[00:01] hggdh, A new euca is landing shortly... upgrade issues fixed [00:02] Daviey: awsome! And... I guess there I go back to Lucid ;-) [00:02] efffin finally [00:02] hggdh, I've been doing it all day - so i do sympathise :) [00:03] so the kernel they tried pushing out today was bad ? [00:03] dominicdinada, huh? [00:04] i was wondering if the update today was a bad kernel [00:05] I know Ubuntu's ruby package is messed, but I can't remember the work around for this error. anyone? http://pastebin.ca/1949976 [00:26] New bug: #649423 in vsftpd (main) "package vsftpd 2.2.2-3ubuntu6 failed to install/upgrade: le sous-processus script post-installation installé a retourné une erreur de sortie d'état 1" [Undecided,New] https://launchpad.net/bugs/649423 === billybigrigger_ is now known as billybigrigger === freeflyi1g is now known as freeflying [00:58] I used to use ubuntu kvm virtualization on hardy, and ran win2k3 instances for a long time with no issues. I have tried everything, and am unable to create a stable win2k3 instance on ubuntu lucid. nothing else i can thing of to try, and i found no specific cause, it just doesnt work. runs unusably slow and throws errors. === ivoks is now known as ivoks-afk [01:05] kirkland, if you have any ideas for how I could troubleshoot win2k3 kvm failures on lucid that would be great and maybe I could make a bug. Im out of ideas, and dont know the cause. [02:11] hi [02:11] im not a troll [02:13] ugh [02:13] anyone know how to do subdomains in Apache? [02:14] I've set my A record for the subdomain with my domain name supplier, but Apache isn't co-operating with making the subdomain >_> [02:15] i want to setup my server for httpd file and database [02:15] what do i choose [02:17] Aliv3: define "httpd file and database" [02:18] web server [02:18] file server [02:18] mysql [02:18] but i dont know the aliases, it has funny names for them on ubuntu server [02:20] Have you read the Ubuntu Server Guide? [02:21] those are the names of the actual software packages that provide those services [02:21] "web server" is a type of service, "apache" is a particular web server application [02:21] as for file server, do you mean, like, ftp? [02:22] most people use apache2 and proftpd for web and ftp respectively. i would second that you should take a look at the server guide though [02:22] vsftpd FTW [02:22] ksx4system: :D yeah, that's what I use [02:23] proftpd seems to be the most popular choice though, probably b/c you can get it to do more stupid tricks out of the box [02:23] it might be better for like a complex shared web hosting server environment [02:23] it certainly isn't bad [02:24] and whats a DNS server do [02:24] can i like make up a tld and let ppl register or something [02:25] imvho BIND9 sucks balls, MaraDNS FTW (easier to configure, less resource hungry) === twb` is now known as twb [02:26] like it gives me these options: [02:26] I use dnsmasq, FWIW. [02:26] Er, for a recursive resolver -- not for hosting my own records. [02:27] [ ] DNS server [02:27] [ ] LAMP server [02:27] [ ] Mail server [02:27] [ ] OpenSSH server [02:27] [ ] PostgreSQL database [02:27] [ ] Print server [02:27] [ ] Samba file server [02:27] [ ] Tomcat Java server [02:27] [ ] Virtual Machine host [02:27] [ ] Manual package selection [02:27] !paste [02:27] For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://tinyurl.com/imagebin | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [03:21] New bug: #649495 in bind9 (main) "package bind9 1:9.7.0.dfsg.P1-1 [modified: usr/share/bind9/bind9-default.md5sum] failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/649495 [04:05] anybody have experience with ps3mediaserver or transcoding? === bastidrazor is now known as bastid_raZor [05:18] a good ubuntu based rescue disk, whats your recommendation? [05:20] All ubuntu disks are rescue disks, what are you rescuing ? [05:21] i need some more out of box, i don't want to install anything :) [05:22] do you have a network or an external drive to put your "more" into ? [05:23] Ubuntu live disks have both install and live modes. Just use the live mode. [05:32] IMO ubuntu desktop CDs make shitty rescue discs, because they don't have stuff like lvm and mdadm pre-installed === knolls_ is now known as knolls [05:58] Hey, I'm wanting to host some small webspace from my house, half to be useful, half to just learn about setting up a server, but first, how to I tell if my ISP has blocked port 80?? [05:59] lincoln: netcat listen to port 80 from inside, and make a request to 80 from outside [06:00] how do I get "outside" though? that's my question. Can I simply do it from another computer inside my network? [06:01] lincoln: ssh into ubuntu.com or so [06:01] I'm assuming that as a sysadmin you already have an account SOMEWHERE [06:02] I'm no sysadmin, just a dude [06:02] haha [06:04] try http://en.wikipedia.org/wiki/SDF_Public_Access_Unix_System [06:07] http://en.wikipedia.org/wiki/Shell_provider contains lists of other providers in the "external links" section. [06:09] I tend to use alioth.debian.org or code.haskell.org or one of my customers' hosts. === FireCrotch_ is now known as FireCrotch [06:12] twb: okay awesome, thanks [06:42] twb, agreed about live CDs being shitty rescue CDs point, thats exactly what i am looking for. [06:42] something out of box, ubuntu based, have basic data forensics, network forensics tools installed [06:42] Historically I use either Knoppix or CentOS live CDs. [06:42] support for more fs out of box like xfs [06:42] Lately I have been using hand-rolled or pre-rolled Debian Live rescue images. [06:43] forensics is a different rôle to rescue [06:43] i have been tempted for http://ubuntu-rescue-remix.org/ and http://grml.org [06:45] somehow i am biased towards grml here :) [06:46] New bug: #649559 in postfix (main) "package postfix 2.7.0-1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 75" [Undecided,New] https://launchpad.net/bugs/649559 [07:18] twb, Error404NotFound, that will teach me for responding before checking which channel I'm on. I use a boot usb with custom stuff installed for server work. duh. [07:20] fluvvell: no worries. [07:31] New bug: #649574 in dovecot (main) "dovecot-postfix incomplete install" [Undecided,New] https://launchpad.net/bugs/649574 [08:03] ttx, ping [08:03] https://bugs.launchpad.net/ubuntu/+source/mountall/+bug/649591 [08:03] Launchpad bug 649591 in mountall "mountall spins eating cpu when 'nobootwait' option exists in fstab" [Critical,Confirmed] [08:07] ttx, I have to get to bed. but i'd really appreciate a ping to Keybuk or someone who could look at that bug. === jjohansen is now known as jj-afk [08:14] ttx, you shuld have access to ubuntu@ec2-184-72-166-196.compute-1.amazonaws.com, but the bug reproduces on reboot of instance of type m1.small of recent maverick. [09:05] Anyone know about the marvin backported kernel for 10.04? [09:06] http://pastebin.com/e8RRyxUW <-- When I try to add it, I still can't find the package, getting # [09:06] W: Failed to fetch http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu/dists/karmic/main/binary-amd64/Packages.gz 403 Forbidden [09:07] mountall is totally broken [09:07] I hope the whole thing is gone by 12.04 [09:11] Why does apt-get update give me that error message, but not aptitute update ? [09:11] aptitude [09:16] sander^work: because in the mean time, someone deleted the file from the server [09:17] twb, who to contact to get it fixed? [09:18] Whoever the PPA belongs to, I suppose [09:18] Try #launchpad or #ubuntu-kernel? [09:18] Ok.. thanks. [09:41] Any idea how to mount virtual machine disk (it is logical volume, not image file) in host system? [09:42] overlord_tm: you mean that it's an LVM LV? [09:42] yes [09:42] mount /dev/mapper/foo-bar /mnt [09:42] If your virtual disk is partitioned, you'll need an -ooffset=N, where N can be obtained by running fdisk -l (or so) on the device [09:43] Typically something like 63 × the block offset. [09:44] twb, it says that i have to pass filesystem :/ it is ntfs drive of win 2k3 machine [09:44] So you probably have a partition table on the virtual disk [09:45] fdisk says there is no valid partition table [09:46] What does file -s say? [09:46] root@mru-server:/home/mru# file -s /dev/mapper/vg_mru-sql--server [09:46] /dev/mapper/vg_mru-sql--server: Qemu Image, Format: Qcow , Version: 2 [09:47] Ah, OK. It's not a raw disk, it's a qcow2 image. [09:47] You can't mount it unless you convert it into a raw disk using qemu-img(1). Note that this will consume 100% of the virtual disk's size on your host system, in addition to whatever the LVM LV already consumes. [09:48] ok, what parameters do i have to pass to virt-install to use raw image? [09:49] I don't know. I don't use silly wrappers like libvirt. [09:50] that are broken in lucid :) [10:42] good morning all...i'm trying to upgrade ubuntu server 9.04 to 10.04LTS. i first upgraded to 9.10, then attempted to go to 10.04LTS and hit an error that aborted the upgrade. hunting round the forums, it appears "the problem is caused by a feature in apt that tries to ensure that all important packages get configured immediately to minimise risks if something goes wrong".... [10:43] so, the workaround i'm trying to use is to manually change to lucid, get the 64bit version of the latest lucid mountall, then force mountall to install without the dependencies... [10:44] my problem is, i don't know which is the latest lucid mountall version...can anyone help? is it 2.15.2? [10:44] kinygos: ask rmadison(1) [10:45] $ rmadison -uubuntu -slucid mountall ==> mountall | 2.14 | lucid | source, amd64, i386 [10:45] twb: thank you very much for the quick reply :) [10:57] Anyone got any experience with Ruby hosting? specifically with RVM and Passenger? [10:57] Not me. #ruby or #ruby- might be a better place to ask about that [10:58] ok, fair enough [10:58] By all means wait around here for an answer, too [10:59] One other question. If I'd like to add an openlikewise user to a group, do I need to include a double \ instead of a single \ because of the domain? [10:59] ...if that makes sense [11:00] Never heard of "openlikewise" [11:00] ok. It's for authentication against active directory [11:01] It looks like a proprietary LDAP client [11:01] #ubuntu-server will probably only provide support for the PADL LDAP client that ships with Ubuntu. [11:02] (AD is basically Kerberized LDAP.) [11:04] ok cool. Out of interest, would PADL do a similar thing? [11:04] AFAIK PADL can talk to AD, though I've never tried it myself. [11:05] sound like it'd probably be a headache making the switch anyway [11:07] Shrug [11:07] IME LDAP and krb are *always* a headache [11:08] lol, true [11:11] Unfortunately nothing else (except NIS) is available, and NIS is a non-starter for networks that have users on them [11:12] And now for IPv6 it looks like I need to learn ipsec, too, which I've been avoiding... :-( [11:13] I'll be honest with you. Most of that went over my head, other than a few buzzwords - mainly IPv6 and ipsec [11:13] I'm a dev first and a lousy admin second [11:14] <_ruben> ipsec isnt all that bad once you get your head around it ;) [11:18] i've installed these packages: apache2 php5-mysql libapache2-mod-php5 mysql-server. how do i begin developing a web site. yes i know php and mysql but have only developed on a windows system. [11:20] rebirth: AFAIK PHP and SQL aren't any different on Unix. [11:21] <_Techie_> theyre exactly the same [11:21] <_Techie_> the only thing that differs is your editor [11:21] _Techie_: emacs works on Windows :P [11:22] <_Techie_> twb, dont get smart with me [11:22] ok, then what is a good editor to use on ubuntu? i used PSPad in windows. [11:22] rebirth: if PSPad includes transparent SFTP support, you could continue using it [11:23] <_Techie_> or you could get really tricky and actualy make changes then upload the file [11:23] rebirth: "programmers" tend to use vim or emacs, with comparatively negligible contingents using other random stuff like kedit and joe. [11:24] i don't see it in the software ceneter, also, i'm not attached to it and open to suggestions for the best software to use in ubuntu [11:24] I guess these days there's a non-negligible group of people who have machines powerful enough to run Eclipse or jEdit without OOMing [11:24] rebirth: you shouldn't be running a GUI on your server [11:24] Eclipse hmm [11:25] <_Techie_> i suggest nano, but thats mainly because i like being a complete asshat [11:25] nano is certainly a good choice of *first* editor. [11:25] <_Techie_> twb, i still use it for working on servers [11:25] does it do syntax highlighting? [11:26] <_Techie_> nope [11:26] i want something with syntax highlighting [11:26] nano has syntax highlighting [11:27] It's pretty shitty, though, even compared to the weak-ass lexical highlighting of vim or emacs [11:27] is vim console based? [11:27] (REAL editors parse code properly to highlight it, raargh!) [11:28] Both vim and Emacs have tty and GUI versions. [11:28] The GUI versions are basically some menus and toolbars around the tty versions [11:28] <_Techie_> twb, http://xkcd.com/378/ [11:28] my system is not going to be a server btw, i'm just developing on it [11:29] _Techie_: unfortunately Emacs' BUTTERFLY function is currently buggy [11:29] <_Techie_> twb, good thing i use nano then [11:29] _Techie_: http://code.haskell.org/~twb/Preferences/.nanorc (re syntax highlighting) [11:30] i'm installing emacs [11:30] http://emacsbugs.donarmstrong.com/4644 (re butterfly) [11:31] what is butterfly? [11:31] <_Techie_> twb, thankyou, you just made my day [11:31] rebirth: an alleged joke [11:32] Since cho seems to be AWOL again, try alioth: http://alioth.debian.org/~twb-guest/Preferences/.nanorc [11:39] <_ruben> speaking of awol ;) [11:50] i have something of a noob question....i used easy_install to install django, but i've heard from many sources that easy_install is baaaaad....easy_install put django into python2.6/dist-packages....which seems to be a ubuntu/debian thing cos normally packages go in site-packages.... === _Techie_ is now known as _TechAway_ [11:51] i will be manually installing this time, so my question is...should i install all my packages in dist-packages or site-packages? === _TechAway_ is now known as _Techie_ [12:03] smoser: i'm here now === _Techie_ is now known as _TechAway_ === _TechAway_ is now known as _Techie_ === _Techie_ is now known as _TechAway_ [12:22] How can I tell grub2 that I want it installed in /dev/sda2 permanently? (So, that it will go there every time there is a kernel update instead of overwriting my MBR) [12:32] ok, found it. It can simply be done with dpkg-reconfigure [12:37] hello [12:37] I'm trying to configure bind and still got this error: [12:38] rndc: connect failed: 127.0.0.1#953: connection refused [12:38] in /var/log/daemon.log I don't see any error ... [12:39] what can be wrong? [12:59] hello to all. i have just installed on my ubuntu server "php5-geoip" but for some reason i still get the erorr message in my script "Fatal error: Call to undefined function geoip_country_code_by_name". What i need to do additional to get GeoIP in php to work on a ubuntu server ? [13:00] apologies for the noob question, but i cannot find an answer that doesn't baffle me online....i have just upgraded a ubuntu 9.04 server to 10.04...the server has 2 physical hard drives so i'd like to install RAID1...is this possible, or do i need to rebuild the server? [13:07] wanted tell only that i solved the problem [13:08] kinygos: upgraded from 9.04 to 10.04? [13:18] pmatulis: i went from 9.04 to 9.10 then to 10.04 === rgreening_ is now known as ghost [13:19] it didn't have RAID configured..i don't think the second drive is even mounted, but i can see it with fdisk -l === ghost is now known as rgreening [13:30] is it possible to create disk partitions once the operating system has been installed? [13:31] kinygos: no, creating a RAID array is a destructive process [13:32] kinygos: the array is like a device in itself (ex: /dev/md0 for software RAID) [13:32] <_ruben> it is possible however to convert an installed system from non-raid to raid, its not quite trivial though ;) [13:33] _ruben: i would like to hear about that [13:34] <_ruben> pmatulis: long story short: use the 2nd drive to build a degraded raid array with the partition of your likings, copy data from 1st disk to raid, fix bootloader to boot from new disk, if everything works, repartition the 1st disk and add it the array for rebuilding [13:35] _ruben: that makes sense...if i have no data to salvage (this is a brand new server)...is there a simpler way? [13:35] <_ruben> kinygos: no data to salvage on either disk? [13:35] i don't have any access to external media either (such as cd-rom or network drive) [13:36] _ruben: no data at all [13:36] dnsmasq is mysteriously failing for me. [13:36] <_ruben> then why did you bother to upgrade the box in the first place, could've just reinstalled it [13:36] I've instructed it to serve DHCP on several interfaces, but it's not reporting any DHCP requests in syslog [13:36] Further, ss (netstat) reports that it's only bound to *:domain, and NOT to *:bootps [13:36] i have no external media...nowhere to store the .iso file [13:37] <_ruben> then again, without external media, that can be troubleshome [13:37] <_ruben> -h [13:38] <_ruben> also, the "story" i outlined above, is kinda tricky without exernal media (no option to boot a recovery cd when you bust your bootloader) [13:39] you've got me thinking about whether i could install 10.04 on the second drive and make that bootable....and if so, whether i can prepare the partitions on that disk for configuring RAID across to the initial drive on reboot [13:41] the sales guy from the hosting company said it was easy to set up RAID on 9.04...how foolish do i feel for accepting what a salesman has said??? [13:43] kinygos: md, fake or hardware RAID? [13:43] i was hoping for md [13:44] md RAID is easy to set up in d-i in 8.04, at least. [13:44] d-i? [13:44] It'd probably be substantially harder if you're working from an already-installed, non-raided disk [13:44] kinygos: d-i is the installer on the Ubuntu Server and Alternate CDs [13:46] twb: ah ok...i'm currently investigating the feasibility of a fresh install of 10.04 on the second physical disk.... [13:46] twb: unfortuntaley the server comes pre-installed with 9.04 :( [13:46] (serious typing failure there...sorry) [13:48] i've come to my senses...i don't know if the second drive is bootable according the server bios :( [13:48] <_ruben> buying