[00:10] should the entries for /etc/event.d/tty1 really be 'start on stopped rc2' ? [00:21] Yes? [00:23] ok, well still, /bin/login isn't coming up [00:23] Takahito's kernel worked for me [00:24] I can ssh into a xen vm no problem [00:25] tell me something interesting you have done with a ubuntu server [00:25] telexicon: I don't know how Xen handles terminals :/ [00:26] AtomicSpark, im running an ejabberd xmpp server on one [00:26] a postgres database on another [00:26] one of them is acting as an nfs gateway to our SAN [00:26] another one is running apt-proxy to speed up updates for our workstation deployments [00:27] i just have a unused hp netserver 4 at home (duel p2 someghz 256MB pc100), just getting ideas for it. :P [00:27] one of them is running squid + dansguardian [00:27] im trying to set up one now, to test java web app frameworks [00:28] after you set up a squid proxy server, do you have to set clients to use that as their proxy server or is it automatic? [00:28] we want it to be transparent [00:29] but the cisco ASA is making that unnecessarily difficult [00:29] yes. transparent would be best. :P [00:29] cisco makes everything difficult imo [00:29] its certainly doable, i have the squid side all set up properly [00:29] but the cisco asa wont cooperate [00:32] what is the asa for? i'm reading the cisco page but it doesn't really explain much [00:32] its a firewall, nat, VPN.. thing [00:33] ah kindof an all-in-one [00:33] sure [00:33] but i dont see whats so special about it [00:34] well they like making themselves feel special. trust me. i tried to get certified. haha. [00:36] i dont like this :-/ [00:36] ssh is letting me login without a password [00:36] and sudo doesnt work and login doesnt work [00:36] i think pam is broken [00:36] * telexicon sighs [00:39] that is very interesting. i'm doing a little side research myself. [00:39] oh nevermind, doh [00:39] i forgot i put my pubkey in the image [00:39] hehe ok that one was my fault [00:45] it happens. i love when you reinstall a server and can't ssh into that machine anymore lol. [00:45] i suppose it's nice for security. no man-in-the middle attacks. [00:46] eh [00:46] but it limits usefulness [00:48] need a better way to remove the server from the .ssh file. i suppose you could just comment out the line. i usually just end up deleting the file. [00:53] sommer: on that ebox url, I was trusting the official ebox documentation, and a hint of a problem from a forum post, but someone in #ebox folks say it is https://yourserver/ebox not https://yourserver/eBox but hasn't responded with a clarification - huh.... anyone have ebox installed? I don't want to reboot to get it up and lose all my web page context.... [00:58] nealmcb: I think it's /ebox, but I don't have it installed any longer [00:59] haha. firedrill. i took my laptop with [00:59] sommer: my firefox url memory concurs. too bad that their doc doesn't.... [01:01] AtomicSpark: Remove the file?!? [01:01] yes? it clears your stored keys [01:01] AtomicSpark: I usually just remove the line manually, but ssh-keygen can do it for you, too. [01:01] AtomicSpark: Yeah. That's a bad plan. [01:02] ah. well i usually do bad things. ill try ssh-keygen next time. [01:02] There's a reason it stores them :) [01:04] true. which is why i said there should be a better way then just removing the file. [01:27] soren: thanks for the tip on ssh-keygen - that is always a pain. good procedures (e.g. when building virtual machines) to just keep the original host key would be even better [01:40] how can i keep the time on my 6.06 LTS server up to date? every few weeks the time is off. [01:41] HS-L: Install ntp. [01:41] is already installed.. :/ [01:44] you manually installed ntp? only ntpdate is installed by default. [01:44] you could try adding another time server if ubuntu.com gets off once in awhile. i never really checked after i insatlled it. [03:41] <--nub [03:41] need some help [03:42] i have setup server. with ebox. i have server running and can connect from other pc. [03:42] now i need to set some shares. [03:42] anybody have a link or command to start me off? [03:46] anybody know where i can find info on setting up shares for samba [03:46] ? [03:47] I have a wonderful system setup here for ssh access to my server, my question is how do I setup a dual boot thin client for a remote X session? [03:48] domare, I gave up on samba and now use ssh....no problems ther [03:48] and your shares are win systems? [03:49] No, I had problems between two linux systems. [03:49] sorry [03:49] ah.. yeah im trying to use this as a media server for my home pcs. [03:49] i need to get it running here before i think of trying at my biz [03:50] domare, I got your idea.....I am kind of doing the same except with all ubuntu systems [03:51] yeah, i can connect to my fileserver.. but i have no shares setup! [03:51] i also have ebox running for web administration [03:52] Yeah, I tried that route before.....to no good end [03:52] ssh was my solution [03:52] that seems a more solid way to do things with this.. i should try that [03:53] domare, I think so! I use it over the internet with no issues as well. [03:53] do u have a link to some instructions? [03:53] hmmm.....geeze. I forget now, let me review my notes [03:54] have u thought about trying VMware for you needs? [03:54] I don't use windows anymore....I just need to remotely run X windows sessions. [03:55] maybe you could