18] yays, it installed properly [00:20] Xperiment62: :) [00:22] nijaba: okay, i've merged screen-profiles [00:22] nijaba: i made a few adjustments [00:23] nijaba: i commented out the welcome screen from the default windows file, for the moment [00:23] nijaba: i hit a few problems with it that i'd like to get resolved first [00:23] nijaba: mainly, it doesn't look/work right on some window geometries [00:23] nijaba: also, i want to work on the keybindings a bit more [00:24] nijaba: i confirmed some of jdstrand's concerns [00:24] nijaba: i think we need to have a policy that a given keybinding must work in both a) gnome-terminal, and b) the console before we put it in keybindings/common [00:24] cxo: because those packages require additional new packages to be installed to perform the upgrade. after you've upgraded the rest of your packages, do a dist-upgrade and it'll show you what new packages it's installing to perform the upgrade [00:25] nijaba: i think we can add others to something like keybindings/xterm, keybindings/gnome-terminal, keybindings/konsole, keybindings/console and so forth [00:25] nijaba: if they only work in one place or another [00:25] nijaba: but i think if something lands in keybindings/common, it should be pretty portable [00:54] how do i set up automounting easily? [00:55] vorian: I'm sort of here now. Should be around in an hour or two. [00:56] ScottK: no problem, i pm'd you a link to look over [00:59] Xperiment62: fstab/autofs/pam_mount [01:27] how do you remove old kernels that are still on the system? [01:27] dpkg -r [01:28] aptitude remove [01:28] apt-get remove [01:31] how do i list the pkg names of all the kernels installed? [01:34] cxo: dpkg -l | grep linux-image [01:34] thanks a lot guys [01:35] i'm new to this apt/deb business [01:44] I have just installed a sata card in our server and need to hotswap drives with it... I found and installed the package scsiadd, but all of the scsi ID's say 0. [01:44] Is there a way to generate scsi ID [01:59] i get a message at boot from the 8139cp module saying that the hardware needs the 8139too driver instead, which after login is already loaded and ready to go, can i just blacklist 8139cp? why does it keep trying to load [02:00] Just blacklist it [02:02] i put it in /etc/modprobe.d/blacklist but it still loads it at bootup [02:03] Thanks mathiaz [02:05] jdstrand: nijaba: okay, updated screen-profiles uploaded to my ppa, and back onto the queue for universe [02:06] nijaba: we need to talk about the keybindings a bit more tomorrow [02:15] how do you get it to save the iptable rules? [02:23] maybe can anyone recommend me a VPS service provider in U.S. East ? [02:24] linode [02:24] hads: do they offer Ubuntu Server? [02:24] Yeah [02:25] linode rocks [02:27] hads: thanks [03:02] does anyone know of VPS provider around virginia or in virginia ? [03:52] Can I add disks to a RAID5 array? [03:52] As in, go from a 4disk raid5 to a 5 disk? [03:57] Yeah should be able to [04:00] http://linux-raid.osdl.org/index.php/Growing [04:07] wow [04:07] ZFS can't do that [04:09] ok i'm logged in into my new linode ubuntu 8.10 32 bit server, but nothing seems to work i guess all the packages are not installed? is it possible for me to update it with one command for it to be equivalent to an installation i would get from a regular ISO install ? [04:43] what's the way to restrict user UIDs being exposed to others? 'top' and 'ps' can be restricted on BSD systems by passing '0' to 'security.bsd.see_other_uids' in src/etc/sysctl.conf. what can be done for linux? [04:47] compengi: "exposed" ? [04:47] Oh, you don't want user to be able to see another user's uid? [04:47] Why's that? [04:47] to prevent harvesting? [04:48] to prevent them picking on each other :D [04:48] each user likes his privacy [04:49] i like root's too ;) [04:49] Give them each a virtual machine ;-) [04:49] nah.. [04:49] if this can be done simple as on bsd [04:57] compengi: Linux != BSD [04:57] reading forums about changing computer name, they say its in /etc/hostname but i dont have this file [04:58] techsupport, Then create it: it ought to exist. [04:58] techsupport: what happens if you type "hostname pancake" ? [04:59] ball, i know that for sure, that's why i'm asking around if there is a similar way on linux [04:59] compengi: fair enough. It's an unusual requirement though afaik. [05:00] techsupport, using ubuntu? [05:00] what is your hostname currently? [05:02] techsupport, yes ubuntu [05:02] blahnana, ubuntu [05:02] and what do you get if you type "hostname" [05:02] by itself [05:02] it should output the current hostname [05:03] unless you typed 'hostname pancake' already [05:04] I like pancakes :-9 [05:04] ...even American ones [05:08] mmmmm pancakes [05:08] see?! [05:10] ;) [05:10] when i type screen i get error Cannot open your terminal '/dev/pts/0' - please check. [05:10] techsupport: then don't type "screen" [05:11] why not? i need it [05:11] Oh. [05:15] sounds like you might be using su [05:15] or at least, that's one common cause [05:16] blahnana, thanx, let me try to logout [05:16] if you use su, make sure you use something like su - [05:17] perfect [05:17] worked [05:17] or sudo [05:17] Or launch screen, and use su inside the screen session. Running screen as root can be dangerous. [05:19] another thing [05:19] paul@linode:/$ sudo chown orudie /home/orudie [05:19] sudo: unable to resolve host linode [05:20] i recently changed hostname by 'sudo hostname linode' [05:20] that just changes an environment variable I think [05:20] you haven't changed your hostname [05:20] not sure if hostname actually does anything else [05:21] looks like it does, but clearly the change hasn't stuck [05:21] blahnana, but how can i change the hostname? [05:21] you need to put your hostname in /etc/hostname [05:21] echo 'linode' > /etc/hostname [05:21] reboot and check it comes up with the correct hostname [05:21] persia, I'm not sure if either is better than the other actually [05:21] /etc/hostname does not exist [05:22] techsupport, The command above will replace it. [05:22] (or create it if it doesn't exist) [05:22] if screen is vulnerable, you could get control of root's screen [05:22] and have a root login [05:23] but if screen is vulnerable, you could get control of a user's screen, and still have a root login [05:23] just did echo 'linode' > /etc/hostname , and /etc/hostname still doesnt exist should i create it ? [05:23] I think the issues with screen tie back to old issues with it [05:23] blahnana, At a deep level, it matters, because one might not be using "files" to determine the hostname, and there are various ways to tweak things. /etc/hostname is used at boot to run hostname. After running hostname, one often needs to reinitialise sessions and services to use the new hostname, as they may only run gethostbyname() at startup. [05:23] techsupport, did you run it as root? [05:23] blahnana, yes [05:23] techsupport, echo linode | sudo tee /etc/hostname [05:23] techsupport: "sudo echo 'linode'>/etc/hostname" [05:23] ? [05:24] ah, what persia said [05:24] ball, Won't work, because the redirect doesn't inherit sudo. [05:24] persia: thanks, good point. I'm not used to sudo [05:24] if I use it, it tends to be "sudo su" [05:24] oops [05:24] ball, It's *much* safer: definitely worth