a server without removable media and not having external removable media around is kinda uhm fail in my book [13:48] it's not that...i can get access to a network drive, but it will cost money that my boss doesn't want to pay :( [13:49] <_ruben> usb external cdrom drives cost nearly nothing, and thumb drives even less [13:50] it's flat-fees from the techs at the hosting company...they want to charge me just to stick a usb stick in the server :( [13:51] <_ruben> ah, so the server's remote as well? hope you got an out-of-band management board in it (ilo/drac/ipmi/etc), then again, considering the rest: i doubt it ;) [13:51] kinygos: if you are confident you can get the install correct FIRST TIME, there's a way to blow away (i.e. install to) both disks with md RAID by booting the installer *from the first disk itself*. [13:52] Er, assuming you have access serial or KVM-over-IP or some other management like _ruben mentions. [13:52] (Or you're *really* confident about doing the network part completely blind and one-shot.) [13:53] <_ruben> then again, lots of such solutions also offer virtual media allowing you to do whatever you want ;) [13:53] _ruben: I wouldn't know -- I only get the really horrible setups :-P [13:53] smoser: looks like the incredible cjwatson fixed it [13:54] i have KVM-over-IP....i'm on the phone to the ISP asking if the second drive is bootable...he still wants money just to give me access to some external disk space [13:54] cjwatson is just about the best feature of Ubuntu [13:54] <_ruben> twb: kinda the samehere , but idrac's are getting ordered more or less by default now ;) [13:55] kinygos: OK, so what you do is download the d-i netinst kernel and ramdisk and put them in /boot, and tell grub about them [13:55] kinygos: then, reboot, and in your KVM-over-IP UI, pick d-i, and then you are into a normal netinst and neither disk is mounted, so you can install to both. [13:56] yes, super watson to the rescue [13:56] twb: awesome, i will give that a shot...thank you and _ruben very much for your time...really appreciate it [13:58] kinygos: one moment and I'll give you the URLs, since you'll probably pick the wrong ones otherwise [13:58] :) i was just trying to find them [13:59] http://archive.ubuntu.com/ubuntu/dists/lucid/main/installer-amd64/current/images/netboot/ubuntu-installer/amd64/linux and .../initrd.img [13:59] Er, and .../initrd.gz [13:59] Swap "lucid" and "amd64" to taste. [14:00] lucid and amd64 are perfect....again, thank you very much :) [14:03] ttx: lemme check to see if im not crazy [14:03] zul: you still want to have a shot at samba maverick bugs, post-RC ? [14:04] ttx: i think so...if they are sane [14:04] no mention of the signon assistant fix in changelog [14:04] ttx: yeah i think i might have gone a bit crazy === sailerboy is now known as zz_sailerboy [14:05] Vaguely apropos of samba --- can one talk SMB2 (instead of CIFS) between samba server and samba guest in a homogeneous lucid network? [14:05] ttx: i cant reproduce the samba upstart nmbd bug though and i would like to get the sign-on bug fixed for maverick [14:05] twb: you need a newer version of samba like samba 4 or 3.5 [14:05] zul: ok, please prepare select fix, upload after RC [14:06] ttx: ack [14:10] zul: so I guess you can, but it isn't production-ready? [14:10] twb: samba 4 isnt [14:13] heh, thanks for the compliments :) [14:13] currently trying to fix bug 569900. I have the exact same size disk in my image but can't seem to quite reproduce the installer setting the wrong size [14:13] Launchpad bug 569900 in mdadm "mount: mounting /dev/md0 on /root/ failed: Invalid argument" [High,Confirmed] https://launchpad.net/bugs/569900 [14:14] * cjwatson tries lucid just in case [14:16] is it safe to just edit /boot/grub/menu.lst (i have grub version 0.98 as i upgraded from 9.04) === jorge_ is now known as jcastro_ [14:19] kinygos: yes, since you're about to destroy all of /boot === jcastro_ is now known as jcastro [14:25] ttx: i did the apache sru for the a2enmod modreqtimeout || true fyi [14:26] zul: cool [14:26] <_ruben> joy .. my local repository mirror + personal repo server got its disk (vmware esxi .vmdk) suffer some corruption .. time to figure out how i'll be able to salvage as much data as possible [14:40] AIDE question: if you do aide.wrapper -i, shoudln't it create a new aide.db? [14:46] or copy aide.db.new to aide.db? [15:02] how do i login in GUI mode in Ubuntu server [15:02] awanti_: usually you don't [15:02] awanti_: the server does not include a GUI by default, and we recommend you do not install one. [15:02] awanti_: did you add the gui stuff? [15:02] awanti_: the default package selection doesn't include an x-server [15:04] may i ask you why we don't install gui? [15:05] awanti_: because it's usually not needed on a server system [15:05] awanti_: why do you want a gui on a server? [15:05] awanti_: it's a waste of resources, no? [15:06] awanti_: if you would like a gui, use Ubuntu Desktop instead [15:07] yes your right. But i am comparing to RHEL. [15:07] zoopster: because more code = more potential vulnerabilities [15:07] RHEL only does it because they're marketing to Windows weenies, who can't comprehend a system without a GUI [15:08] (Sorry, bad completion there.) [15:09] why would my virtual network interface vanished after apt-get upgrade just now, how can I repair this? [15:09] ok i have one more question... I want install the packages offline. instead of downloading can i use download Server cd image [15:09] jmitchell: what provided the virtual network interface? [15:10] awanti_: all my servers, be them AIX or Solaris or freebsd or ubuntu or gentoo, are in a server room. They might be machines on their own but could also be vms. When I work on them I simply ssh to them and do my thing. Adding X to that just means more bandwidth I could be using for something else. [15:10] awanti_: just aptitude install --download-only while online, then rerun it without that option later [15:10] a physical interface [15:10] twb, physical interface was bridged to the virtual one [15:10] jmitchell: so it's defined in /etc/network/interfaces? [15:11] twb, i'm running hyper-v as the hyperviser [15:11] twn seth0 was and still is defined yes [15:11] however the if does not come up [15:11] odd [15:11] Isn't hyper-v a Windows thing? [15:11] correct [15:12] That part's out-of-scope for me, then [15:12] But pastebin your interfaces(5) file [15:12] ok, 1sec [15:13] ok can i use that OS Ubuntu server image for installing other application which i missed during installation time [15:14] twb, fek... cant copy and paste [15:14] it's just auto seth0 [15:15] <_ruben> awanti_: maybe, the online repositories offer waaay more packages than the cd has to offer [15:15] There should also be an "iface seth0 inet dhcp" or so [15:15] and iface seth0 inet dhcp [15:15] OK, then I have no idea what went wrong [15:15] <_ruben> awanti_: in /etc/apt/sources.list there's probably a commented line referring to your cd [15:15] soz, was typing still [15:15] twb, a fresh install and it works [15:15] twb, do the upgrade and bleh [15:16] and without the upgrade the vm is unstable [15:16] wtf, vmware here i come [15:16] screw this [15:16] twb, thanks for your time anyways ;) [15:16] awanti_, if your installation was complete enough to allow the server to go online, just add the other packages using aptitude or apt-get. [15:16] no worries [15:16] I usually do a basic server install + openssh. [15:17] twb, no clone option for vm's in hyper-v... it's just insane... :( [15:17] You'd have to go out of your way to get a system that can't get a netboot post-install. [15:18] well, i was thinking netboot [15:18] jmitchell: personally I'd prefer a jail to a VM [15:18] but why... seems like reinventing the wheel for something that should be easy [15:18] Oops, s/netboot/network/ [15:18] chrooted jail? [15:19] chroots have virtual filesystems. jails also have virtual network stack and process trees and other cgroupy things. [15:19] But they share a kernel with the host OS, so syscalls are still really fast [15:19] i feel like a noob twb, never heard of jails [15:19] Well, FreeBSD calls them jails [15:19] Solaris calls them zones [15:20] twb: I would think a zone is more than a jail [15:20] will do some reading maybe i can find a better way to do what i need [15:20] In linux they're vserver or openvz/virtuozzo or lxc [15:20] raubvogel: I'm hand-waving the details [15:20] mucho thanko ;) [15:20] twb: gotcha :D [15:21] raubvogel: suffice to say they're trivially circumventable like a chroot, but they're an order of magnitude more efficient than hardware vm [15:21] That is, the class including jails and zones and what have you [15:22] twb, how would you circumvent a jail? I always thought the entire idea for it was to make it harder to do that. [15:23] Sorry, I meant "they're NOT trivially ..." [15:23] You scared me there ;) [15:23] It's 0023 hours here [15:24] twb, sounds about right [15:24] and if u'r work isnt done by 2mrw U"R FIRED [15:24] Over here where lovebugs are our state bird, it is 10:25 [15:24] i'm in the crime center of the world, south africa [15:25] 4:24pm here [15:25] it's nice and sunny here though, gonna go mountain biking in 34min and counting :) [15:25] jmitchell, I crashed last night at 2am trying to do some work. Just to get up at 6 again. [15:25] ttx, what do you think of https://bugs.launchpad.net/bugs/649833 [15:25] Very nice! [15:25] Launchpad bug 649833 in cloud-init "uec images motd suggests tasksel, but tasksel not installed" [Undecided,New] [15:26] Anyone uses AIDE? [15:26] smoser: looking [15:26] raubvogel, we have just finnished a huge project and every1 is in chilled mode :D [15:26] raubvogel, AIDE? [15:27] smoser: since when do you think we don't have tasksel ? [15:27] jmitchell, http://sourceforge.net/projects/aide/ [15:27] * jmitchell sniffs at the url [15:27] smoser: in other words, does this affect lucid as well ? [15:27] $ sudo tasksel --section server [15:27] sudo: tasksel: command not found [15:28] i dont know [15:28] Intrusion detection [15:28] checking [15:28] no need for that anymore, i just use mikrotik and it's handy torch feature, scripting and firewalls take care of the rest [15:28] IMHO, kinda useless because it only checks for changed stuff once a day. OSSEC seems to be more active [15:28] ttx, lucid does have it: http://uec-images.ubuntu.com/releases/lucid/release-20100923/ubuntu-10.04-server-uec-amd64.manifest [15:29] I am also using fail2ban for the firewall stuff [15:29] i like that u can teach fail2ban [15:30] smoser: so I think we should fix the seed [15:30] it's looking after my bind9 too [15:30] Agreed [15:30] smoser: I'll comment to that effect [15:30] stopping twits from doing spoofed . lookups [15:30] smoser: ideally, before we generate the RC cloud images [15:30] Agreed. [15:30] Meeting. BBL [15:31] ciao, nice chatting [15:31] ttx, so which of these two bugs do you think i was alluding to yesterday (when i said i knew of some, but wasn't telling) [15:31] :) [15:31] I... need... lots... of... coffee [15:31] i have a biiiig mug in front of me now too :) [15:32] smoser: we have a specific seed for cloud images ? [15:32] yes. [15:32] http://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu.maverick/annotate/head%3A/uec [15:33] * ttx wonders when that was booted off [15:34] some other dependency probably [15:34] it was never explicit [15:34] * ttx just doublechecked the server ISO was ok :) [15:36] smoser: I'm on it [15:36] well, i just pushed a branch [15:36] if you want to wait for that [15:36] ah [15:37] on its way to lp:~smoser/ubuntu-seeds/ubuntu.maverick.bug649833 [15:39] smoser: merged [15:40] :) [15:40] nice [15:40] before i even proposed [15:40] how fast to those seed things take affect ? [15:40] i thiknk some process has to run to update seeds, right ? [15:40] smoser: no clue, will ask on #ubuntu-release [15:40] yep [15:40] germinate [15:42] whoowhoo. a .4 build today :) [15:48] smoser: so, it looks like you need to ask for a manual push after the next germinate.cron run, in ~1h, and then trigger a cloud image build [16:00] ttx: I think I have a handle on bug 569900 at last [16:00] Launchpad bug 569900 in mdadm "mount: mounting /dev/md0 on /root/ failed: Invalid argument" [High,Confirmed] https://launchpad.net/bugs/569900 [16:01] cjwatson: a risky handle ? [16:03] ttx: if I'm right, it's an adjustment to my previous attempt to fix the problem [16:03] simply increasing the chunk of space we ensure is clear at end of disk [16:04] I'm trying to establish a bit more certainty about this though [16:10] hey, what would be the best way to catch kernel panics and write them to disk? - linux-crashdump? [16:14] ttx: and, if I'm right, the reason it resisted analysis up to now is that it's very sensitive to the exact size of the disk, and if you use a round number of MiB then you'll never see it ... [16:14] (and my habit involves things like 'kvm-img create -f qcow2 foo.img 10G') === mathiaz_ is now known as mathiaz [17:02] New bug: #649930 in vsftpd (main) "package vsftpd 2.2.2-3ubuntu7.1 failed to install/upgrade: ErrorMessage: subprocess installed post-installation script returned error exit status 3" [Undecided,New] https://launchpad.net/bugs/649930 [17:06] Hi i'm running ubuntu server 10.04 on virtualbox and network is fine samba configured. I can access my www folder ok,but on windows 7 im running html editor nvu when i try to save the page on www it says: "save failed" someone could help? === ivoks-afk is now known as ivoks [17:15] anyone? what would be the best way to catch kernel panics and write them to disk? - linux-crashdump? [17:16] I think the old and well-proven way is a serial console, a panic may cause I/O to fail (intentionally or not) [17:17] problem is no physical access to server :( [17:17] * SpamapS recalls his first I.T. job where the SunOS voicemail system had a printer hooked up to the console for just such debugging duties. [17:17] hackeron: are there anyone with physical access to it you can call? [17:18] hackeron: there are some very affordable modems that you can purchase that will call you whenever there is activity on their serial port. ;) [17:18] RoyK: yeh, to press the power button but I added kernel.panic = 2 to /etc/sysctl.conf so havent needed to ask - just want to know why its kernel panicing every month or so [17:18] heck I bet there are *thousands* just sitting in peoples' closets ;) [17:19] RoyK: cant afford downtime so I get the server to reboot asap [17:19] ic [17:19] and nothing in /var/log/messages? [17:19] I think it's the dvr card thats causing it so it wont affect IO I dont think [17:19] iirc linux will log a panic to syslog if it can [17:19] hackeron: if you can't afford downtime, you should have a remote console of some sort hooked up so you can read those dumps. [17:20] lol.. a DVR that can't afford downtime? ;) [17:20] hackeron: have you tried to enable syslogging to another host? [17:20] RoyK: hmm? - I've never had a panic logged to syslog, even when I cause one [17:20] SpamapS: it's a CCTV DVR - I just get the server to reboot [17:21] hackeron: I may be thinking of OOPSes - they usually go to the syslog [17:21] RoyK: yeh, I have kernel.panic_on_oops = 1 so they dont [17:22] hackeron: but if you can call someone, mail them a null modem cable and ask them to connect a cable to somewhere else from where you can access it - or even - configure a network console to some box [17:22] !network console [17:22] dumb bot [17:23] !network console is https://help.ubuntu.com/community/Installation/NetworkConsole [17:23] Error: I am only a bot, please don't think I'm intelligent :) [17:23] ubottu: network console is https://help.ubuntu.com/community/Installation/NetworkConsole [17:23] Error: I am only a bot, please don't think I'm intelligent :) [17:23] serial consoles are good for this kind of thing, yes [17:23] fsck [17:23] I will look into network console, thanks - but anything I can do at all with software? [17:23] RoyK: I doubt it likes the factoid name being multiple words [17:24] !networkconsole [17:24] ubottu: networkconsole is https://help.ubuntu.com/community/Installation/NetworkConsole [17:24] the purpose of linux-crashdump is for this kind of thing, yes [17:24] it's a more complex pile of stuff and therefore more prone to failure [17:24] (just by immutable natural laws) [17:25] hackeron: https://wiki.ubuntu.com/Kernel/CrashdumpRecipe may help you [17:25] analysing the results can be a bit challenging though [17:26] cjwatson: yeh I get that but if it's a hardware fault chances are a serial console wont work anyway if the box cant reboot itself, and if it can reboot itself, surely there's a way to log the panic? [17:26] you don't just get the panic, you get a complete dump of memory state [17:26] panics get written to serial console very reliably, IME [17:26] filesystem I/O is SIGNIFICANTLY more complicated than rebooting [17:27] or than writing to a console (serial or otherwise) [17:27] writing to a filesystem in a panicked state is really not a good plan [17:27] either some kind of console arrangement or linux-crashdump is likely to be the best you'll get [17:28] cjwatson: hmm seems to work on mac and windows - get a nice report problem dialog on reboot with the panic -- I understand it won't work every time, but it's something I'd like to try :) [17:29] linux-crashdump (LKCD) is a lot simpler than filesystem I/O - it basically dumps memory state to the swap partition, which is retrieved afterwards [17:29] hackeron: linux-crashdump is the best building block for this you are likely to get [17:29] I suspect the panic is coming from a v4l driver so dont think thst would affect filesystem IO? [17:29] EOT [17:29] cjwatson: ok, cool - seems last release was 2005 though - it still works? [17:30] see the wiki page I gave you earlier [17:31] cjwatson: yeh no mention of maverick ther [17:31] cjwatson: and says problems even in lucid with apport [17:34] sure, those are relatively minor problems as I understand it [17:38] cjwatson: tried it, installed, rebooted, then ran 'echo c > /proc/sysrq-trigger' - system rebooted, nothing in /var/crash [17:38] cjwatson: any ideas? [17:39] sorry, no [17:40] I have to withdraw from this conversation now - I intended to just provide quick pointers, I'm afraid I don't have enough time to get involved in details [17:49] cjwatson: ok, thanks, I'll ask on ubuntu+1 [17:50] Ubuntu JeOS, I'm unable to get programs that segfaults to produce core dumps. I've tried to "ulimit -c unlimited", but this does not come into effect as coredump is stuck at 0. Is there a way to fix this? [17:51] New bug: #650439 in mysql-dfsg-5.1 (main) "package mysql-server-5.1 5.1.37-1ubuntu5.4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/650439 [17:51] hi all. from ps -ef, I get 'cassiani 23637 1 33 17:40 ? 