load whatever it is your trying to duel boot [03:55] I finally kicked the windows habit as of yesterday [03:55] Ah, I want my thin clients to have the option to run a local version of ubuntu with ssh, or a remote X session. [03:56] Get the idea? [03:57] yeah, similar to what im thinking in the end [03:57] i would like to try some different setups in VMs also.. not sure if i can do that from server though [03:58] like a terminal server in one.. file in another [03:59] im getting the "following features not present on the CPU" after the server install [03:59] i know i need a generic kernel [03:59] network is not linking up from rescue mode, any ideas? [04:00] i think everybody here is asleep here man [04:00] ha, good stuff [04:01] domare, you could be right. [04:01] heres somebody who got VMware up.. http://www.bauer-power.net/2008/04/installing-vmware-server-on-ubuntu-804.html [04:02] <{Roger}> vmware server is sweet [04:02] under xp i have small business server and some other VMs running nicely. Im kinda trying to replicate that with better performance results [04:02] i can get it working in parallels and vmware, the laptop i'm doing this on won't connect with the network, so i need different options [04:02] yeah... fun stuff [04:02] I guess I just don't understand how vmare would help my application out [04:02] moving from the virtual environment to the real thing [04:03] oh, not me.. nevermind [04:04] u guys seen the new cisco modules that u can install linux on.. [04:04] runs on the backbone of your network [04:05] yeah, for like a million dollars [04:05] yeah.. heh.. but the network virtualization is very coo. [04:05] i just got a quote for two 48port switches with 10GbE uplinks, $25k [04:06] damn [04:06] probably the cx4 modules that threw it over the top [04:12] this is a good read for some 101 network virtualization info [04:12] http://www.petri.co.il/server-virtualization-network-virtualization-storage-virtualization.htm [04:13] i guess 3com already has some modules that run a linux system.. u can set sniffers and whatnot right on backbone [07:49] nealmcb: "the original host key"? [08:07] i connect to my ubuntu server through VPN since connection is established i loose route to internet to everyhting, anybody can help in soloving such prob [08:15] I'm having some trouble with 'automiunt' during reboot. I have checked my /dev/fstab file, but it all looks fine (except from all the encrypted shit concerning the /boot partition etc. ). Should I look anywhere else ?! [08:18] torben: You could perhaps explain what your problem is.. [08:18] You fix "some trouble" by "changing some stuff". [08:19] If the problem description becomes more clear, the solution can too. [08:20] soren : *Cough*, well... It seems that wen I reboot my box it automounts two partitions 'on top' of each other. Very annoying !?... [08:21] ... It does run 24/7 as a server, and I can offcourse manually unmount the stuff etc., but it would be nice if I could 'just' reboot and have a working system :D [08:21] When you say "automounts"... Are you talking about nfs automounting or just regular mounting done at boot? [08:22] Hrm... Good question !?. [08:22] If you don't know, then it's not about the nfs things. [08:22] ... I'll just check dmesg... [08:22] You don't accidentally set up nfs automounting. Really. [08:23] What do you mean by "'on top' of each other"? Can you put your fstab somewhere? [08:23] !pastebin [08:23] pastebin is a service to post multiple-lined texts so you don't flood the channel. The Ubuntu pastebin is at http://paste.ubuntu.com (make sure you give us the URL for your paste - see also the channel topic) [08:26] soren, Well, thanx a bunch... I am bit of a noob on IRC (as you probably guessed)... The link is : http://paste.ubuntu.com/9413/ [08:27] Hm... nfs... ?! [08:28] And what seems to be the problem, exactly? [08:28] ...and you really should be using UUID's for the last filesystems, too. [08:30] Well, the problem is that hda1 is mounted in /home/apach2/domains, and not /home /ftp/data/webservices... [08:31] torben: How is that "'on top' of each other"? [08:31] Heh, my To-do list regarding Linux, and my server is really full ;o) [08:34] when i connect to my server through VPn i loose connection can anybody guide me [08:34] hmmm [08:37] soren, Well, I run Ubuntu server from a USB-stick [08:37] Ok.. [08:38] ... On the stick I have a directory for Apache, containing the simple (and small) parts of a website [08:40] ... I also have a IDE-HD containing some ftp stuff etc, and experimental websites... When I boot the server mounts my 'ftp-directory' (on the IDE drive) on top of my 'Apache-directory' (on the stick) [08:40] That makes no sense. Your fstab doesn't even mention the apache directory. [08:40] You say you can fix this by running some command.. Which one is that? [08:40] soren, that's what puzzles me [08:41] 'umount /home/apache2/domains' [08:42] offcourse that removes my 'ftp-directory' from the mountpoint at my stick.... [08:42] ... Revealing the underlying 'domains'-directory'... [08:43] ... It's not a real 'big' problem... But had to reboot a couple of times yesterday, and it began to bother me alot :) [08:47] <_ruben> i'd check if thre's a boot script that does the mounting .. grep mount /etc/init.d/* or so :) [08:48] torben: So your ftp directory is mounted twice, then? Once on /home/apache2/domains and once somewhere else? [08:52] _ruben, Hm... Actually