[04:12] hi guys. I have a sort query. I have a list of IP's I'd like to sort in sequence, but 192.168.0.100 preceeds .10 preceeds .1 [04:12] what argument can I use to put them in order? [04:15] kieppie1: try -g [04:15] or -n [04:16] did try numeric. found an eventual solution: sort -t . -k 1,1n -k 2,2n -k 3,3n -k 4,4n [04:16] thanks anyway [04:17] np === ujjain is now known as ujjain|afk === incidenc1 is now known as incidence [08:33] I'm getting lots of MCE's and reboots (primarly at night). But how to collect them? 'mcelog' doesn't seem to log anything anywhere. === airtonix_ is now known as airtonix === ujjain|afk is now known as ujjain [09:33] whats best vmware server 2 or virtualbox on ubuntu 8.04 Hardy? === cjwatson_ is now known as cjwatson [09:41] LinSkyrate: that's going to depend on your goals, I prefer virtualbox personally [09:41] greppy: oki.. just downloaded it.. i wil give it a run [09:43] greppy: how can i install virt box with dependencies? i use dpkg -i but that fails on deps [09:52] LinSkyrate: use apt-get install instead of dpkg, that will grab the dependancies. or you can try 'sudo apt-get -f install' to see if it will fix what it needs [09:53] greppy: jupp.. but i need to set the repo for hardy [09:55] why not use lucid? [09:56] LinSkyrate: why not use 10.04 aka lucid? [09:58] because im running other things that demands 8.04 [09:59] ouch :( [09:59] but what package is this in : libxcursor1 (and a lot of others that are missing? [10:02] i get this error: Avhenger av: libxcursor1 (> 1.1.2) men skal ikke installeres Avhenger av: libxrandr2 (>= 2:1.2.0) men skal ikke installeres Avhenger av: python2.5 (>= 2.5) men skal ikke installeres Anbefaler: libhal1 (>= 0.5) men skal ikke installeres Anbefaler: libpulse0 men skal ikke installeres Anbefaler: libsdl-ttf2.0-0 men skal ikke installeres Anbefaler: pdf-viewer [10:05] no idea what the errors mean ( since I only read english ) bt I think it is complaining about versions not being available [10:06] greppy: the version is ok but miss all that installed to go on [10:07] do i need ubuntu-desktop installed? [10:07] you might [10:08] if my norwegian is correct, something depends on libxcursor1, but that it cant be installed. [10:08] my box with virtualbox running on it is in another building right now, so I can't really check. [10:09] EricJ: it just says that it needs this installed but i have not done so [10:09] LinSkyrate: run LANG=C apt-get ... [10:09] LinSkyrate: so if you just try to explicitly install those dependencies? [10:10] EricJ: jupp [10:10] im installing ubuntu-desktop to see if it contains all that [10:10] Oh dear. [10:10] he he [10:10] LinSkyrate: erm - that probably won't help [10:10] LinSkyrate: which version is this? [10:11] RoyK: 8.04 Hardy [10:11] oh [10:11] could you try Lucid? [10:11] RoyK: i need that version for a stable Ulteo [10:12] what's an ulteo? [10:12] Ulteo is VDI [10:12] Virtual Desktop via web [10:13] then you're probably talking about a desktop installation, right? if so, strictly speaking, this is the wrong channel... [10:13] im going to make a simulation for some money people, and they need to see some eyecandy to put the money on the table for my projects [10:13] I wonder why they still stick with 8.04, though [10:14] it's not getting newer, and the eyecandy will be better with Lucid [10:14] they have a 3.0 version, but thats to new and buggy [10:15] can i do a command line upgrade to Lucid? [10:15] im sitting remote for this server [10:15] do-release-upgrade [10:16] RoyK: ?? [10:16] just that? [10:16] check /etc/update-manager/release-upgrades first [10:16] that should be set to Prompt=lts [10:16] and yes, just that :) [10:17] will it not get me up to 11.04 then? [10:17] it'll take you to the next LTS, meaning 10.04 Lucid [10:17] or if Prompt=normal, it'll take you to the next release, 8.10 [10:20] RoyK: will do that, but must finish ubuntu-desktop install first as i already started that [10:35] RoyK: its says No new release found? [10:35] i changed to lts [10:37] erm... [10:37] RoyK: should i put anything else there than lts? [10:37] can you doublecheck the version? lsb_release -a [10:37] no, lts is fine [10:39] RoyK: shit.. this release are in 11.04 for some reason [10:39] heh [10:39] I've seen some rather messy dependency issues in 11.04 [10:40] RoyK: can i downgrade do Lucid? [10:40] reinstall [10:40] to [10:40] æsj [10:40] jau [10:40] thats shitty its a remote server [10:41] didn't you say it was running Hardy? [10:41] it doesn't upgrade all on its own... [10:43] must have mixed the cds [10:43] because this should be 8.04 [10:43] did a plain server install yesterday [10:45] I thought you were talking about some virtual installation? [10:45] nope [10:45] thats inside [10:45] ok [10:45] its not virtual.. its a controlpanel for virtual desktops [10:47] ok [10:49] and you don't have something like