00:23:45 ./flexpart80_average_yy.exe' - is there a way to find the full path of the file running? [17:56] RoyK: try "ls -l /proc/23637/cwd" [17:56] Laverne: thanks - I found it [17:57] your welcome === jj-afk is now known as jjohansen [17:58] Hi, I'm not sure if this is more a question for #ubuntu or #ubuntu-server but I have a headless machine that users use VNC/NX to connect to, sort of like a terminal server with a GNOME session for users. Since this is a shared machine, I want to disallow users from restarting the machine from GNOME. How do I remove this ability from GNOME session users? [18:01] New bug: #650443 in mysql-dfsg-5.1 (main) "package mysql-server-5.1 5.1.41-3ubuntu12.6 failed to install/upgrade: subprocess new pre-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/650443 [18:01] New bug: #650446 in eucalyptus (main) "Cloud exhausting Permanent Generation space" [Undecided,New] https://launchpad.net/bugs/650446 [18:01] heh? don't you have to have root permisions to restart? [18:01] Hello, on ubuntu server 8.04 how to run the setup script (the one that lets you choose which servers to install) from prompt? [18:03] dselect? [18:04] No, theres a ncurser selections that is available during install... [18:04] *ncurses [18:05] patdk-wk: Apparently not when NX/VNC spawns a session using /etc/X11/xinitrc or something like that. There is a restart button but I'm scared to test it out to see if it works. [18:08] https://answers.launchpad.net/ubuntu/+question/83825 [18:15] patdk-wk: Thanks, I'm looking at it. Most of those solutions are about removing the entries from the GNOME panel but users would still get a prompt to restart the machine after updates, right? So I'm looking at policykit-gnome but the instructions seem to not apply to me. [18:17] heh? upgrades wouldn't touch that [18:18] I mean, after doing a dist-upgrade, users get a prompt when you log in, "This machine needs to be restarted. Restart now/restart later". I'm assuming they get this regardless of whether the buttons are there in the panel applet. So I'm trying to forbid it through policykit. [18:22] ttx: http://people.canonical.com/~pitti/workitems/maverick/canonical-server.html ??? [18:24] kim0: there (#ubuntu is too confused...) [18:46] hello ppl please someone help me with bind9 i have rndc issue i have run rndc-confgen and updated both named.conf and rndc.conf file the key is same but i have this error: http://pastebin.com/hZV5TAH8 [18:47] Im sick fixing this error i really need some help :-) [18:57] SpamapS: nice, next time we should aim for rainbow colors [19:01] ttx: blue is too calming to motivate us [19:01] SpamapS: hopefully the next db run will fix it [19:02] SpamapS: it was ok a few hours ago [19:04] hggdh: ping ? you needed something ? [19:05] kim0: after the server meeting, if you do not mind [19:05] hggdh: sure [19:36] is it possible to do both usermod networking and bridge mode networking at the same time? [19:36] *usermode [19:36] <|rt|> hey guys I just got a new server that's using a brand new areca raid controller which the built in driver fails to detect....so I downloaded and compiled the driver source from areca on another box and put the ko file on a usb stick...but in the installer when I tell it to look on removable media for the driver its still not working [19:36] with libvirt [19:36] <|rt|> could it be b/c the kernel version is different on the install cd compared to the server I built the driver on? [19:37] <|rt|> there's no errors that give me any clues in dmesg [19:43] Back for a quick sanity check: [19:43] Does the ubuntu bootloader still have problems booting from a Raid-5? [19:44] <|rt|> assuming your talking about md raid boot has to be on raid0 or raid1 [19:44] ah shazbot. [19:44] <_ruben> i always use a small raid1 array for /boot [19:44] I assume I can take a little sliver out of my raid-5 for that? [19:44] <_ruben> wasnt even aware of the raid5 issue, im used to doing that because of lvm ;) [19:45] 100MB from each drive or something ridiculously small? [19:45] <_ruben> 100MB should be plenty [19:45] <_ruben> i tend to do 128MB due to its power of 2 ;) [19:45] <|rt|> 100MB works...all depends on how many copies of the older kernels you want to keep around [19:45] <|rt|> _ruben: I do the same thing [19:45] Right -- so I'd have a raid-1 running /boot, 128MB in size, then another Raid-5 many terrabytes in size. [19:46] <_ruben> JavaAtom: that's how i do, scratch that, did it ;) [19:46] gotcha [19:46] _ruben: Much appreciated sir. [19:46] <_ruben> no my / including /boot is on a 2G flashdisk ;) [19:46] <_ruben> s/no/now/ [19:46] heh. [19:46] I'd do that, but I've already capped the SATA connections on this box. [19:47] and it has enough of a time booting what I've got. [19:47] <_ruben> the flashdisk is pata ;) [19:47] heh [19:47] No reason not to do it later. [19:47] Thanks [19:47] <_ruben> doesnt even have a cable, the disk sits in the connector [19:47] heh. That's pretty sweet. [19:47] <_ruben> (does have a (tiny) power cable obviously) [19:47] << runs off to install server. [19:48] <_ruben> though with default -server installs, 2G tends to be just enough, gets tricky with full version upgrades though [19:52] smoser: Any link with more info on the pv-grub feature ? [19:52] hggdh: so did you need something ? [19:52] kim0: yes, I will PM you [19:53] kim0, yes. [19:53] smoser: is there any other major cloud related feature to maverick (besides cloud-{init,utils}) ? [19:53] http://aws.typepad.com/aws/2010/07/use-your-own-kernel-with-amazon-ec2.html [19:54] i'd say some improvements in cloud-init (and acceptance of cloud-init into amazon ami) [19:54] plus the use of pv-grub [19:54] sounds great :) [19:54] and our images also can service their own kernels on uec [19:55] thanks man [19:55] jjohansen: what's the best linux kernel book in your opinion? [19:55] uhmm, depends [19:55] ldd3 is fair