playing about with. [05:24] "sudo /bin/sh" [05:24] persia, I meant "I'm not sure if one is better than other... running screen as root or running screen and then opening a root shell" [05:25] ball: 'sudo -i' is the nicest way to do that [05:25] persia: I usually live somewhere sudo doesn't exist [05:25] jtaji: thanks [05:25] still dont exist [05:25] sudo -i seems better to me than sudo su [05:26] I meant "sudo sh", "su" was a typo [05:26] but, that's still handy to know why there are limitation with changing the hostname on the fly [05:26] so if i create it , what do i put in /etc/hostname ? [05:26] blahnana, running screen and then running programs as root is safer than running screen as root, because there's no possibility to attach directly to the root process from arbitrary input: it is mitigated by screen (which doesn't have a lot of security, but that's a separate thing). [05:27] root@linode:/# echo linode | sudo tee /etc/hostname [05:27] sudo: unable to resolve host linode [05:27] linode [05:27] root@linode:/# [05:27] techsupport: you need to edit /etc/hosts also [05:27] or you break sudo [05:27] persia, not sure what you mean [05:28] blahnana, Right. screen can spawn stuff, so if you're running screen, I can attach to your screen (yes, screen allows sharing), and spawn another shell, and do stuff, and detach, without you seeing it. [05:29] If you're not running screen as root, I can't get root by doing this: only access to your user stuff. I'd have to switch to control of your root shell, which I hope you only have open when you're actively doing stuff. [05:29] jtaji, sorry i never edited /etc/hosts what should i change there ? [05:29] Also, since screen supports multiple sessions, if one doesn't run screen as root, one can run a mix of root and non-root processes from screen. [05:29] we were weighing up the merits of one vs the other though [05:29] what modes does /etc/hostname need? [05:29] 644? [05:30] I suppose there's no need to hide it [05:30] jtaji, http://pastebin.com/m67e848de this is my /etc/hosts [05:30] so you'll know that there's a root shell [05:30] techsupport: second line should be something like 'xxx.xxx.xxx.xxx foo.domain.tld foo' [05:30] where xxx.xxx.... is your external IP [05:30] there's a great chance that you'll have access to my user screen anyway [05:30] it's more likely I might have used the password in an insecure way [05:30] whereas my root password will be more protected [05:31] most likely [05:31] so change 127.0.0.1 localhost to 207.192.72.15 linode ? [05:32] no leave that line [05:32] alone [05:32] Right, and if you're careful, you'll run a user shell, spawn root when you need it, and go back to user before you leave the shell unattended and detach. [05:32] haha [05:32] i'm a noob [05:32] well, I feel that we weren't talking about that [05:33] techsupport: http://www.linode.com/wiki/index.php/Configure_Static_IP [05:33] I agree with your point, but I think we're getting into operating procedure, which I perhaps assumed was beyond the scope of the initial comments [05:34] whereby I assumed that if you a) must have a root shell and b) must run screen you have already considered the things you're pointing out [05:34] OK. Then sticking strictly to screen, aside from operating procedures, there's also that it's easier to run a mix of root and non-root programs from a screen session if screen itself isn't running as root. [05:35] maybe, maybe not... if I'm running root screens I'd rather keep them together protected [05:35] And it's easier to establish operating procedures with redundant logs for mandatory unattended long-running root shells. [05:35] and not have a user having access to them necessarily [05:35] Personal preference, I suppose. Wouldn't be my recommendation. [05:35] lessens the chance that I do something stupid with my terminal [05:36] and I'd use my gui to split up the sets of terminals [05:36] yeah, personal preference [05:36] ok fixed [05:36] I think that if you're running root shells, and you've got them in a screen, I see it as better that you're running the screen as root in the first place [05:37] also, when i do logout it returns [05:37] paul@linode:/$ logout [05:37] bash: logout: not login shell: use `exit' [05:38] but in terms of which is more vulnerable... I wondered if there was a code reason for what you suggested, rather than just best practices type thing [05:38] techsupport: that's correct, use exit or ctrl+D [05:39] blahnana, Not code, just principle of least-privilege. [05:39] realistically, I'd think that your normal user account is more likely to be broken through running naughty things, or needed things that have bugs [05:40] I think that principle doesn't hold true when you're protecting a mechanism for full privileges behind a lower-privelege thing [05:46] blahnana, i archived a directory with tar -cvzf , but how can i extract it now ? [05:47] use x instead of c [05:47] use t instead of x or c if you want to see a list of files in the archive [05:48] tar -xvzf q2server.tar.gz [05:48] returns no such file or derictory error exit delay from previouse errors [05:49] blahnana, ping [05:49] is there a file called that? [05:49] in the current directory? [05:49] orudie@linode:~$ ls [05:49] q2server.tar.gz [05:49] orudie@linode:~$ tar -xvzf q2server.tar.gz [05:50] yes [05:50] try gzip -d q2server.tar.gz [05:50] does that work? [05:52] needed permissions [05:52] :) [05:52] ls [05:58] hello Techie, owh [05:58] sup dude [06:00] not much [06:00] yourself? [06:02] Hiya [06:03] not much here... just downing a bunch of tunez [06:09] I'm just playing a few that I just downloaded === erichammond is now known as ehammond1 [06:11] everythime i hear a good song i dont have, i grab a whole lot by the artist [06:12] I might do that if I had more time. [06:12] ...though I did with Airtone [06:12] i have 1041 mp3's [06:12] ...because I really liked their work [06:12] I don't know how many I have [06:14] i gotta chuck all mine onto a removable drive to plug into my server for LAN's [06:15] downside is that combined with video and ausio [06:15] audio [06:15] i have over 6 gig of media [06:15] and my largest thumb drive is 2 gig [06:15] 6 Gbytes isn't bad at all === ehammond1 is now known as erichammond [06:16] I've had 8 Gbyte stripe across two 4Gbyte flash sticks [06:16] its slowly getting larger [06:16] jbod ftw [06:16] brb [06:16] thats not including dvdrips and screeners [06:16] thats just general media [06:19] * ball nods [06:19] if we count all that stuff in [06:19] theres easily over 300gigs [06:22] * ball shrugs [06:22] I don't have any of that stuff [06:23] anyway, im gonna restart my computer and check this hard drive [06:23] if its not dead [06:23] i have 320 gig more internal storage [06:23] * ball nods [06:40] New bug: #314620 in logwatch (universe) "Please merge logwatch_7.3.6.cvs20080702-2(main) from Debian unstable (main)." [Undecided,Confirmed] https://launchpad.net/bugs/314620 [07:00] kirkland, nijaba, jdstrand: termcapinfo xterm* ti@:te@ was working for me, but just a bad workaround (when reattaching the screen again, the "backlog" when dicing is the different tab I used :/). So, I this there is a real issue about that [07:02] ok, this seems to be the intended behavior (reading the rest of discussion) :) === lukehasno1 