a BMC/IPMI/ILO/ILOM/ALOM in the machine so you can control it remotely? [10:49] nope... :( [10:50] * RoyK likes SuperMicro's IPMI - supports installing a server from local ISO images :) [10:51] took me 20 minutes to install four Lucid servers at once (a quad thing - four mobos in 2U) [10:51] dont have that in this machine [10:52] still get the missing libs when installing virtbox 4 [10:53] RoyK: Using ext4? [10:54] yeah [10:54] why? [10:57] RoyK: previously ext4 has suffered from slow installation, due to the fsync's between every operation. [11:07] Daviey: dunno about that === Guest26577 is now known as Adri2000 [11:53] ping Daviey [11:54] lol - seems the guy working on fsck for btrfs works for oracle [11:58] RoyK: hah :) === kim0_ is now known as kim0 [12:26] New bug: #796551 in openssh (main) "package openssh-server 1:5.8p1-1ubuntu3 failed to install/upgrade: Unterprozess neues pre-installation-Skript gab den Fehlerwert 1 zurück" [Undecided,New] https://launchpad.net/bugs/796551 [12:26] New bug: #796552 in mysql-5.1 (main) "package mysql-server-5.1 5.1.54-1ubuntu4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/796552 === med_out is now known as med === med is now known as medberry === NG_ is now known as ng_ === ng_ is now known as NG_ [13:53] I apt-getted roundcube, and noticed that index.ph in /usr/share/roundcube is owned by root (which is the the indexfile /var/lib/roundcube/index.php (docroot) links to). Why isn' t this www-data? [13:53] *index.php and the same goes for .htaccess [13:54] the reason I ask this is that I get a "SoftException in Application.cpp:296: UID of script "/var/lib/roundcube/index.php" is smaller than min_uid" [13:55] which is of course true if the owner (and group) are root [13:59] foo-nix: that looks like a packaging issue [14:00] foo-nix: I don't see it on the bug report https://bugs.launchpad.net/ubuntu/+source/roundcube [14:00] foo-nix: you would like to fill a bug for it? :) [14:03] lynxman, sure, but I have no experience with ww-related packages [14:03] i.e. which install roundcube, squirelmail, phpmyadmin, etc [14:03] morning, and happy Monday [14:13] foo-nix: ooh I see [14:13] foo-nix: well your guess is right, it should be www-data, I'll try to reproduce the problem later and see if I have the same issue as you [14:14] lynxman, unfortunately, after chmodding I have the same problem, I will restart apache now [14:14] although I wouldn't understand why :P [14:14] foo-nix: hmm what story do the logs say? [14:15] oh, same error [14:15] SoftException in Application.cpp:296: UID of script "/var/lib/roundcube/index.php" is smaller than min_uid [14:15] foo-nix: /var/lib/roundcube... [14:15] foo-nix: then you're talking also about the index.php in /usr/share/roundcube [14:16] yeah [14:16] foo-nix: is there anything in /var/lib/roundcube? there should be right? [14:16] the index.php is linked there [14:16] foo-nix: is the link made by root? [14:17] lynxman, I think so, I don't really know how linking works with userrights [14:17] wait [14:18] http://imagebin.org/158053 [14:18] lynxman, that's a listing [14:18] foo-nix: okay but that's the one in /usr/share/roundcube [14:19] foo-nix: what's the content of /var/lib/roundcube [14:19] lynxman, wait, I just chmodded the files to www-data [14:19] initially most of them were roor:root owned [14:19] *root:root [14:20] foo-nix: if they were there must be a reason behind it... [14:20] lynxman, I think so, but then why do I get this error? [14:20] and why wouyld an index.php be owned by root? [14:21] foo-nix: must be something else related [14:21] foo-nix: I would need to dedicate some time to try to reproduce it [14:21] lynxman, I don't know much about how to configure www-packages and where ubuntu puts stuff [14:21] foo-nix: I would recommend you though, uninstall the package "apt-get purge packagename" [14:22] foo-nix: and then install it again, don't chown anything [14:22] foo-nix: copy/paste the problem into Google, see if someone else had it as well [14:22] lynxman, yes, I find it disturbing to do so, either :) [14:22] foo-nix: since I'm sure there's something else involved, never installed roundcube myself so... :) [14:22] lynxman, others also have it, but they don' t state the owners of the files and moreover none installed from portage [14:23] lynxman, thank you very much [14:23] I have guests arriving in a bit more than 30 minutes [14:23] foo-nix: no problem, let me know how the purge and reinstall goes [14:23] foo-nix: oops, hurry up then ;) [14:23] hehe [14:23] so I will be less availiable [14:23] lynxman, but if you would like to have some info, feel free to e-mail me at herbert@fmf.nl [14:24] I typically respond quite fast. [14:24] foo-nix: cool :) [14:25] indeed :) === jcastro_ is now known as jcastro [14:28] foo-nix: http://www.crazysquirrel.com/computing/debian/servers/ubuntu-mail-server.jspx [14:28] Down