U might be ritht there... [08:53] soren : Exactly :) [08:55] _ruben, well there it was :o)... ThanX a bunch [08:56] I've just logged into a VPS with Ubuntu for the first time. For starters it's running with a root account. How do I find out how else this "virgin install" differs from a standard ubuntu-server installation? [08:57] Other non-standard things are that it's running ReiserFS, has wifi tools installed etc. [08:58] <_ruben> wifi is standard i think [08:58] <_ruben> torben: glad to have helped :) [08:58] _ruben: Not on the hardy install on my local vmware instance. [08:59] hello _ruben [08:59] This VPS also has x11-common marked as 'rc' and comes with things like make etc. [09:00] <_ruben> owh: well, atleast the wpa_* tools are standard [09:01] <_ruben> owh: they probably tried to make things easy for their customers by providing extra stuff by default [09:01] So, back to the original question, how do I actually compare it with a ubuntu-server install, or can I "install" ubuntu-server and remove all the rest in some sane way? [09:02] <_ruben> take "dpkg -l" from the vps and from a clean install and run diff over it .. probably requires some extra scripting to "ignore" the versions [09:03] <_ruben> dont think there's a "remove all but the core" option :) [09:03] _ruben: Yeah, but that won't catch any "manual" editing. [09:03] <_ruben> as that would kill ssh and stuff as well [09:04] <_ruben> owh: to cactch manual editing you'd first have to get a reference system with the exact same versions installed .. and then diff/md5sum each file to see if it has been changed .. not that easy ;) [09:04] _ruben: Wonderful. Just what I didn't need today. [09:05] <_ruben> owh: depending on the disk layout .. you *could* debootstrap a clean install on it :p [09:05] _ruben: Yeah, but that really feels like a scary option :) [09:06] owh: That's what I've done with all my colocated servers. [09:06] soren: The last time I did a debootstrap it ended in tears. [09:07] I suppose the out I have here is that I can just reinstall another OS over the top if I'm lucky. I suppose I could try that before I really break it :) [09:14] Is there a way, like lsb_release, that would show me which installation was completed? [09:15] As in, Ubuntu Server, Desktop, Alternate, JEOS, etc? [09:20] <_ruben> there might be some hints in /var/log/installer [09:20] Bloody good idea, tah. [09:21] _ruben: Crap, no longer exists. [09:21] Hi guys, maybe this is a silly question, but here it is anyways. Can a (d)dos attack cause a server to reboot? [09:21] _ruben: And dpkg.loh is empty, yay :| [09:22] omnz0r: Well, it is possible. [09:22] owh: Do you know of any "documentation" or examples of this, that is accessible via web? [09:23] <_ruben> dos is denial of service .. so getting a box to crash sure qualifies as being a dos [09:24] omnz0r: Not specifically no, but a ddos could cause all manner of issues, caches filling, drives filling, memory filling, network services crashing, each of those could cause a reboot in itself. [09:24] _ruben: sure :) The question was not what a dos attack is, but more what a dos attack can cause :) [09:25] owh: thank you very much, it seems intuitive when you put it that way :) [09:28] In this VPS I need to receive small emails that are sent to a single account where a .forward file pipes the email to a script. What email package is likely best suited - probably in memory footprint more than anything - to achieve this? [09:29] <_ruben> owh: i think i'd just go for the default: postfix [09:30] <_ruben> then again, smtp is rather simple, could write your own daemon for it ;) [09:30] _ruben: I'm canny, but not stupid :) [09:32] <_ruben> hehe [09:32] _ruben: Next you'll be suggesting that I get out the soldering iron and the magnifying glasses to repair any broken memory cores :) [09:33] you'll need a spare core for that ;) [09:34] ewook: You raise a valid point, spares for that would be hard to come by :) [09:36] ewook: Mind you, finding the diagrams and the actual board would be probably just as complicated >:-) [09:38] owh: true enough =o). [09:39] <_ruben> hehe [09:43] Does apt support wild-cards, like apt-get --purge remove *mysql* ? [09:45] In case anyone else was wondering, not as far as I can tell. [09:48] Workaround: apt-get --purge remove mysql-common [09:58] mysql I am really interested in learning that owh! [10:01] <\sh> owh, apt should support regexp...apt-get install ^claws-mail e.g. works [10:11] good morning [10:13] <_ruben> g'day [10:13] hi _ruben [10:13] quick question: [10:13] i was wondering if there was an easy way to swap from the desktop version of ubuntu to the server version [10:14] <_ruben> apt-get remove ubuntu-desktop ; apt-get install ubuntu-server .. might do the trick .. but doing a clean install sure is prefered [10:14] <_ruben> what's the idea behind to move from dekstop to server? [10:15] i've been playing around with dekiwiki on my desktop ubuntu - and will shortly have a 64bit server free to move into production [10:16] i just wanted to have a look at the server version while waiting for the server to arrive [10:17] apt-get install ubuntu-server, and shit+alt+f1 to switch to a console [10:19] thanks Deeps [10:20] <_ruben> Payo: i usualy use virtualization software (vmware/xen/kvm/vbox/etc) for that [10:21] <_ruben> for testriding software/os's that is [10:21] yeah - i should get to grips with virtualisation === c1|freaky_ is now known as c1|freaky [10:22] which of those packages are open-source [10:22] <_ruben> xen and kvm are afaik, vmware and vbox arent, but are free (depending on which exact product) [10:23] <_ruben> i prefer vmware myself [10:27] hi all. is someone familiar with openvpn? i own a dedicated server. it only has one ethernet card. can i still use ethernet bridging? the howto says i shouldnt use it with a normal ethernet card which is connected to the internet but we want to be able to play lan games and use windows filesharing and that's not really possible without the ethernet bridging as it wont send any broadcasts and stuff without ethernet bridging. [10:29] mindtouch offer a vmware image of dekiwiki - time to start learning! [10:31] <_ruben> c1|freaky: i am familiar with openvpn, but never tried it on a single homed machine [10:31] <_ruben> or actually, i think i did [10:32] it has 6 IP Addresses, - 6 virtual interfacces on eth0 but only one NIC [10:32] (the server) [10:32] clients should only connect to it and see each other [10:32] there is no LAN behind the server only the vpn network [10:33] <_ruben> c1|freaky: i think that should be doable, its unlike any setup i've done though [10:34] _ruben: i have currently set up the server using tap but still using the normal server config directive and not the ethernet-bridging stuff [10:34] it's working [10:34] but it says i must use ethernet bridging not the server directive [10:34] if i do that it wont work anymore [10:35] so now i have no idea if i can put that bridge up on the interface or if i should because the howto says i'd open a security hole [10:35] and in #openvpn noone answers [10:36] <_ruben> the mentioned security hole is probably that the clients can reach eachother, which is exactly what you're trying to achieve [10:37] ok thank you [10:38] :D [10:40] <_ruben> wonder if i shoudl create a ppa for my open-vm-tools packages for hardy/gutsy [10:42] umm, if i just put a ethernet bridge on that interface, will it mean other clients from the internet can send netbios packages? [10:43] and wont it take the interface down? oO [10:43] _ruben ;D [10:43] <_ruben> wouldnt dare to say, never worked with bridgin (other than with vmware) .. and a decent firewall should keep "the internet" out of your bridge [10:44] <_ruben> i dont think you need to create a bridge though [10:44] <_ruben> since you only want inter-client connectivity [10:44] <_ruben> which is up to openvpn to take care of [10:44] Make sure to use an interface which is private and which is connected to a LAN which is protected from the internet by a firewall. <-- this is waht the howto says, nothing applies to that ethernet card [10:45] ok and what's the bridge for then? [10:46] i really need the tunnelling of bridging but it really works allready [10:46] just unsure what the bridge then does if it's usually required [10:48] This example will guide you in configuring an OpenVPN server-side ethernet bridge. Multiple clients will be able to connect to the bridge, and each client's TAP interface will be assigned an IP address that is part of the server's LAN. <-- maybe only for this then [10:48] but as i have no lan behind the server only through openvpn [10:48] i think it's ok to leave it like that [10:48] i hope [10:49] <_ruben> c1|freaky: i *think* so too, but dont know for sure, try and find out ;) [10:50] what should i try to find out? i've allready played lan games, used windows filesharing etc. [10:50] well I finally found information on how to set up a gui for ubuntu server edition [10:50] so that's working [10:50] :) [10:51] :) [10:52] how do i set up samba to not require user authentication - ill search google [10:52] <_ruben> c1|freaky: if it was already working, then what was the problem? :P [10:53] <_ruben> gui on server .. yuck [10:53] _ruben: the problem was, im using ethernet bridging, so i wondered why it works without the ethernet bridge. ive setup without the bridge before. i thought maybe it was using that interface in a completely wrong way [10:53] <_ruben> ah [10:57] New bug: #225608 in postfix (main) "Upgrade to 8.04" [Undecided,New] https://launchpad.net/bugs/225608 [10:57] <_ruben> nice bug description [10:58] hehe [11:23] well I like using a gui but also as well just the command line to comes in quite handy too there _ruben [11:26] <_ruben> gui's are a waste of resources on a server imo [11:27] <_ruben> not to mention the security implications [11:27] true [11:29] resources is what alot of people need these days esspecally when it comes to demanding applications [11:30] there are always thouse ups and downs to things [11:31] I think I well have a better home for using ubuntu then microsoft there _ruben [11:32] :) [12:02] hello all [12:03] iam searching a time-daemon, which willbe run as daemon, and willbe check and control time all view minuts with my own timeserver. which program should i use for that? [12:05] <_ruben> ctx144k: not really sure what you mean, but i think the answer is ntpd? [12:06] _ruben: apport is _so_ helpful sometimes, isn't it? [12:08] _ruben, ill look for it, thx [12:08] <_ruben> huh? .. ow that bugreport i guess, didnt look at it, so wouldnt know the 'source' either :) [12:09] <_ruben> lamont: ^^ [12:12] _ruben: yeah 225608 is a "upgrade failed, kthx" report. of course, I haven't managed to duplicate it in my gutsy->hardy transitions, so the lack of information, mixed with the