is now known as lukehasnoname [08:06] moin [09:11] New bug: #314657 in samba (main) "smbd segfault in assert_uid " [Undecided,New] https://launchpad.net/bugs/314657 [10:10] hiya [10:10] can somebody please take a look at bug 181948? [10:10] Launchpad bug 181948 in exim4 "exiqgrep: error on messages w/o size" [Medium,Incomplete] https://launchpad.net/bugs/181948 [10:11] it's been sitting in the sponsoring queue for a while [10:27] hi [10:27] I am using Ubuntu 8.04 Server [10:27] I have already installed mysql-server-5.0 [10:27] I am faced with [10:27] http://rafb.net/p/M18ESh55.html [10:31] apparmour hates you, apparantly. [10:32] Well, not "apparantly", but "probably". [10:33] soren, ok [10:34] although i did not understand [10:35] Well, it might be something different. I just remember someone having problems with mysql because they wanted their mysql databases outside /var/lib/mysql [10:35] Can you pastebin the output of: [10:35] ls -ld /mnt/data/mysql/ /mnt/data/ /mnt ? [10:36] sure [10:36] Without the question mark at the end. [10:36] soren, please give me a moment [10:37] Sure. [10:37] <_ruben> if its apparmor interfering, that should apparent from the (sys)logs [10:37] Indeed. [10:37] I don't remember when we added apparmor profiles for mysql. [10:38] I think it was hardy, but ICBW. [10:38] <_ruben> i really need an update for my acronym engine .. this one took several seconds to figure out :p [10:38] It was hardy. [10:41] _ruben: :) [10:42] soren, http://rafb.net/p/iR8kDY93.html [10:42] kaushal: Ok, that looks fine. It's probably an apparmour thing, then. [10:42] Sorry, that's apparmour. [10:42] ok [10:43] Gah. apparmor. [10:43] That 'u' keeps sneaking in there. [10:51] soren i fixed [10:51] it [10:52] but when i start/stop mysql [10:53] How did you fix it? [11:08] soren, when i stop/start mysql [11:08] i get [11:08] http://rafb.net/p/rymbHy64.html [11:09] Well, yes. [11:09] You discarded all the stuff Ubuntu provides. [11:10] i provided http://ubuntuforums.org/archive/index.php/t-27156.html [11:10] but it still does not work [11:12] That's a completely different issue. [11:14] soren, i added in usr.sbin.mysqld [11:15] under /etc/apparmor.d [11:15] /mnt/data/ r, [11:15] /mnt/data/** rwk, [11:15] and then restarted apparmor [11:15] that issue got resolved [11:15] but now i am faced with this issue [11:21] soren, hurray it worked [11:21] i did it on my own [11:21] * kaushal patting my back :-) [11:49] hi all, I am looking to install a DLT drive on LTS. Guide i am looking at show to use mt-st... is this correct? === Shoopuf1 is now known as Shoopuf === Brimstar1 is now known as Brimstar [12:55] I installed a HighPoint 4 port sata card in our server last night, is there a way to make the drives hot swappable? [12:57] i doubt hpt is capable of hot swap [12:57] that's a 10$ raid controller, right? [12:59] It was like $200 [13:00] I found the scsiadd program and it works to remove sda, but that's it [13:00] 200$ for hpt? [13:00] last time i buyed hpt, it was very cheap :) [13:01] doing scsiadd -p to see the list; the list states all scsi drives are ID: 00 [13:01] there are like 10 diff models [13:01] hot swap should be supported by controler [13:01] i don't think you could do anything with controller that doesn't support that [13:02] hot swap = OS isn't aware of missing disks [13:02] so it doesn't know when it is out and back in again [13:02] but i could be wrong :) [13:03] ivoks, scsiadd -s rescans for scsi devices and adds them [13:03] scsiadd -r 0 will remove the 1st device [13:04] scsiadd -r 1 should remove the second, but all the ID's are 0, so I do not know how to, or if I can, assign ID's to the scsi drives [13:05] well, if you are using raid controller, you should see only one drive, right? [13:05] scsiadd is for adding/removing scsi devices [13:05] logical raid disk should be one device [13:06] so, if you have non-raid scsi controller that supports hot swap, you would use this tool [13:06] i'm not sure this should (or is designed to) work with raid controllers [13:06] The card has 4 ports with drives in each. The goal is to run 2 mirrored with mdadm, and 2 as singles for removable backup [13:07] ah... mdadm raid, not hardware raid [13:07] yes, I like the idea of being able to move the raid drives to another machine in case of problems [13:07] no controller required... The only reason for the card is to use the sata ports [13:08] well, still, controller should support hot swap [13:09] (i never used scsiadd) [13:09] i'm just reading this: [13:09] This is *NOT* a substitute for powering down to connect or disconnect hardware unless it's specifically designed to be hot swappable. [13:10] hi, i have a problem in choosing a virtualization technology. I'd like to go with kvm as i think it is the choice for the future, but in the few time i had for testing I noticed a consistent performance difference with xen. Probably it is only because of my configs, but i have too choose and have no more time for investigating. Can i have your points? [13:10] i always use kvm, when i'm not using vmware [13:10] recently, i droped vmware for kvm [13:11] ivoks, from HighPoint description: Hot swap and hot spare [13:11] ivoks: what about vm process cpu usage? [13:11] eolo999: vm uses cpu depending on what you are doing in vm :) [13:12] heath|work: then, hot swap should work :) [13:12] ivoks: when idle still my kvm machine was using ~= 20 % of Host CPU! [13:14] eolo999: mine uses 0-1 when idle [13:14] ****! [13:14] so it seems it's MY problem. [13:14] does anybody has ebook/PDF file with book title "Ubuntu 8 Server Administration and Reference" http://books.google.com.my/books?id=LhMFril7EBgC&dq=Ubuntu+8+Server+Administration+and+Reference+torrent&source=gbs_summary_s&cad=0 [13:16] a ovaj vostro ti se nije svidio? [13:17] http://www.kodeks.hr/Dell/vostro-a860,PID-3,P-1665,G-370.aspx [13:18] sorry [13:18] :) [13:19] wrong channel :D [13:30] how can I view the size of a dir? [13:30] du -hs [13:30] du -hs /bin [13:31] thanks [13:31] replace h with k or m or leave it out, depending on what you want [13:41] hi all, I am looking to install a DLT drive on LTS. Guide i am looking at show to use mt-st... is this correct? [13:44] you can use gnu mt too [13:45] (provided by cpio package) [13:48] ivoks: cheers [13:52] ivoks: i can see that the tape drive is available on /dev/st0 I have made no change to the system, is there anything else I need do? [13:53] selinuxium: nope, you can use it [13:53] selinuxium: there's also /dev/nst0 ;) [13:55] ivoks: I am trying to run a VM and connect the DLT to it... In VMware passthrough SCSI device all the physical SCSI devices start /dev/sgX... Any ideas? [13:55] ivoks: the host is Ubuntu [13:56] Clients, sadly Suse... but I am not at the Suse point... [13:56] selinuxium: i never did that, so i don't know anything about it [13:56] ivoks: m] [13:56] ??? [13:56] np! [13:57] exporting /dev/sg won't be enought though [13:58] yann2: ping [14:00] pong [14:00] hello ivoks :) [14:00] hi :) [14:12] ivoks: what is /dev/sgX connect to? how is it used? [14:22] selinuxium: that's generic scsi device [14:22] selinuxium: you can't use it as a tape device [14:23] ivoks: oh... And I cannot link sym link one to another then either... [14:24] vorian: Progress - Bug 312661 [14:24] Launchpad