in the roundcube section. [14:28] ;) [14:28] Might help, sounds related. [14:31] "Sounds like a good idea but unfortunately the packaging of RoundCube (and I'm guessing a lot of other stuff) hasn't caught up with the change yet. When I installed RoundCube it with Aptitude it, of course, went in owned by root." [14:31] thanx [14:33] There should be nothing wrong with a web app being owned by root. Did I miss something? === Adri2000 is now known as Guest87619 [14:34] Oh, suphp.. or "the worst idea ever" ;) [14:34] SpamapS, suphp runs these scrips ar the file owner [14:34] SpamapS, it sounds uite secure if you ask me. [14:34] *quite [14:35] SpamapS, this allows for a website to own it' s own file [14:35] makeing sure the one site can' t fiddle or piggy-back on the other site (' s security issue) [14:35] hmm, you want to run roundcube per user? [14:35] shared hosting is a fail [14:35] why not run it per server, and be done [14:36] We have KVM to solve this problem for us now. [14:36] Are you suggesting to install a new server each time you setup a new site? Or some virtualserver configuration? That' s complete overkill [14:36] I would put shared hosting on a different box, than the email server [14:36] yes [14:36] foo-nix: its a single command. :) [14:37] SpamapS, rm -rf ~/ is also a single command [14:37] SpamapS, I like the idea of virtual machines [14:37] SpamapS, but I also like the idea of not having too much of them. [14:38] foo-nix: w/ appropriate config management you can manage thousands of them at once. [14:39] They're as easy to manage as processes now, and without all of the hacks that we built around isolating processes from eachother over the last 20 years. [14:39] SpamapS, sure, but that doesn' t make virtualhosts bad [14:39] no, it makes virtualhosts obsolete. obsolete still has value. [14:40] just like ipv6 should, some day, make nat obsolete. :) [14:40] SpamapS, anyhow, the server of my university does it the vhost way :) [14:40] I would say that suphp/suexec should have options for refusing to run as root. [14:41] I'm not going to allow php scripts to run as root from a http-request [14:41] foo-nix: we're actually working on a system for making deploying and managing a service like roundcube as easy as 'apt-get install roundcube' .. [14:41] gday [14:41] foo-nix: http://ensemble.ubuntu.com/ [14:42] this stuff is cool, although it is too bad that the first lecture on clod computing I attended was so bad === jetole_ is now known as jetole [14:42] hahaha clod computing [14:42] *cloud [14:43] I think we should call it clod. :) [14:43] I hurt my back working out a few days ago, and now typing and even sitting at a desk is agony :P [14:43] foo-nix: cloud is just the easy way to do it now. We'll be making it work w/ local vms and containers too. [14:43] foo-nix: maybe try a standing desk [14:43] cloud => let others fix it for you :) [14:44] SpamapS, thanx :) [14:45] How do you start/stop the heimdal kerberos? === Guest87619 is now known as Adri2000 [14:50] foo-nix: one could say that about server co-location too.. I mean.. its not like you're running your own MAP on the net, right? [14:51] SpamapS, what is a MAP? [14:51] sorry, I dont know :P [14:52] and disambiguation is hard with map :P [14:52] Hey... Does anyone have a windows 2003 server? :) [14:53] foo-nix: I meant MAE ;) [14:53] I could :) [14:53] patdk-wk: really? [14:53] ya, about 70 or so [14:54] patdk-wk: Are you running an ntp server? [14:54] yep [14:54] patdk-wk: fancy presenting it to me, want to test something? :) [14:55] heh? [14:55] :P I don' t know MAE either, is it some kind of bussiness word for network? [14:55] foo-nix: http://www.linktionary.com/m/mae.html [14:56] they are all joined into AD [14:56] patdk-wk: seriously, want to reproduce a bug. :) [14:57] sure, but I can't give you rdp access [14:57] SpamapS, ah, now I get it, sorry [14:57] patdk-wk: No, i just want to access your ntpd from UDP. [14:57] (as a consumer) [14:57] my ntp? or ntp on win2003? [14:57] Daviey: the montreal people should have access to win23k box [14:57] I run my ntp servers in linux [14:57] well, co-location meaning the server is on a different location then the bussines/users? [14:57] patdk-wk: wise :) [14:58] zul: good thinking [15:04] foo-nix: yes. that is the norm for pretty much all web properties that aren't google/amazon size. [15:06] there' s chrome laptop coming which operates in the cloud [15:06] Heh.. ChromeOS has its roots in Ubuntu. [15:08] SpamapS, you didnt know that? [15:08] zul: I was noting it for foo-nix [15:08] SpamapS, ok maybe im not quite awake yet [15:08] SpamapS, ok, why? [15:11] foo-nix: why what? === smoser` is now known as smoser [15:18] Daviey, do you have an updated version of http://pb.daviey.com/Cy6D/ ? or is one easy to produce ? [15:25] When I boot up 11.04 Server, I get "Screen can not display, not recommended