non-english error message makes it, uh, a conversation starter at best [12:15] <_ruben> lamont: heh, indeed [12:50] hello everybody =) now I successfully installed ubuntuserver because nothing else is working. but how do I get into X and xfce or something else? or is it only CLI? [12:51] ubuntu server doesn't feature a desktop component [12:51] a GUI component, even [12:51] apt-get install ubuntu-dektop if you want the standard ubuntu desktop environment [12:51] ubuntu-desktop even [12:51] ok Ill try that thanx [12:55] everyting I try with sudo apt-get is giving me "the package could not be found" [12:55] maybe I have no connection...? [13:02] <_ruben> try sudo apt-get update [13:05] I did ifconfig and there is only info about lo nothing about eth0 [13:06] <_ruben> and what does "ip addr" show? ifconfig is ancient [13:14] I have no ipaddress but now eth0 is shown in ifconfig [13:18] morning [13:21] New bug: #225661 in samba (main) "Typo in /etc/samba/smb.conf (nsuccessful)" [Undecided,New] https://launchpad.net/bugs/225661 [13:26] dendrobates: morning [14:01] soren: "original host key" - i.e. the problem is that people reinstall an os, use the same name for the machine, but don't save the old /etc/ssh/ssh_host_dsa_key and/or /etc/ssh/ssh_host_rsa_key and reuse that on the new os. maybe ubuntu should make it easier to preserve these across reinstall. [14:01] lucent home? [14:05] <\sh> zul, ping bacula :) [14:06] <\sh> zul, could you add the drop-mysql-tables script from the bacula source to bacula-director-mysql ? it's missing somehow, but it would make sense, when we ship create-tables-mysql script from the very same dir [14:06] \sh: sure [14:06] patches also welcomed :) [14:06] <\sh> zul, I'm just on my way to a working build infra at home again ;) [14:07] heh [14:08] <\sh> zul, another thing is, I'll try to fix the qt4 bacula console, too...it has some issues when editing the media files (e.g. it's not possible to set a file to max 100GB size, the input line has wrong max values somehow) [14:08] \sh: thats weird can you open issues in launchpad so we can track them for a possible SRU as well [14:09] <\sh> zul, yes..I'm working on a fix, as said, I stepped upon this bug last week when I setup bacula for the company ;) [14:09] ok [14:09] <\sh> zul, manually editing the volmaxsize inside the db tables works :) but not from the qt4 console [14:10] interesting we will have to talk to upstream as well [14:11] <\sh> zul, for sure...when I have a patch ready for applying I'll push all the stuff towards debian+upstream [14:11] thanks [14:28] mathiaz: Ive added the dovecot bug to the stablereleasetracker wiki page [14:31] zul: great - thanks [14:39] hmm - I hadn't seen this before: https://help.ubuntu.com/community/WebMin seems like some information on why it was dropped based on https://answers.edge.launchpad.net/ubuntu/+question/2873 would help, along with an ebox pointer [14:39] !ebox [14:39] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/eBox [14:39] !webmin [14:39] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. See !ebox instead. [14:41] * nealmcb starts taking a stab at it.... [15:03] comments welcomed on https://help.ubuntu.com/community/WebMin and/or my post just now to the list, with more specifics on the config file issues requested [15:07] if I do a data dvd with gnomebaker, will that burn image files as an image? or just as data? [15:09] * nealmcb also notes WebminWithoutARootAccount [15:09] https://help.ubuntu.com/community/WebminWithoutARootAccount [15:12] nealmcb: are you hunting down all the webmin occurence on w.u.c ? [15:12] nealmcb: and help.u.c [15:12] nealmcb: ? [15:13] slowly.... [15:13] * mathiaz hugs nealmcb [15:14] :-) [15:15] but it would really help to have something more substantial to point at when folks talk of incompatibility with policy - I hate just passing on pretty vague complaints [15:17] nealmcb: agreed [15:17] nealmcb: you could do search on irclogs.ubuntu.com about webmin and #ubuntu-server [15:18] nealmcb: sometime ago there was a discussion about it and IIRC infinity gave some reasons about the configuration files issues [15:18] mathiaz: that's a painful prospect.... much heat, little light [15:18] nealmcb: I agree with you - we should really come up with a good answer on this one [15:20] if we collectively spent as much time actually improving ebox (or even fixing webmin) that we do complaining about webmin, we'd be much better off :-) and have fewer complaints.... [15:21] * mathiaz nods [15:23] nealmcb: eBox should be getting better since they are basing their stuff off of hardy starting their next development cycle [15:25] Hello I am trying to use IP based hosting on apache2 but my DocumentRoot is not being used within the configuration file within sites-available, it is using the default path from the default site. Can anyonehelp me with this? [15:26] excuse me, I meant DocumentRoot [15:27] <_ruben> eghjaytee: did you enable the site (symlink in sites-enabled) [15:27] _ruben: yes I did [15:27] this is so weird [15:27] this is for a second apache instance [15:27] I copied over the entire /etc/apache2 dir and changed the IPs [15:27] <_ruben> 2 instances on same host? [15:27] no different hosts [15:27] <_ruben> ah ok [15:27] load balanced [15:27] nealmcb: it seems that the only reason why webmin was dropped was the lack of maintainers [15:28] _ruben: can you thinkn of a reason why DocumentRoot would