bug 312661 in db4.3 "Please remove db4.3 from Jaunty" [Wishlist,Fix released] https://launchpad.net/bugs/312661 [14:25] what do i need to do to upgrade from 7.10 to 8.04 lts? [14:25] install update-manager-core [14:25] and start do-release-upgrade [14:26] is it like upgading in one huge step or does it go trough every version? [14:26] ow [14:26] there's only one step from 7.10 to 8.04 [14:26] ScottK: I uploaded partimage last night, so db4.3 is clear for removal [14:26] ivoks, lol yeah that helps [14:26] vorian: Good, because StevenK just removed it. [14:27] ;-) [14:27] One down. [14:27] \o/ [14:27] ivoks, i really need to get better at this ubuntu bussines [14:28] a_ok: before the dot is the year 200*7* and 200*8* [14:28] a_ok: after the dot is the month... [14:28] so, 7.10 is 2007, november [14:28] 8.04 is 2008, april [14:28] ivoks, now that is goed to know. is there a steady release schedule? [14:29] every six months for regular version, and every two years for LTS [14:29] thanks [14:29] ScottK: so, what's the plan for mail stack? [14:30] :) [14:31] selinuxium: try modprobing 'st' module in guest [14:31] selinuxium: maybe it will discover tape, if it can see generic device [14:32] ivoks: Sadly it doesn't work that way. you have to define it first in the host then the guests can use it... [14:33] ScottK: it looks like db4.4 can also be removed [14:34] vorian: File a removal request then .... [14:34] okie [14:36] vorian: I think that during the Jaunty sync period from Debian a lot of stuff got move, so it's no so much left. [14:36] ivoks: First someone approves the spec and then maybe I find some time to guilt you into working on it. [14:36] :D [14:37] i don't mind working on that [14:37] OK. Glad to have you do it .... [14:49] ScottK: on the packages that depend on db4.6, should I bump them to 4.7? [14:55] New bug: #314623 in likewise-open (main) "likewise-open: allows lockout while disconnected" [Undecided,New] https://launchpad.net/bugs/314623 [14:57] that bot is not identified to services btw [15:09] vorian: I'd say let's wait until Jaunty +1 for that. [15:09] alrighty [15:09] ScottK: same for 4.2 then? [15:09] If we can clear out the older ones for Jaunty, that'll be a clear win. [15:09] alrighty [15:09] 4.2 is a bit special. [15:10] openldap (sladp) has severe performance problems with later versions of libdb. [15:10] ok then, i filed requests for 4.4 and 4.5. That will leave only 4.2 and 4.6 [15:10] So as long as openldap is stuck on 4.2, we can't get rid of it. [15:11] noted, I would be happy to talk to upstream about changing that :) [15:11] Working on porting the other stuff to 4.6/4.7 would be useful so that it can be removed once slapd is fixed. [15:11] vorian: It's a libdb problem, IIRC. Slangasek (who is also active in openldap maintenance in Debian) has been working on it. [15:12] I pinged him on #ubuntu-devel a bit ago to ask. [15:12] i see [15:12] how about cyrus nvi and kolab? [15:14] I think cyrus and kolab have to be done toghether. [15:14] I'd say look into it and do it if it works. [15:15] * vorian is on it [15:27] is it possible to have a usb drive in /etc/fstab mount only when available? [15:30] hi i'm having problems when pinging from my ubuntu i got !DUP - duplicates. happens only from that machine and it was not happening before. can somebody be so cind to help me out. [15:30] kind* [15:36] espacious, did you try booting from the live cd and trying like I suggested yesterday? [15:37] hi heath|work no not yet, but i excluded the switch.and looking now on my ubuntu setup [15:37] im also getting the live cd to test... [15:40] what else is possible to be? [15:41] howly! [15:41] heath|work ! found out where the problem is! [15:41] ? [15:41] i disabled our VPN (win 2008 server) and no duplicates more. [15:42] ubuntu resides on that server. [15:42] is a vmware. [15:42] nat or host networking? [15:42] going to #win-server :D [15:42] host. [15:43] must be sth wrong in ruting and remote access on win server. [15:43] have an idea what? [15:52] espacious, what does route print have? [15:52] in windows [15:56] just a sec. [16:03] nijaba: hi, around? [16:04] kirkland: yepper [16:04] nijaba: hey, okay, i merged your tree [16:04] nijaba: and released [16:04] \o/ [16:04] nijaba: there's a couple of things i wanted to run by you [16:04] nijaba: for one thing, i juggled the keybindings a little bit [16:04] good [16:05] nijaba: can you pull the latest lp:screen-profiles? [16:05] doing it [16:05] nijaba: i think we need to establish a minimum criterion for keybindings to land in keybindings/common [16:06] nijaba: i would suggest that those requirements are works identically in a) gnome-terminal, and b) tty console [16:06] nijaba: but i'm open to ideas on that one, and perhaps adding KDE's konsole to the list [16:06] sounds faire [16:06] (without an e) [16:06] nijaba: additional keybinding sets should be added to other files in that dir, which indicate *where* they're known to work [16:06] :-) [16:07] coolset_gnome for example [16:07] nijaba: something like that, yeah [16:07] nijaba: and we'd work into your framework helper a way for users to turn those on/off at their discretion [16:08] right === jjesse_ is now known as jjesse [16:08] heath|work what do u mean? [16:08] nijaba: so, i found the F-keys to be the ones that seemed to work best across gnome and console [16:08] what do you mean? [16:09] nijaba: open keybindings/common [16:09] nijaba: moving from left to right .... [16:09] nijaba: F1 was already taken (help in gnome) [16:09] hey people. I am running hardy server edition in a xen virtual machine (fresh install of the virtual machine). When I try to update my packages, I get a segmentation fault when apt tries to setup module-init-tools [16:09] nijaba: F2 was open, so I made that "create new window" ... probably the most frequently used? [16:10] nijaba: F3/F4 were also unused, so I mad those PrevWindow and NextWindow [16:10] nijaba: that makes F2/F3/F4 all very tightly associated [16:10] nijaba: oh, and F5 being "kill window" [16:10] sounds good [16:10] apt output on http://pastebin.com/m4992b7c6 [16:10] nijaba: i know F5 is usually 'refresh'.... [16:11] yes, that's why I chose it that way [16:11] nijaba: unfortunate, but i think we just go with education on that one [16:11] nijaba: yup, i understand [16:11] nijaba: F6/F7/F8 are still available for our mapping [16:11] nijaba: and F12 is open [16:12] nijaba: i thought we'd try to see what else is *most commonly used* [16:12] nijaba: i was thinking about scrollback, but i'm undecided [16:12] nijaba: F8 is advanced help [16:12] nijaba: F9 is your welcome screen [16:12] nijaba: i was thinking F12 might be your welcome screen [16:12] nijaba: in that it's on the end, and easy to find [16:13] nijaba: but i couldn't get that to work immediately [16:13] my keyboard goes up to f19 [16:13] nijaba: whoa [16:13] I think apple did not know what to do with the extra space on their slim kb ;) [16:14] nijaba: :-D [16:14] nijaba: okay, cool, i need to update the help.txt accordingly [16:14] which is, btw, one of the best kb I have ever used [16:14] nijaba: i missed that last night, oops ... [16:14] eh, np [16:14] * kirkland likes his loud, clicky IBM keyboard [16:15] nijaba: i was actually thinking we might throw the detach onto one of