but its not a good intro [19:56] jjohansen: right now i'm uising LInux Kernel development by Robert Love (google employee) [19:56] I always found robert loves book a better read, better introduction but its getting old [19:56] yeah, its good but quite old [19:56] jjohansen: is there something better? [19:57] hrmmm, not that I know of but I really haven't been looking [19:57] jjohansen: alright thanks :) [19:57] I heard a rumor he was updating his book though [19:57] cool :) [20:00] <|rt|> anyone ever had to use external media for a raid controller during installation? [20:00] jjohansen: btw.. the update you talking about would be the 4th edition right, since the 3erd was published july this year and thtat;s the one im using [20:00] <|rt|> or know if there's any instructions on how to do that anywhere [20:00] ah, no I didn't know the 3rd had been published [20:01] I haven't looked since hrmm, about a year ago at ALF [20:01] resolving DNS is really slow when the primary dns is down (probably because it keeps attempting to use the primary for seconds with every query) is there a way to speed things up? [20:02] jjohansen: oh ok :) we are using this book for a course actually and it is pretty good [20:03] yeah, it is the best of the one I have read, my only complaint was that it was getting dated [20:03] but so is ldd3 [20:03] * RoyK really needs zfs [20:04] jjohansen: :) Thanks for the input btw :) [20:06] * RoyK has an old server named fenris at work, dual 360MHz SPARC on sol8 [20:06] never mind found the timout:n option :) [20:08] here is my rndc.conf : http://pastebin.com/zwa6QbBp and here is my my rndc.key source http://pastebin.com/P5Mrfwac [20:08] and this is in my named.conf : http://pastebin.com/HygjiMyv [20:08] and rndc -c /var/chroot/named/etc/bind/rndc.conf -s 127.0.0.1 -p 953 status [20:08] prints this error http://pastebin.com/k21gYQ4M please help me figure out whats wrong [20:08] alright. 20100928.4 build for uec images started. a new high number for .X for me. === ivoks is now known as ivoks-afk [20:46] mathiaz: the print-server task fails, see bug 650566 [20:46] Launchpad bug 650566 in foomatic-db-engine "print-server task is not installable on Maverick RC" [High,Confirmed] https://launchpad.net/bugs/650566 [20:53] mathiaz: If you ack me, I'll commit the seed fix. [20:53] mathiaz: then we can wait for euca to hit, and ask for a respin. [20:57] how do i build a virtual machine with two network interfaces? [20:58] can i bridge both of them? === ivoks-afk is now known as ivoks [21:06] mathiaz: got acked from Colin, so that's done [21:08] mathiaz: now the idea would be to trigger a respin when euca is uploaded and published, and when the seed is ready [21:08] * mathiaz nods [21:10] ttx: I'll sponsor eucalyptus [21:13] # Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8) [21:13] # DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN [21:14] Is there any other way to alter VM than undefining it in virsh and then importing modified xml? [21:14] anyone know how to set nameservers in this newfangled dynamic resolv.conf world? [21:15] <_ruben> pwnguin: uninstall resolvconf and edit that file anyway? ;) or figure out how to influence resolvconf(8) [21:15] heh [21:20] does anyone know which kernel versions include pv-on-hvm drivers? [21:20] if any [21:22] scrllock: You mean virtio? [21:23] ttx: Daviey: eucalytpus ubuntu4 sponsored to maverick [21:23] soren: no, one with the CONFIG_XEN_PVHVM=y [21:23] So, I'm having trouble with cron. It's somewhat related to this bug: https://bugs.launchpad.net/ubuntu/+source/upstart/+bug/592114 but not quite the same symptoms. Cron starts at boot, but will not run the user cron jobs, even after editing or touching, until I restart cron after boot. Has anyone seen anything like this? [21:23] option [21:23] mathiaz: I'm about to stop [21:23] Launchpad bug 592114 in upstart "upstart doesn't start cron automatically on boot in lucid for server on amd64" [Undecided,New] [21:24] ttx: ok - with whom should I talk to get an iso respin? [21:24] ttx: just ask in #ubuntu-release? [21:24] mathiaz: yes [21:24] mathiaz: not sure who will be up at that point [21:24] you need to wait for build/publication [21:24] that should take a couple hours [21:24] * mathiaz nods [21:26] scrllock, that option is not even present in the default kernels [21:26] apw: yeah, that's why i was asking in -kernel [21:27] it's in the mainline 2.6.36-rcX, but that kernel is a bit too buggy [21:30] <_ruben> pwnguin: the manpage seems pretty decent though ;) [21:33] _ruben: i ended up just removing resolvconf =| [21:33] <_ruben> pwnguin: probably the easiest "solution" ;) [21:34] <_ruben> must i prefer to be in control of my config files myself as well [21:34] <_ruben> +admit === _TechAway_ is now known as _Techie_ [21:35] mathiaz, thanks === lifeless_ is now known as lifeless [22:18] is there a package that contains dos2unix/unix2dos on 10.04 ? [22:20] found it, tofrodos [22:21] to Frodo's ;) [22:25] we go [22:25] :) [22:28] hey folks, what is the best way to do exactly what active directory can do? thanks!!! [22:28] in ubuntu server 10.04 sorry [22:29] <|rt|> I need to supply a driver for my raid controller to the installer. do I need to compile the module against the headers for the kernel version on the install cd or would the newest server headers work? b/c when I go to load the module I get an error saying it's not a valid module format [22:30] <|rt|> illytacos: the easiest way is to use a windows server with AD [22:30] lol I don;t have a windows server [22:32] <|rt|> illytacos: AD does a lot of things. Basically it's a ldap, kerbose, and dns mash up [22:35] <|rt|> so you can approximate most of it's functionality using those protocols...but it's a pretty daunting task as each of those are things are pretty deep subjects [22:43] Personally I think thats something that most corporate/large organizations find daunting about Linux migration. That they actually have to setup krb/ldap/dns instead of just installing a server and going forward. [22:47] Samba 4 is still alpha.. wow.. its been years. :-P [23:14] does ubuntu has its own virtio drivers for win 2k3, or do i have to use red hat's?