be picked up from within a VirtualHost conf file? [15:29] <_ruben> eghjaytee: apparently apache thinks the default is the best match for the request .. no errors or wahtever in the logs? (during both restart/reload and request of the url) [15:30] nealmcb: webmin upstream provide packages for debian/ubuntu - http://www.webmin.com/deb.html [15:30] _ruben: no it is so weird, it is like the VirtualHost isn't being picked up [15:30] mathiaz: for debian that seems the reason. for ubuntu it seems that some admins hated it with a passion. but the concrete reasons for hating it are hard to find. [15:30] and so it uses the default document root [15:31] I will up the logging [15:31] nealmcb: right - we'll wait for infinity to wake up and we'll know why ;) [15:31] mathiaz: yeah - that is the point of that wiki page on webmin. I'd love for someone who knows debian policy to review the current webmin and point out some issues (or maybe they fixed them) [15:31] oh dear god, help! [15:32] I have a presentation in an hour or so, [15:32] nealmcb: I'm not totally convinced by "webmin munges your config files" [15:32] and my linux server which was working fine when I left last night, [15:32] it is painful going thru the complaints here.... http://www.google.com/search?num=50&hl=en&safe=off&client=firefox-a&rls=com.ubuntu%3Aen-US%3Aofficial&hs=dUb&q=site%3Airclogs.ubuntu.com++webmin++%23ubuntu-server+infinity&btnG=Search [15:32] I come in and am able to connect to it through SSH but yet it cannot get any outside access. [15:32] <_ruben> eghjaytee: try: apache2 -S [15:32] So then I turn around and reboot the box, [15:32] and now it can't even see it's eth0 or IP! [15:32] nealmcb: that's the role of webmin - editing configuration files [15:33] nealmcb: I fail to see how ebox and webmin are different on this point [15:33] nealmcb: expect that ebox takes care of not overwritting localy modified files [15:33] i can't even use webmin on mine as I cannot connect to it! [15:33] i have made a raid 5 of 3 disks and will soon expand to 6 (sata 3G), how can i tell if all disk working in 3GB or 1.5 GB i.e. Sata I or II [15:33] the only specific I got was from ScottK:... There was a funny bit today about an Ubuntu user writing in with a Postfix problem that turned out to be webmin adding 'sudo' to their smtpd_recipient_restrictions. - http://archives.neohapsis.com/archives/postfix/2007-11/1330.html [15:34] heh [15:34] but I don't know what that is about - lamont? [15:35] I'm not sure why it added that, but fundamentally it corrupted the config file. [15:35] New bug: #225741 in mysql-dfsg-5.0 (main) "/usr/bin/mysql_config --libs_r reports incorrect link flags" [Undecided,New] https://launchpad.net/bugs/225741 [15:36] ScottK: can you clarify the implications of that a bit more? [15:36] _ruben: hey man, I thought theat a2ensite created a link [15:36] this is creating a file [15:36] so it seems that webmin adds options when it modifies a configuration file [15:36] man that is messed up [15:36] nealmcb: Use webmin and eventually you will cry is my summary. [15:37] nealmcb: you'd edit a service, and the generated configuration file ends up with non-viable options [15:37] I've never used it, so I have no idea exactly how that happened, but it's consistent with the don't use it advice I've seen given. [15:37] jesus that was the problem [15:37] ScottK: right - that's what we've always heard - [15:37] <_ruben> eghjaytee: my understanding is that a2ensite would create symlink as well.. [15:37] thanks _ruben [15:37] I dont knwo why it isnt linking [15:37] I will look into it [15:37] <_ruben> eghjaytee: heh, no problem :) [15:37] ScottK: I'd like to know what are the problems exactly - if this hasn't turned into an urban legend [15:38] i.e. postfix won't start up? or won't work? remember we're trying to clarify this for the newbies, not hardened postfix wizards [15:38] mathiaz: In that particular case it corrupted the postfix config file and broke Postfix. I've no idea how or why. [15:39] "DO NOT USE WEBMIN" there.. that's simple. :0) [15:39] I think webmin ensnares baby penguins and eats them also [15:39] If the postfix module in webmin generates a incorrect postfix configuration file, the postfix module in webmin should be fixed [15:39] mathiaz: sure. [15:39] but that's not a postfix bug [15:39] mathiaz: Sure. Have fun with that. [15:39] and it makes all the people who can help walk away from you [15:39] ScottK: lamont: do you think ebox will be better? [15:40] nfc [15:40] nealmcb: Same with me. At least ebox are working with a Debian environment in mind. Odds are better, but it's a hard problem. [15:41] <[LMJ]> hi [15:41] lamont: nfc? [15:42] hi [LMJ]. [15:42] ScottK: have fun with that -> do you suggest that the webmin code is of bad quality or hard to work with ? [15:43] mathiaz: I'm suggesting it's not a problem I'm interested in solving. I think it's better to invest time in documentation and training to make it easier for people to properly administer their systems. [15:44] <[LMJ]> I'm really stuck right now : I can't find a way to see my 2 network card on my fresh Hardy server install : 1 eth is nvidia gigabit (forcedeth) onboard who works fine, the other one, an old 100Mbits D-link one is not seen by the kernel, nor during the boot or via lspci. I through the card was day, i've tried another one who I know it works, same issue, I can get this second eth card working please ? [15:44] <[LMJ]> hi ScottK [15:49] lamont: Any suggestions on what to ask for (in terms of more information) in Bug 225608? [15:49] Launchpad bug 225608 in postfix "Upgrade to 8.04" [Undecided,New] https://launchpad.net/bugs/225608 [15:49] ScottK: I encourage you to work on what you like, and that is the sort of thing I like too. But I think documentation actually has limited impact on solving bug #1. Microsoft didn't get where it is by having good documentation - they leveraged people's preference for guis. I think it would be nice if we could expect that a significant fraction of the folks we want to be adopters of ubuntu server to read documentation and become wise in the [15:49] Launchpad bug 1 in ubuntu "Microsoft has a majority market share" [Critical,Confirmed] https://launchpad.net/bugs/1 [15:50] nealmcb: nfc is short for "no clue" [15:50] nealmcb: the problem with webmin may be that it gives to much power to the targeted audience [15:50] lamont: Ahh :) :) [15:50] ScottK: for starters "exactly how did it fail" :-) [15:50] mathiaz: good point [15:51] lamont: Yeah. I was hoping you might have some suspects I could point them at. [15:51] nealmcb: if we assume that the target audience for a web interface is up to junior sysadmin, they should not be able to edit secondary dns zone in bind9 for example. [15:51] lamont: nfc about what "smtpd_recipient_restrictions" are or what "sudo" would mean there? [15:52] ScottK: well, a translation of the error message to something I know would be helpful [15:52] lamont: OK. [15:52] nealmcb: about whether ebox would work better than webmin [15:52] as in "what's ebox?" [15:52] nealmcb: If we replicate Windows sysadmin culture in Ubuntu, we won't have accomplished much. [15:52] ScottK: agreed. [15:52] nealmcb: while knowing enough of an answer to know that, like webmin, it assumes a level of clue on the part of the user that is not generally present in the target audience [15:53] nealmcb: so we could say that standard tasks should be available through a web interface [15:54] lamont: ok. anyone care to help me flesh out that one reported webmin error I've heard about? Is it a random thing to do, or does adding sudo make it insecure, or is it a syntax error? [15:54] Any ace translators recognize what language "paketti postfix on jo asennettu ja konfiguroitu" is in? [15:54] nealmcb: Syntax error. Postfix will die, restart, and die without delivering mail (IIRC). [15:55] The issue isn't that it's sudo, but that it's some unrecognized restriction. [15:55] ScottK: finish maybe? [15:55] If it had been 'bob' it would fail the same way. [15:55] nealmcb: for this particular bug, the webmin module generated a wrong configuration file [15:56] nealmcb: I don't think it's a good example why webmin is bad - ebox could lead to the same error, or a any other document [15:56] <_ruben> indonesian and the likes are an option as well i think :p [15:56] ScottK: google concurs with zul, Finnish [15:56] ScottK: thanks! ouch!! [15:56] seriously, I need help. I just went through a rescue mode boot and could get it working, [15:56] mathiaz: right [15:56] but when finished and rebooted it stops functioning again. [15:56] kirkland: Thanks. [15:56] When I ifconfig I am getting hex, NOip [15:56] if I check the interfaces file, it shows everythign fine. [15:57] lamont: That's the reporters first Ubuntu bug report ever, so I don't hold out a lot of hope. [15:57] I cannot get an IP through DHCP so I must set it by hand which it works fine. [15:57] however after reboot, nothing. [15:57] Anyone, for the love of god. [15:57] <_ruben> CannibalM: probably an error in the interfaces file .. no errors on sudo invoke-rc.d networking restart ? [15:57] ScottK: atm, it's +moreinfo :-( [15:57] dear oo.o why do you take focus? [15:57] I didn't put my mouse in your window [15:58] _ruben, the interface file is mint, even checked it against three other systems. no issues, the networking restart never gets an IP, just loops forever. Can't even break it. [15:59] <_ruben> loops ? put the output on a pastebin [16:01] can't copy and paste. [16:01] lol no net [16:01] lamont: The error message translates at "the package postfix is already installed and configured". [16:02] <_ruben> CannibalM: then describe in what way it's looping [16:02] zul and kirkland: Thanks. Finnish it is. [16:03] lamont: Does that mean it's update-manager and we shift the blame to mvo? [16:04] hello. how's hardy doing in server deployment? anybody has a feedback? [16:05] How can i tell if disks working in 3GB or 1.5 GB i.e. Sata I or II [16:05] New bug: #225600 in php5 (main) "[Hardy][Regression] PHP 5.2.4 symlink bug breaks TYPO3 default setup" [Medium,Confirmed] https://launchpad.net/bugs/225600 [16:09] <_ruben> Asad2005: check output of dmesg .. there should smth like: [ 19.298661] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300) [16:09] <_ruben> might depend on driver tho [16:11] _ruben, and if it is showing 1.5 while drive and controller is 3 GB is there away to alter it from linux or bios [16:12] <_ruben> Asad2005: i know certain maxtor drives have a jumper for sata1/sata2 mode [16:12] <_ruben> other than that i dunno [16:12] Server deployment = the suxor. [16:13] anger comes from it when networking randomly works [16:13] grrrrrr [16:13] * _ruben uses his psychic powers to see which errors CannibalM is getting [16:14] lol [16:14] in short, DHCP never can pull an IP while everything