the F-keys [16:15] nijaba: anyway, with only 3 more available to work with, i thought we'd be conservative with those [16:15] yes, might be simpler than my 3 key thing [16:15] nijaba: yeah, perhaps [16:16] nijaba: i think 2 and 3 key things are fine ... we should just put them in another file, outside of common [16:16] nijaba: and make them easy to enable/disable as a lot [16:16] but C-shift-del is still close to something people tend to know [16:16] nijaba: it's *really* close to ctrl-alt-del for my liking :-) [16:16] right, that's part of my plan for the next few nights [16:17] nijaba: okay, next thing .... [16:17] nijaba: windows/common [16:17] nijaba: i renamed that from default -> common [16:17] sure [16:17] nijaba: i commented out the welcome screen for this release [16:17] :/ [16:17] nijaba: i'd like to get it back in [16:18] nijaba: i had some issues with resized windows [16:18] nijaba: it didn't display properly [16:18] talk with newt... [16:18] ;) [16:19] heath|work http://pastebin.com/ma791262 [16:20] nijaba: yeah, i saw your math in there :-) [16:21] nijaba: well, it's still easy to get to, while we test/fix it, by hitting F9 [16:21] right [16:21] nijaba: and i'd like to get the functionality working where you can tell it "don't show me this again" [16:22] in my list as well [16:22] nijaba: then i think we're cool to turn it back on by default [16:22] nijaba: excellent [16:23] kirkland: still unsure what/where you changed about $profile [16:23] nijaba: oh, right [16:23] nijaba: i made a few changes there [16:24] nijaba: okay, looking at select-screen-profile [16:24] nijaba: i created a assert_symlink() function [16:24] nijaba: which, i'm fixing a bug in right now ... [16:25] nijaba: the -L test should be against $1 [16:25] ahasenack, right, I was wondering [16:25] nijaba: and i dropped the .screenrc-windows code, as it wasn't working correctly [16:25] ahasenack: sorry, autocomplete on ah! [16:25] ;) [16:25] Nicke: ok :) [16:25] oooops :D [16:26] nijaba: the listprofiles() function itself gathers the available profiles [16:26] nijaba: and that's called in prompt() and setprofile() [16:26] nijaba: we're still, unfortunately, using a global $SELECTED variable [16:26] nijaba: i want to get away from that [16:26] nijaba: but i couldn't do that easily in this iteration [16:27] ok [16:27] nijaba: however, at the very least, it's ALL_CAPS, which indicates global [16:27] nijaba: i also replaced all indentation with tabs [16:27] nijaba: and I added your copyright statement to the top of those other two helper files [16:27] kirkland: that part is a pain, as I configured my vi for python [16:28] nijaba: sure [16:28] nijaba: i'm not going to touch your python indentation [16:28] nijaba: i'll follow your standard there [16:28] kirkland: for python, tab is to be banned [16:28] * kirkland grumbles about python [16:29] kirkland: I need to tweak my vimrc some more to change this based on the code type [16:29] nijaba: yeah, have it exclude shell for that [16:30] ubuntu-minimal wanted to isntall it and that made it crash. Now removed it and installed libc6-xen. Problems gone [16:30] found the culprit: libc6-i686 causes apt to crash when installing module-init-tools [16:30] when running in a xen virtual machine [16:43] nijaba: what do you think about F6 as a simple detach? [16:44] kirkland: good for me [16:50] I have removed a scsi device with scsiadd -r 3 0 0 0 [16:51] this was successful, however I can not get it back on line [16:51] scsiadd -a 3 0 0 0 is not working and rescan-scsi-bus.sh does not grab it [16:54] heath|work: These are actual scsi drives? [16:54] hey people. should I run my iptables script pre-up or post-up in network/inerfaces [16:55] genii, sata [16:58] I have 2 disks that contained 2 partitions each, both part of an mdadm raid0 array [16:59] the disks were orgiinally in a NAS, there was a bodged firmware upgrade and the software tool to repair attempted to format the disks fresh [16:59] it was stopped promptly before it got to the data partition (tool was still formatting system partitoin) [17:00] mdadm --assemble now doesn't work on the data partitions, complaining about missing superblocks [17:00] google suggests that mdadm's smart enough to recognise an underlying filesystem on a disk and reallow access to it [17:00] ie, mdadm --create wont trash all the data on the disks [17:06] heath|work: They are attached to a a SCSI host adapter by way of some adapter like so: http://www.addonics.com/products/io/adsalvd160.asp or just on their own sata controllers, etc? [17:06] Deeps, what does cat /proc/mdstat have? [17:07] genii, just sata drives connected to the MB [17:08] good news [17:08] tar upstream is seriously considering including proposed acl/selinux/xattrs patches [17:09] heath|work: Did you do: scsiadd -p after you added the drive back in? [17:10] genii, yes... even with scsitools rescan-scsi-bus.sh I get nothing [17:10] it's very weird [17:11] heath|work: Does the computer see the drive on ahardware level? (bios scan etc) [17:11] heath|work: was empty [17:12] genii, thanks for you help with this... I have not rebooted yet, Iwas actually trying to avoid a reboot [17:12] heath|work: looks like mdadm is clever though, i did the create, it detected the e2fs and warned me, i told it to carry on creating anyway, and now an e2fsck hasn't complained *yet* on the newly created md1 device [17:13] Deeps, after creates you should have something in /etc/mdstat [17:13] http://paste.ubuntu.com/101755/ [17:13] now yeah [17:14] it's a raid0 though, so it wont have much of interest [17:14] and infact, doesn't [17:14] heath|work: OK. Does tail dmesg have anything like: scsi 3:0:0:0: rejecting I/O to dead device or so after physical insertion of drive? [17:15] anyway now just to wait for e2fsck to finish it's thing. 1.5tb array, so it wont be speedy hehe [17:17] genii, http://paste.ubuntu.com/101757/ after doing scsiadd -a 3 0 0 0 [17:18] but nothing shows in fdisk -l [17:18] there's no disk [17:18] that's sata channel [17:19] ahhh... I wounder if the drive is not powering up then [17:25] where is the cache of uuid's to dev [17:25] if there is one [17:25] cache? [17:25] /dev/disks/by-uuid/ [17:25] there are links [17:25] hey people. Is it recommended to run apparmor on a LAMP server (running inside a xen virtual machine) running hardy? [17:26] SmokeyD: apparmor runs by default [17:26] ivoks: yeah I know [17:26] SmokeyD: there's only mysql profile [17:26] which is ok, if you don't do non-standard setup [17:27] ivoks: I am wondering more whether it is usefull to keep it running [17:27] i would [17:27] SmokeyD: it is usefull [17:27] ok, then I'll dig into the docs [17:27] SmokeyD: but it can also be problematic, depending on your knowledge :) [17:27] heath|work: "ata4" suggests it would be responding as scsi device 5 ... have you tried to scsiadd that designation in? [17:33] ivoks: I am reading http://developer.novell.com/wiki/index.php/Apparmor_FAQ now, but in the mean time, are problems caused by too restrictive apparmor settings easily recognized? [17:34] SmokeyD: since only mysql is apparmored, non working mysql would be only possible problem [17:34] ok [17:35] other programs are not apparmored? SO the use is still quite limited? If I wanted apache for instance apparmored, I need to