resolution". Is there a way I can change the resolution of my server? [15:33] smoser: yes! [15:33] smoser: http://people.canonical.com/~davewalker/delta.txt (updates every 4 hours). Also now includes packages ~ubuntu-server is subscribed to and the packageset. [15:37] Daviey: every 4 hours the info gets stale ;) [15:38] Daviey: also can you blacklist some of the packages on your script [15:39] zul: wel i thought more frequent might be a waste of time. [15:39] Daviey: yeah i was pulling your leg [15:40] zul: Yes, could blacklist some.. But the fact they are in the packageset and subscribed to should mean we need to care about them [15:40] Daviey: well im thinking about cups since its more desktopy [15:41] zul: hah, and gimp is totally server centric :) [15:41] Daviey: well yes i need gimp to my network diagrams ;) [15:45] hello, how do I guarantee a certain order in sites-enabled? I'm using name based vhosts and I need one of the hosts to work as catch all [15:46] katol, use numbers as prefixes. [15:46] they are ordered in an alphanumeric order. [15:46] Daviey: oh ok, is that considered a good practice? [15:47] and just to be sure, there is not other way to define a catch all without overriding the other vhosts from the configuration itself, right? [15:48] heh? [15:48] using apache? [15:49] the catchall is defined using default [15:49] not via order [15:51] patdk-lap: yes apache, default as in _default_ ? I think that does not work for name based vhosts, only IP/port based [15:51] oops [15:51] i meant patdk-wk [15:52] ah it doesn't, hmm [15:59] oh well, I'll just use the number prefixes [15:59] thank you Daviey and patdk-wk ;) [16:06] smoser: ping [16:06] here [16:07] smoser: can i trade patch pilot days with you [16:08] patch pilot days and cash [16:08] what is your day? [16:08] tomorrow :( [16:08] zul, you don't need to 'swap', just change to a day that suites. [16:09] Daviey: k [16:09] why did Daviey feel he needed to stop zul from buying my patch pilot day [16:09] :-( [16:12] smoser, lol [16:14] lol [16:28] smoser: I see you've just retired the day from ebay... [16:36] Daviey: did you chair last weeks server meet? [16:36] jamespage, i did indeed. [16:36] marvellous - that means its me in the seat for tommorow.... [16:36] jamespage, splendid [16:36] however sprinting so will try to swap with RoAkSoAx: how about it? [16:38] RoAkSoAx, ? [16:39] Daviey: hmm - did I miss the email that said we don't need to writing up meeting minutes anymore? [16:40] * Daviey also offers out chairing to anyone else.. it is open to community contributors aswell :) [16:40] jamespage, no.. [16:40] jamespage, zul is doing week before last at the moment. Then i'll do last weeks. [16:40] * jamespage pokes Daviey with a stick some more [16:40] :-) [16:41] jamespage, sarcasm doesn't become you :) [16:41] Daviey: ??? [16:41] RoAkSoAx, jamespage asked if you could take his chair spot tomorrow. [16:41] Daviey: yeah just read [16:41] jamespage: sure [16:41] Daviey: must be all those java library forks I've been reviewing today :-) [16:41] jamespage, you signed up for it :P [16:41] RoAkSoAx: excellent - thankyou [16:42] yeah - its just been a bit grinding today ;-) [16:42] jamespage: ;) [16:43] jamespage: can you point me in the direction of documentation to write/create/try out automated tests? [16:43] SpamapS: yeah sure - lemme just dig something out [16:44] where do you want todo it? in the cloud or in kvm images? [16:44] kvm, its a lot of boot testing.. [16:46] Right - so I'd recommend you branch lp:ubuntu-server-iso-testing [16:46] thats the framework that currently tests our ISO images across server, desktop and alternate [16:46] cool [16:47] README? man page? [16:47] ok [16:48] "hello" [16:49] 10 PRINT "HELLO #UBUNTU+SERVERÆ [16:50] s/.$/\"/ [16:50] SpamapS: just reminding myself [16:51] :D [16:53] SpamapS: docs/README does give you a bit of insight but its more geared towards running the existing tests [16:53] * jamespage takes an action to add some developer documentation for writing test-cases [16:55] * jamespage also takes an action to make the templates directory configurable from the command line [16:55] good job I'm at the QA automation sprint for the next two days - might actually find time [16:58] jamespage: I'll give it a whack and see what I come up with. :) [16:58] SpamapS: ack - the run-test.py script and the templates directory should be most interesting [17:00] * jamespage thinks that it may need a little refactoring to support SpamapS requirements [17:00] jamespage: the only requirement I'll have is that I can reboot 2 or 3 times [17:01] SpamapS: a limitation of the current framework [17:03] I think you will need to specialise the templates/run_test - this is the script that gets injected into the vm [17:03] it only deals with boot -> run tests -> shutdown at the moment [17:08] jamespage: well technically