do it myself? [17:36] There are others that are apparmored too. [17:36] IIRC there's a list on the Ubuntu wiki. [17:37] ok, I'll just read the docs and stop being lazy :) [17:37] thanks [17:41] ScottK: we are talking about LAMP :) [17:41] oh... he asked :D [17:42] sorry [17:42] SmokeyD: yes, armoring apache is up to you [17:42] ok [17:42] SmokeyD: since there are so many ways to use and set up apache, providing sane profile for it is quite hard [17:43] yeah I can imagine [17:43] virtual hosts, location of files, etc [18:15] I'm trying Wireshark on Ubuntu Server... and flailing :-) [18:32] ball: Wireshark is a GUI app. [18:32] No GUI on Ubuntu Server. [18:32] ball: Use tcpdump instead. [18:32] hello everyone [18:33] I am getting used to the tool that comes with ubuntu server for firewalling [18:33] ufw [18:33] ScottK: ah, I installed VNCserver [18:33] ...so I can run GUI tools if I have to [18:33] but how do I enable prerouting rules? [18:33] Icouldnt find any info on the docs [18:34] bmon seems to be showing me the traffic through an interface. [18:35] come to think of it, that'll work provided I run it on a box that all the traffic is going through [18:35] ...what I was really hoping for was something like xload, but showing me LAN traffic rather than cpu load [18:36] ball: nload? [18:38] nick58b: that's showing me the traffic on one interface. Presumably that won't show me traffic between two other nodes. [18:38] (general volume of traffic on the LAN) [18:39] oops [19:00] ball: oh, yeah, i have switches with snmp and use cacti for that [19:01] I have a hub :-) [19:01] ntop? [19:01] aurigus: I'll try that, thanks [19:03] hub, eek :) [19:03] :-) [19:04] aurigus: don't make me break out the coax ;-) [19:05] if you do that i'll have to break out my pigeons [19:06] * ball gets out the cw key [19:24] Can anyone help me add prerouting rules on ufw? [19:24] I am having a real hard time making a certain rule work [19:24] -A PREROUTING -p tcp -d 201.70.77.6 --dport 2222 -j DNAT --to-destination 192.168.10.8:22 [19:27] do I have to enable PREROUTING like I need to enable POSTROUTING? [19:36] if I set up and alias of eth0:0 192.168.2.31, can I route have traffic going to an ip of 192.168.2.200 look as if it is coming from the alias not the regular ip? [19:37] Why would you do that? [19:37] heath|work: Yes [19:38] ball, we have a cisco router that only listens to 1 ip [19:38] genii, what do I need to look into to do that [19:39] heath|work: odd. [19:40] heath|work: Just use a netmask for the alias which excludes the IP of the non-aliased eth0 (or other aliases). [19:40] ball, it's a security thing [19:41] heath|work: Then 192.168.2.31 for example becomes the ony route to 192.168.2.200 because it's only one on the segment for that adapter [19:41] heath|work: ip route add 192.168.2.200/32 src 192.168.2.31 [19:41] heath|work: might do what you need [19:42] heath|work: oh, and the use of eth0:0 and the like is deprecated, in favour of ip address add [19:43] hmmm... thanks... I will look into that! [19:56] Deeps, I cannot find any tutorials or info on how to setup aliases with the ip command. Would you happen to have any references? [20:01] sudo ip addr add eth0 192.168.1.10 [20:04] heath|work: ip addr add 192.168.2.31/32 dev eth0 [20:07] wow that's easy... thanks! [20:11] `ip` is the tool of the future! [20:12] does it save that on reboot? [20:13] nope [20:13] unfortunately, due to the nature of how the interfaces file works, there's no 'clean' way to add multiple IPs to an interface [20:14] outside of lots of up or post-up lines [20:16] but you can use loops in (post-)up lines [20:16] e.g. up for i in `seq 2 10`; do ip addr add x.x.x.$i/32 dev eth0; done [20:23] Hi [20:24] Does the server version of Intrepid still respect /etc/network/interfaces or does it also have some nm weirdness? [20:25] RainCT: by default there shouldn't be any nm weirdness... at least that I've noticed [20:25] "nm weirdness"? [20:25] OK, thanks [20:26] Deeps: network manager [20:26] did nm get installed as a dependency of something maybe? [20:26] oh [20:26] sommer: I have no problems (using Hardy here), just asking :) [20:48] Deeps, using your above command for adding an alias, I can now ping the alias, but I cannot view it with ifconfig or route. Is this normal? [20:49] indeed, you can see it in ip addr [20:49] ifconfig is also deprecatedin favour of ip, as is route [20:49] (ip addr, ip route) [20:50] so what you are telling me is to drop ifconfig [20:52] now that I have run ip addr add 192.168.2.37/32 dev eth0 how can I make a route through that ip to 192.168.2.1 [20:52] to that particular ip [20:52] or to use 192.168.2.37/32 as your src ip for all outbound connections? [20:54] I need to use 192.168.2.37 to 192.168.2.1 everything else stay the same [20:54] ip route add 192.168.2.1/32 src 192.168.2.37 [20:54] might do what you need [20:55] yeah... that's what I tried I got "No such device" [20:56] hey people. If I want my server to stay synchronized with the time, should I use ntp or ntpdate? [20:56] ip route add 192.168.2.1/32 dev eth0 src 192.168.2.37 [20:56] well... no error that time. Let me see if I can connect!! [20:58] Deeps, you are the man!! Thanks [20:58] source based routing [20:58] someithng you cant do using the old tools like 'route' [21:00] that explain all the problems I was running into. Damn google [21:08] SmokeyD: use ntpd [21:08] SmokeyD: ntpdate is a one off time sync. To run it the ntpd service needs to be stopped [21:08] SmokeyD: see /etc/ntp.conf [21:09] maw_: ok, thanks [21:09] I usually first run ntpdate against my time server, then I start the ntp service [21:14] I usually run ntpdate once a day using cron, for this I stop ntpd first, update time and then start ntpd again [21:15] if you need to do that you're doing it wrong [21:22] another question: would you for a server recommend the binary snort 2.7.0 from ubuntu hardy or a manually compiled snort 2.8.3 (latest stable)? [21:23] depends on whether you'll keep up to date with snort developments [21:24] incase of any security vulns and the like [21:24] and remember to keep your snort patched up [21:24] i'd never recommend using anything outside of the packaging system simply as it's more effort to maintain [21:25] Deeps: but the ubuntu binary version is quite a lot behind the stable snort [21:25] and that would post a security risk right? [21:25] indeed, because latest and greatest != stable [21:25] not necessarily [21:25] security fixes are often backported [21:25] new features, however, are not [21:25] look at debian stable, packages in there are even older! [21:26] Deeps: so it is possible that although snort in repos is version 2.7.0, the vulns which were solved in later versions were patched back in 2.7.0 by ubuntu maintainers? [21:26] indeed [21:26] you can confirm that by looking at the package's changelog [21:27] ok [21:27] cool [21:27] so you would only recommend hand compilation if you really need a new feature from a later version, but not for security issues? [21:28] snort is special : you need the oinkmaster to keep up with rules updates, and it may break an older version of snort from worling. It happened to me once. [21:29] so i'd recommend to compile the latest, which is easy to do. [21:30] i'd recommend either creating your own package from the source, or grabbing the latest package from the testing repos (in this case, jaunty) [21:30] I think I will go with the latest source (stable of course) and indeed create a deb package [21:32] nijaba: around? [21:32] nijaba: https://wiki.ubuntu.com/ScreenProfiles [21:32] kirkland: yep [21:32] looking [21:34] kirkland: that's a serious reverse spec :) [21:36] nijaba: ;-) [21:36] nijaba: well, this should be more information that others can use, if we get some more assistance [21:36] kirkland: missing a todo table? [21:37] * kirkland hands nijaba the todo to make a todo table :-) [21:37] nijaba: good idea ;-) [21:37] nijaba: well, actually, i think those would be better tracked in Launchpad, right? [21:37] nijaba: as wishlist bugs? [21:38] nijaba: maybe highlevel items in the wiki page [21:38] nijaba: specific implementation items as bugs [21:38] nijaba: and then people can bite off bits and pieces, bug by bug [21:38] kirkland: yes, high level one in the wiki page is better [21:38] s/one/ones [21:38] nijaba: cool [21:39] nijaba: okay, one more thing, then i need to shift my focus away from screen-profiles for a bit ... [21:39] nijaba: in that wiki page, i called your python utility "screen-configurator" [21:39] ok [21:39] nijaba: i think that's a more accurate term, than "welcome screen" [21:39] nijaba: what do you think of that overarching direction? [21:40] nijaba: your utility becomes the central place for configuring screen [21:40] kirkland: I had called it screen-profiles-helper, but that's fine too [21:40] nijaba: initially, that means choosing your profile [21:40] nijaba: eventually it means defining what windows to open at startup [21:40] nijaba: whether to run screen automatically on login (or not) [21:40] nijaba: perhaps eventually mangling the keybindings [21:40] yes, that sounds nice [21:40] nijaba: and configuring the applets in the status bar [21:41] nijaba: those last two are "hard", i think [21:41] nijaba: perhaps not a jaunty item, unless we got some help [21:41] nijaba: but good for the "to do" list, i suppose [21:41] applets sound much harder than keybindings [21:41] nijaba: yeah, probably so [21:45] nijaba: okay, i just released version 1.3 [21:45] nijaba: pushing to my ppa, and to universe [21:45] wooohooo!!! [21:46] Since Hardy (Ibex too) nautilus cannot access my password protected samba shared folders, hosted on a Dapper server. I'd like to figure out why. Gutsy and older had no problems to do that. Is it because samba/dapper is outdated? [21:49] non protected folders mount just fine, btw [21:52] the samba server version is 3.0.22-1ubuntu3.8 [22:02] hi all [22:02] please take a look at this issue [22:02] http://ubuntuforums.org/showthread.php?t=1033442 [22:02] need help [22:02] tearing my hair out here [22:04] disable apparmour, if the problems are resolved, you need to edit your configs to play with files in the correct places, or alter the apparmour profile for bind to allow read/write in your custom locations [22:05] thanks for that [22:05] Deeps: why do you recommend disabling? Putting in complain mode is MUCH better [22:05] or that [22:05] complain mode would be better, yes [22:05] https://help.ubuntu.com/community/AppArmor#Put%20all%20profiles%20into%20complain%20mode [22:05] ah thanks [22:06] nijaba: ignorance, rather than malice, i'm afraid [22:07] Deeps: np, it is just that it makes you progress a bit quicker in your troubleshooting and avoids leaving a security hole afterward... [22:07] yepp [22:07] completely agree [22:09] it made no difference [22:09] do you see anything obviously wrong with my config? [22:09] fduplessis: then put it back in enforce before looking somewher else ;) [22:10] yep it's running again [22:10] logging for bind9 works fine though [22:10] when i try to ping my domain i get: [22:11] client 127.0.0.1#57929: query: mydomain.com.mydomain.com IN A + [22:13] fduplessis: can you try with dig? [22:15] vndadmin@alpha:/etc/bind$ dig mydomain.com [22:15] ; <<>> DiG 9.5.0-P2 <<>> mydomain.com [22:15] ;; global options: printcmd [22:15] ;; Got answer: [22:15] ;; ->>HEADER<<- opcode: QUERY, status: SERVFAIL, id: 56607 [22:15] ;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0 [22:15] ;; QUESTION SECTION: [22:15] ;mydomain. IN A [22:15] ;; Query time: 3 msec [22:15] ;; SERVER: 127.0.0.1#53(127.0.0.1) [22:15] ;; WHEN: Wed Jan 7 16:14:47 2009 [22:15] ;; MSG SIZE rcvd: 30 [22:15] paste.ubuntu.com [22:16] what? [22:16] in future, please use a pastebin for anything more than 2-3 lines [22:16] fduplessis: Please use the pastebin next time for large texts. [22:16] ok [22:16] sorry [22:17] fduplessis: SERVFAIL looks bad. Any backup of your previous config to do a diff? [22:17] yeah i can go back. [22:17] brb [22:18] fduplessis: seems like the shortest path to find the issue [22:18] blog IN CNAME example.wordpress.com [22:18] and the line [22:18] like [22:18] are all missing full stops at the end of the domains [22:18] in example.com zone file [22:19] i tried with full stops, didn't work either. but thanks now at least i know what NOT to do... [22:19] FQDNs always need to be terminated with a period [22:19] blah.domain.tld. IN CNAME blah.domain2.tld. [22:20] blah IN CNAME blah.domain2.tld. [22:20] both are legal (i think) [22:20] ah thanks [22:20] i also dont think you can CNAME to a non-FQDN [22:20] that helps [22:20] so: fileserver IN CNAME vasvr01 isnt legal [22:21] but fileserver IN CNAME vasvr01.mydomain.com is? [22:21] fileserver IN CNAME vasvr01.mydomain.com. [22:21] correct [22:21] gotcha [22:21] rolling back now [22:22] when you start bind9, if you look in /var/log/syslog it'll announce any errors in loading zone files too [22:22] ok will lok, thanks [22:22] so you know straight away if you've got errors in your zone files [22:22] which would lead to SERVFAIL like you're seeing above [22:22] of course, in your forum post, you're using example.com and not mydomain.com [22:23] i'm assuming you're just hiding your real domain or whatever and keeping consistent at your end [22:23] yep, i am confusing to myself and other, i know [22:25] ok, rolled back, it's working 100% now [22:25] i must have madde some mistake when adding those records [22:27] i am gettin this tho: Jan 7 16:24:39 alpha console-kit-daemon[4156]: CRITICAL: cannot initialize libpolkit [22:32] Deeps, could you take another look at my forum post? [22:33] jdstrand: hi, you around? [22:33] link? [22:34] jdstrand: i published updated screen-profiles packages to https://edge.launchpad.net/~kirkland/+archive [22:34] jdstrand: i'm wondering if you might test, and give some more feedback [22:34] jdstrand: i juggled the keybindings a bit, to use some that work in both tty consoles and gnome-terminal [22:35] kirkland: you gave me the screen tweaking bug now [22:35] if there's anyone out there that can test in Konsole, i'd be appreciative ;-) [22:35] I've been tweaking my own custom thing now [22:35] greenfly: ? [22:35] * ScottK looks at vorian [22:35] ... for konsole testing. [22:35] * kirkland was looking at ScottK :-) [22:35] http://ubuntuforums.org/showthread.php?p=6513073#post6513073 [22:35] * ScottK needs to go make dinner. [22:36] greenfly: cool, let me know if you find any gems you'd like to share [22:36] kirkland: just I never had thought about it much, and