I'm not shutting down until after a few reboots. :) [17:11] SpamapS: yeah - so that should deal with the 'testing complete' monitoring (which is when the vm shuts itself down) [17:11] you will need to adapt the way the test's are executed to deal with reboots [17:15] Hey folks, I now have: http://ubuntu.esotericsystems.at/ with my rebuilds of OpenLDAP (with OpenSSL support) -- how do I make this a full blown repo? with pool/dist/blah directories? [17:16] kirkland: ping? [17:17] is there a way to see a list of changelog entries with your name in it? [17:17] (ie, for developer application) [17:18] smoser: yes [17:18] smoser: lp:~andreserl/+junk/search_uploads --> originally created by Daviey [17:19] hm.. [17:19] i dont like using things that Daviey wrote [17:19] smoser: lol... i tried to improve it though xD [17:25] Hi... I've setup ntop, and finally got it working with my connection, but now all my attempts to access my websites fails... lewiscawte.info for exampl [17:26] Does this look sane to people: http://people.canonical.com/~chucks/libxen-libvirt.patch [17:31] Anyone have any idea why this is happening [17:35] smoser: i'm not proud of that... and it's damm slow. [17:36] Sadly, the LP api doesn't extend to doing what you want. [17:37] smoser: if you want to fix, LP #610491 - that would fix your requirement :) [17:37] Launchpad bug 610491 in launchpad "[API] Please expose getPublishedSources(package_creator,package_signer)" [Low,Triaged] https://launchpad.net/bugs/610491 [17:37] (Although doesn't extend to [YUR NAME] entries in changelog.) [17:38] smoser: Talking of which, what is a CoC friendly way of nack'ing and applciation? [17:38] *an === lullabud is now known as warzauwynn [17:48] Daviey: "sod off" :) [17:54] * RoAkSoAx goes to lunch === medberry is now known as med_out [19:24] Hi :) [19:25] Ah, okay my server appears to have crashed several tiems now, I've left it running on the console mode and I now have a big page of text but what Im curious is how to I. use this information [19:25] It looks like a kernel stack-trace? I think.. [19:25] Oh and Caps and Scroll-lock are blinking on the keyboard [19:26] Incase thats somehow relevent [19:26] slyboots, take a picture of it and submit a bug [19:26] well, I sort of need to get the system backup now.. and my boss is going to want to know what caused the crash [19:27] So, Is there any sort of general advice that would be helpful in this situation in prehaps preventing it gain rather than "Yea it just crashed" [19:31] well, unless you know how to read that info, there is nothing we can really do to help you [19:31] New bug: #796750 in apache2 (main) "apache2 startup fails with missing log directory" [Undecided,New] https://launchpad.net/bugs/796750 [19:31] but if it's kernel stuff, you have a very limited set of issues [19:32] I .. *think* its a kernel issue [19:32] hardware failing, kernel driver failing, a kern param ran out of stack/mem/... [19:32] Its taking down the entire system so.. [19:32] well, I dunno how you want someone over irc to tell you what it is [19:32] :P [19:33] this is when serial console with serial logger is nice [19:57] WinstonSmith: last [19:57] oops, sorry [19:58] I just spent a little time googling and searching apt but did not find the answer I was looking for.. how does one add the standard man pages to ubuntu-server [20:00] CrazyGir: sudo apt-get install man-db [20:01] which should be on the default install.. but still. [20:05] hrm [20:05] thank you :) [20:06] also depends on what you mean by "standard man pages" [20:08] e.g. if you need the manpages for the POSIX standard, there is 'manpages-posix' (I find those occasionally useful) [20:09] Hm.. [20:09] Is there a way to test if a serial console is waiting [20:09] Uh.. working even x.x [20:10] I added grub_terminal=serial to the defaults file, ran grub install.. but Im not getting anything on the terminal [20:10] hello i was now i need to setup virtual host i use webmin but forgot how to setup so it work 100% i have setup 2 virtual host now, but one answer for all call's [20:11] JanC: by standard, I mean manpages should be on any system you go to [20:11] those manpages ;) [20:11] I can barely wrap my head around ubuntu separating them into a package [20:11] CrazyGir: embeded systems often remove all manpages ;) [20:12] and AFAIK Ubuntu doesn't separate all manpages in one package? [20:12] sure, and embedded systems often build from scratch [20:12] and I don't know one/more packages [20:12] I'm just a BSD guy and looking at this like wtf [20:13] 'man-db' contains teh man command & infrastructure [20:13] not the manpages [20:13] manpages should be in their respective packages [20:13] how do you mean respective packages? [20:14] the manpage for bash is in the bash package, etc.? [20:14] like if I am lookingfor into on the syntax for net interfaces.. [20:14] sure, but the core system? [20:14] ah, programming manpages? [20:15] or about the kernel? [20:15] maybe you want 'manpages-dev' [20:15] ..only in ubuntu [20:15] and/or 'glibc-doc' [20:15] >< [20:16] is there a package that will bring it all down? [20:16] :P [20:16] I think the BSD people are more "anal" about manpages ;) [20:16] in what sense? [20:17] I'm a BSD guy and never had an issue with manpages in Open/Free [20:17] e.g. I remember a BSD sysadmin patching the "hier" manpage to reflect the local machine ;) [20:17] they're there. period. [20:17] * CrazyGir shrug dunno about that.. [20:17] or maybe that was just him ;) [20:17] yea, probably :) [20:18] butstill, IMO utilities that come from the BSDs often have better & more up-to-date manpages [20:19] oh yea, for the most page, documentation in linux is atrocious [20:19] IIRC he told me patching something in FreeBSD is only allowed if you also fix the manpage (if that's needed) [20:19] yea, the OpenBSD devs consider a bug in doc on the same page with security bugs [20:20] which I would agree with [20:20] as they're telling you the wrong thing [20:20] CrazyGir: I don't know any meta package that pulls in every possibly useful documentation though [20:20] yea [20:20] so for typical system docs, manpages-dev? [20:20] or 'manpages'? [20:21] they contain other sections in the man system (see the package description) [20:22] no/win 17 [20:22] arggggghh [20:22] I think manpages should be installed by default? [20:23] and manpages-dev should be installed when you install a build system [20:23] (as no compilers etc. are installed by default) === med_out is now known as medberry [20:23] k, I need to get networking fixed in this VM, then I should be able to confirm [20:23] speaking of compilers.. [20:23] :P [20:24] compiler packages are on the CD ;) [20:24] never mind my ? i have fix it [20:24] but maybe you used the mini ISO? [20:26] Hi all [20:27] people, why ubuntu recommend to install ubuntu server when try to download it? [20:27] sorry [20:27] plm: what do you mean? [20:27] people, why ubuntu *64 bits* recommend to install ubuntu server when try to download it? [20:27] are 64 more stable than 32? [20:27] you mean why ubuntu-server is 64-bit by default? [20:27] sorry for my confuse sentence agaim [20:28] JanC: yes [20:28] most servers have tons of ram [20:28] most server hardware is 64-bit nowadays [20:28] like 32GB is common [20:28] but if you want 32-bits, it's available too (e.g. for small VMs) [20:28] JanC: how I know if my server is 32 or 64? just see in setup if is adm64 emt64? [20:29] plm: what sort of server is this? [20:29] my is EM64T in setup [20:30] JanC: is a mounted server.. [20:30] eh? [20:30] yes [20:30] setup show EM64T [20:30] CrazyGir: 32GB is a bit above normal [20:30] I mean, what sort of hardware are you using, how much RAM, etc. ? [20:31] RoyK: I suppose that depends on what you are doing with your servers ;) [20:31] JanC: intel core 2 duo, 3GB ram [20:31] plm: just install from the amd64 iso [20:31] plm: that CPU supports 64-bit, but you don't really need it [20:31] JanC: I have a few boxes with 64GB - but it depends of use [20:32] 64 or 32? :-) [20:32] or just 4 [20:32] or need I think : if is more than 4gb, use 64, else 32 [20:32] Hm.. Im confused by this [20:32] plm: what are you going to do with it? [20:32] most webservers will live happily with just 4GB [20:32] JanC: postgresql database server [20:32] Im trying to log all information related to this kernel crash over Serial.. [20:33] Now I know the serial cable works ( by starting a getty session on ttyS0) but the instructions I have for logging dont seem to work [20:33] I don't think postgres is (much) faster with 64-it, so maybe better use the 32-bit version then (that uses less memory) [20:33] Or at least, they are for logging into a system over serial [20:33] 64-bit [20:33] what is the proper way to _fully_ restart networking (such that all devices reinit/request dhcp if needed/etc)? [20:33] They all list adding Grub_console=serial to the grub-defaults file [20:33] service networking stop/start doesn't really do that [20:34] JanC: ok, I will use 32-bit [20:34] JanC: but i don't have problems too if use 64-bit right? [20:34] JanC: my question is really when use 32 or 64... [20:34] if just RAM more 4GB is the case [20:35] both will work [20:35] make sure you use the PAE 32-bit kernel though (it's about 0.05% slower, but offers better security) [20:36] or maybe 0.005% slower, I don't know exactly, but isn't really important ;) [20:36] the PAE kernel can also use > 4 GiB RAM with 32-bits, if you ever upgrade it [20:37] Anyone any idea? [20:38] slyboots: serial consoles are a pain in linux :( [20:38] CrazyGir: restarting "networking" should work? [20:38] JanC: for now I will install 10.4 32-bit, because I already using a old 7.4 (feisty) 32-bit. And when I will get a "real" server with 