now I can't stop [22:36] kirkland: vorian has expressed a desire to get more inolved with Server Team, so here's s chance .... [22:36] kirkland: I like the %l for system load [22:36] greenfly: k [22:36] ScottK: sounds good ;-) [22:36] * vorian waves [22:36] kirkland: and I wrote a little thing that parsed fetchmail -c output to show many how many unread messages I have [22:37] kirkland: I'd be happy to test it-- is it in your ppa? [22:37] jdstrand: yup [22:41] kirkland: is help supposed to be F1 still? [22:41] ScottK, kirkland, i'll be able to test it later this evening [22:41] jdstrand: no [22:41] jdstrand: F1 is help in gnome [22:41] vorian: cool, thanks. [22:42] kirkland: 'morning Dustin :) [22:42] jdstrand: f10 is the toolbar in gnome, and f11 is fullscreen ... i avoided all of those [22:42] owh: howdy [22:42] no problemo [22:42] kirkland: help.txt isn't shown to me on startup so I have no idea what is what ;) [22:42] jdstrand: yeah ... that's commented out at the moment [22:42] kirkland: where is help? [22:42] jdstrand: we're going to add it back in, as soon as we have a way to disable it, permanently, by choice [22:43] F9 [22:43] kirkland: Just out of curiosity, are my emails getting to the server list? I've asked some questions over the past few months with zero replies and I was wondering if I was unloved, or just not getting to the list :) [22:43] kirkland: TBH, I like it-- it is not unlike what you would get in normal screen-- though I agree having a way to disable would be nice [22:43] owh: i see a question from you at 1:48am this morning, regarding vmbuilder under intrepid [22:44] * owh notes that gmail's mail servers don't send copies of list postings back - RFC's, must be something new:) [22:44] owh: it has been a slow few weeks, xmas and all [22:44] kirkland: So, just unloved then :) [22:44] jdstrand: yeah, the intention is to put it back in, as soon as we get a couple of issues solved [22:45] newt is pretty [22:45] jdstrand: nijaba did a great job with that [22:45] props to nijaba ! [22:45] \o/ [22:45] * nijaba blushes [22:45] kirkland: BTW, did you see the patch submitted for bug #238444 [22:45] Launchpad bug 238444 in kdebase-runtime "Index building for initial launch is overly complex" [Undecided,Invalid] https://launchpad.net/bugs/238444 [22:45] 5 [22:45] o/ [22:45] jdstrand: that's going to become our all-around "screen-configurator" [22:45] kirkland: Whoops wrong bug. [22:46] kirkland: Bug #203169 [22:46] Launchpad bug 203169 in samba ""status" function for init scripts" [Wishlist,In progress] https://launchpad.net/bugs/203169 [22:46] Crap [22:47] Not that one either, let me hunt :( [22:49] kirkland: Uh, it was the last one: https://bugs.launchpad.net/ubuntu/+source/lsb/+bug/203169/comments/46 - <- nis patch [22:49] Launchpad bug 203169 in samba ""status" function for init scripts" [Wishlist,In progress] [22:50] owh: hmm, someone marked the nis task invalid ... [22:51] kirkland: The person submitting the patch did that. [22:51] oh, i see [22:51] I'm not sure what he was doing. [22:51] The patch itself is mostly right, but I'm not sure that it's all correct. [22:51] owh: thanks, i'll chase it down [22:52] owh: and test it [22:52] gosh, i haven't run nis in *years* [22:52] kirkland: I think the -p in the first pid of proc is borked [22:52] s/proc/proc-call/ [22:53] Deeps? [22:53] kirkland: Uh status_of_proc, rather :) [22:54] kirkland: Doesn't supplying the pid defeat the purpose? [22:55] owh: no, it's a good thing ... it ensures that we're checking the status of the daemon started by the init process [22:55] owh: and not just some other one that the admin might have started [22:56] kirkland: Uhm, not really. The file is one stored in the /var/run tree, that could have been put there by anything. [22:57] owh: well, you'd have to look at the rest of the init script [22:57] owh: specifically, the start action [22:58] kirkland: I understand how it stores the file, but if the admin for some reason starts it twice with the init script, then the file will be overwritten. So we might be looking for a pid that is no longer there if the stuff gets out of sync. Wasn't that the whole point of not using the pid in the first place/ [23:00] kirkland: And is the definition of nis running not that a process is running, not if the process with the pid that was stored is running? For example, perhaps the nis process fails, detects itself failing and starts a new instance of itself. [23:00] owh: we were specifically asked to add pid support to status_of_proc() [23:00] kirkland: I don't recall, but I trust your statement. [23:01] kirkland: In that case, looks good to me :) [23:01] owh: if i were less busy, i'd dig out the mail archives [23:01] mathiaz: thanks for the rabbitmq revu [23:01] owh: it's in the debian bug tracker though [23:01] kirkland: I'll have a squiz. I don't doubt your statement though. [23:04] kirkland: There was a debate about the pid stuff using kill -0 which would only work for the process owner and root. I did not see anything else. [23:07] kirkland: Don't spend any time on this, I was just drawing your attention to the patch. The debate can last for another time. [23:07] owh: thanks for that, i'll chase down the patch for commit [23:07] Cool [23:17] jcastro: np - I've made some comments [23:17] jcastro: once they've been addressed I'll take another look at it [23:17] nod [23:21] jdstrand: there? [23:23] jdstrand: btw, I will have some time tomorrow evening to take some tests against ufw trunk. Do you want me to try to run it against jaunty or hardy? [23:24] I think I asked yesterday, but I didn't bookmark it: Can you add disks to a RAID? Like adding 2 disks to an existing RAID5 or 6 [23:30] hello. can anyone tell me how to make automatic file backups to an external usb hdd on my ubuntu-fileserver? [23:31] Koprolyt: Use any command-line backup tool of your choice and schedule the backups using cron. [23:31] can u please give me a name of any cmdline backup-tool pls [23:32] nijaba: tab mangling fixed! [23:33] rsync, rdiff-backup ... use Google for this kind of question, it is more comprehensive! [23:33] Koprolyt: you can use tar+cron [23:33] rdiff-backup, rsync, or something more advanced as bacula [23:33] unison ... there are so many... [23:35] thx! what is the most popular? so i can google help. my server is running now for 1 day. it was my first contact with linux. [23:36] New bug: #293233 in smbnetfs (universe) "crashing with SIGABRT while scanning network or reading some files" [Undecided,New] https://launchpad.net/bugs/293233 [23:36] New bug: #314899 in openssh (main) "User is prompted for password for irrelevant public key" [Undecided,New] https://launchpad.net/bugs/314899 [23:43] rsnapshots neat. [23:44] aiui, similar to rdiff-backup [23:45] New bug: #314776 in bind9 (main) "OpenSSL signature verification API misuses" [High,In progress] https://launchpad.net/bugs/314776 [23:54] to install LAMP which command do i run? sudo apt-get install lamp ? === liberfiasco is now known as libervisco [23:56] i think theres a tasksel task for lamp under ubuntu [23:56] otherwise, install the components manually [23:58] Kamping_Kaiser: Yes, it's a tasksel [23:58] techsupport: sudo apt-get update && sudo apt-get install tasksel && sudo tasksel ...then choose LAMP install