10GB ram I will install ubuntu 64-bit and does dump/import data :-) [20:38] JanC: like service networking restart? [20:38] have you tried that? :P [20:39] CrazyGir: I have done that in the past ;) [20:40] not going to mess up my local networking now halfway a download :p [20:43] what is url for choice manually ubuntu mirros to download ISO? [20:43] http://www.ubuntu.com/download/server/download is automatic [20:44] found : https://launchpad.net/ubuntu/+cdmirrors :-) === PascalFR is now known as boxershort [20:51] JanC: you cant' anymore [20:51] JanC: Restart: unknown instance [20:51] eh [20:52] or am I wrong here? [20:54] "sudo status networking" says "unknown instance"? [20:54] JanC: networking isn't the service you want [20:55] JanC: are all of your network interfaces configured in /etc/network/interfaces ? [20:57] SpamapS: it's CrazyGir who has the problem [20:57] yeah I was reading backwards [20:57] CrazyGir: whats the issue? [20:57] I guess network-manager is managing the network then? [20:58] zul: ppa:gandelman-a/puppet-module-rsyslog there are some test manifests you can run/apply in /etc/puppet/modules/rsyslog/tests [20:58] well the 'networking' upstart task is just a thing to run 'ifup -a' .. its not actually managing "networking" [20:58] adam_g: cool ill have a look tonight [20:58] I thought it triggered the other network stuff too, but maybe that's not true anymore ☺ [21:11] people, I would like to use postgresql 9.0, but 10.4 LTS has just 8.4.. what you sugest? [21:11] postgresql 8.4 is old.. [21:12] plm: I think there's a PPA with 9.0 for Lucid [21:12] plm: I used one at least :) [21:12] lynxman: but is stable..? [21:13] plm: as stable as 9.0 can be [21:13] lynxman: and secure.. and one apt-get -f dist-upgrade will not break the system? [21:13] plm: I used it in a prod enviroment, but of course your mileage can vary [21:14] lynxman: using this PPA, one apt-get -f dist-upgrade will not break the system? [21:14] plm: it didn't on mine, but again... use at your own risk, I'm not a certifier of any kind ;) [21:15] Can anyone recommend a program I can use to watch a user who is logged via SSH in realtime? I want to set up a few apache things and I need to show him what I am doing and explain it while I do that on the phone. [21:16] I know in freebsd there was a watch tty thing, i just cant find an equiv for UBuntu [21:16] lynxman: right, I know, just getting information :-) curious.. if the advantage of LTS is just a long support.. so is install 10.4 with PPA postgresql 9 is the sabe of install 11.4 with default 9.0 postgresql :-) not? [21:16] *same [21:16] tdelam: you can share a byobu session with the other user if you both login under the same user [21:16] plm: Except that the rest of your OS will reach end of life sooner than the LTS [21:17] lynxman: oh? [21:17] plm: and the postgres 9.0 ppa is community based, so clearly no support on that one [21:17] oh wait, byobu is? [21:17] tdelam: byobu is a newer "screen" [21:18] It's not a GUI thing? [21:18] tdelam: it'll setup a virtual TTY that both users can write and see at the same time [21:18] tdelam: nope [21:18] interesting, I'll try it out on my server first. [21:18] tdelam: https://help.ubuntu.com/10.04/serverguide/C/byobu.html [21:18] lynxman: right, so in this case use 11.4 with default 9.0 is better.. because I have oficial support of ubuntu for postgresql [21:18] thanks lynxman [21:18] lynxman: tdelam: even better: https://help.ubuntu.com/community/Byobu [21:18] plm: I wouldn't risk it if I were you, unless you plan to replace that server in less than 2 years [21:18] kirkland: oh thanks :) [21:20] plm: so yes you wouldn't get support for the private PPA but on the other hand you'll get support from the PostgreSQL community anyway [21:20] plm: so it's not very risque [21:20] lynxman: right, but for example, using 10.4 with PPA postgres 9.0. in 2012 are there a ubgfix of 9.0... this PPA will receive this update like as another package? [21:20] wow, byobu is awesome [21:20] so simple [21:20] tdelam: it's very nice :) [21:20] tdelam: byoby is a very fancy config for screen ;) [21:21] plm: if the community maintainer does, yes, and they're very good [21:21] ya, tested it out, logged in from work to my server then logged into some other server i have and tried it [21:21] it's exactly what I need [21:21] plm: worst case scenario you can build your own postgres packages for 9.0 [21:21] lynxman: right, I will use 10.4 with PPA :-) [21:22] plm: it's the safest choice for a prod server :) [21:22] should work if the user has putty right? [21:22] * tdelam doesn't see why not but stranger things happened with windows users [21:23] lynxman: thanks :-) [21:23] tdelam: yes, should work like a charm, you can resize the byobu window to fit the user with the smallest terminal [21:23] plm: no problem ;) [21:24] i love this and it's only 3 mins of usage === Xago_ is now known as Xago [21:27] IMO for PPAs, you always need to check out who is maintaining it ;) [21:27] and maybe contact them [21:32] JanC: good point [21:33] e.g. I know the cherokee PPA is maintained very well & very actively [21:33] with new packages 1-2 days after every official release [21:34] Martin Pitt maintains his own PostgreSQL PPA http://www.piware.de/2010/09/postgresql-9-0-final-released/ [21:34] that's the one I used [21:34] Martin Pitt is one of the main Ubuntu developers [21:34] JanC: sactly, so I reckon it'll be in good shape [21:35] so that should be okay, although a company could ask him if he plans to maintain it for the next X years of course ;) [21:35] JanC: yeah but PostgreSQL is not affiliated to any company, well maybe except 2nd Quadrant, but it's a community driven corp [21:36] well, Canonical can guarantee security updates to a PPA or backport for example [21:36] but you'll have to pay for that maybe ;) [21:36] JanC: heh [21:59] hi,, i get java error3 when opening my website... is this a port problem? [21:59] apache2 and ubuntu 8.04 [22:32] kirkland: ping [22:32] RoAkSoAx: pong [22:33] kirkland: howdy!! Have a second to discuss powernap stuff ? [22:34] ping Daviey [22:34] RoAkSoAx: sure, wassup [22:35] kirkland: so I'm working on the client/server thing already [22:36] kirkland: and first looking into auto-registration for client machines [22:36] with powerwaked [22:36] so that powerwaked will know all the hostnames/mac/ip's of the clients if they register [22:36] kirkland: now, since you guys worked on that for eucalyptus [22:37] kirkland: I was wonderingf what's the best way to go here. either 1. Embbed python code for avahi, both on client and server or 2. do as separate components with powerwaked-client-listener and powernap-client-publisher or similar [22:37] kirkland: and the publisher only uses avahi-publish rather than embbed the code in powernap's daemon [22:38] 1. will involve checking the config file for "AUTO_PUBLISH=y" if y, then publish, otherwise, do not [22:39] RoAkSoAx: hmm [22:39] kirkland: and 2. will involve publishing using just an upstart script [22:39] RoAkSoAx: you gotta be real careful here, from a security perspective [22:39] RoAkSoAx: imagine that you're running powernap on your laptop (as I am) [22:39] RoAkSoAx: and you go to UDS [22:40] RoAkSoAx: and your laptop is broadcasting that it's a powernap client, hello, i'm ready to be managed ... [22:40] RoAkSoAx: and then someone decides that Andres' machine needs to go to sleep :-) [22:42] RoAkSoAx: i think I'd prefer an approach closer to ssh-copy-id / ssh-import-id [22:42] RoAkSoAx: rather than autoregistration [22:42] RoAkSoAx: for autoregistration, you should use orchestra + puppet/ensemble [22:42] RoAkSoAx: where what you're really doing is federating clients to a server [22:42] RoAkSoAx: that shouldn't be unique to powernap, IMHO [22:43] RoAkSoAx: especially while we're working on other infrastructure in the Platform/Server to do this in a general way [22:43] SpamapS: congrads on getting ensemble into the NEW queue ;-) [22:43] kirkland: right, yeah I was planning on using avahi to do something like what was done with UEC for auto-registration features, such as: MachineA publishes itself to the network sending its IP, MAC and Hostname. Then powerwaked server obtains that info, and checks the local database if it's already there or not, if not, then just register it by adding the info to the powerwake cache [22:44] RoAkSoAx: sorry, i'm not a big fan of that part of UEC's design/implementation [22:44] RoAkSoAx: i'd hate to see it repeated for powernap :-( [22:44] kirkland: yeah I guess if using with puppet/ensemble the idea would simply be to interact with some powerwaked API to tell it "Register this machine if it is not already registered" [22:45] kirkland: yeah I'm also not really fun of broadcasting that info and would rather do a real client/server approach where the client know's who the server is, and tries to register when daemon is started [22:45] s/fun/fan [22:52] kirkland: what about the usage of rabbitmq [22:53] RoAkSoAx: what about it? [22:53] kirkland: I mean, use rabbitmq to do registration, notification to the server, stuff like that [22:54] kirkland: for i.e. if powernap-client is "managed" then whenever it changes of state, powernap-client sends messages to the server, or noties that's gonna execute and action, or has recover from an action and stuff like that [23:05] hey anybody have some expereince with dl380 g4 servers? [23:19] Eitan: what kind of experience? I used to use them quite a bit but only with centos. [23:20] kirkland: thanks. :) Hopefully an archive admin is as congratulatory as you. ;) [23:21] SpamapS: hah! [23:31] New bug: #796854 in euca2ools (main) "euca-bundle-vol fails on when running tune_cmd" [Undecided,New] https://launchpad.net/bugs/796854