else can, this is on most hardware we have tried and trialed the software on. [16:14] so setting by static it works, [16:15] worked beautifully, came in this morning and went to update repos and download a tarball onto the system through webmin, [16:15] no access to the net. [16:15] the internal network, no problem. [16:15] wtf? So I check the interfaces file, check ifconfig etc etc. [16:15] all looks fine. [16:15] I reboot, and now no network found [16:15] ScottK: heh [16:15] Ihave to use rescue mode and hand configure the info, [16:16] (Which matches the interfaces file EXACTLY) [16:16] lamont: I don't think Postfix can make that error. [16:16] ScottK: ok. so what exactly did it didn't do? [16:16] * nealmcb doesn't need psychic powers to guess what the reaction here will be to the mention of webmin and random networking problems.... [16:16] Dunno. [16:16] and then start all the services through the shell loaded through the rescue boot. [16:16] <_ruben> i have no idea what you mean by : DHCP never can pull an IP while everything else can, this is on most hardware we have tried and trialed the software on. [16:16] Webmin has been great for me. [16:16] Well, we use a DHCP network here for most items, [16:17] from there we set all static IP's on the reserved IP's for servers etc. [16:17] well, with Ubuntu, it can never get an IP [16:17] ever. [16:17] I use DHCP almost exclusively on the network I'm sitting on, with no issues. [16:17] I have to always set it by hand. [16:17] but only since warty [16:17] <_ruben> dhcp works like a charm here as well [16:17] Yeah, its really wierd because I have used a couple different ubuntu releases, [16:17] on other machines with no issues at all. [16:17] but for the last two, [16:17] even desktop machines can't get it. [16:18] CannibalM: I've been using ubuntu with dhcp since before it released. [16:18] both upgraded and freshly installed machines for feisty/gutsy/hardy, and probably earlier [16:18] CannibalM, i've been having issues with a hardy system and dhcp too [16:19] its wierd, just the last two releases. Our servers, ranging from old DL350's up through new dual quad cores. [16:19] I'm starting to think its our DHCP server. [16:19] but everything else works mint so I'm boggled. [16:19] * Kamping_Kaiser blames nm, just because [16:19] Kamping_Kaiser, I took your advice before not to do with hardy yet :) [16:20] and at the same time, the desktop's for dev work with the desktop edition, [16:20] same thing. [16:20] Well, through rescue boot I got apache and everything else running with the network functioning again, so I'm gonna do a drop of all my files onto something else, [16:20] rhineheart_m, :) hopefully its going to be helpful advice [16:20] then do a full wipe and see if the last LTS server can handle the task of holding a network interface and IP to function properly. [16:21] it was 8.04-server that wasnt working. my 8.04 laptop is fine [16:21] Oh, [16:21] * Kamping_Kaiser has had issues with 8.04 on both server installs hes done [16:21] and any other early adopters download the desktop version of 8 and have it not be a bootable disk? [16:21] Kamping_Kaiser, yeah.. I just play around with debian etch.. to tell you honestly.. ubuntu helped me a lot to learn the debian way :) [16:21] yea, I was thinking of Debian, not sure what would be the best distro for a server setup [16:21] I would hate to move to RH [16:22] * ScottK has done one fresh install of 8.04 and several upgrades without problems. [16:22] althought we already have an enterprise boc going. [16:22] *box [16:22] ScottK, oh good, it works for someone then ;) [16:22] I suspect it works for almost everybody. [16:22] great OS, amazing distro, just small things that happen. [16:23] me to. but for once i'm the one with problems, so i get to be the whiny one [16:23] mmm.. I'm running virtualmin on a fresh debian box... and so far it works like a charm! I love it! :) [16:23] I just can't have somehting go wrong at random times, [16:23] its a test box and I can't be ready to have a presentation and have something randomly go out on me. [16:24] ScottK, is network manager installed on server, or just desktop? [16:24] Just desktop. [16:24] I'm pretty sure. [16:24] All my servers are on static IP, so I don't use dhcp on servers. [16:24] thats what i thought. [16:25] i've tried one dhcp system (because its on a desk next to me if stuff breaks), and one static. i should really get the hdd from the "doent work" install and file bugs [16:25] can you then say that the release of hardy is premature? [16:26] rhineheart_m: Anyone who wants to can say that. I think they'd be wrong, but they can say it. [16:26] !ldap [16:26] LDAP is the Lightweight Directory Access Protocol. For more information and installation instructions, see https://help.ubuntu.com/community/OpenLDAPServer [16:26] * Kamping_Kaiser says it (wrongly :P) [16:33] someone want to comment on this *cough* old page and how it relates to the latest mail advice? https://help.ubuntu.com/community/MailScanner [16:34] nealmcb: I suggest just delete the page. [16:35] MailScanner is defective by design when integrated with Postfix. [16:41] ScottK: what part of the design are you referring to? [16:42] The part where it integrates with postfix via unpublished postfix internal interfaces (queue file format) and so each time postfix makes internal changes, it has to be updated, but will lose mail in the mean time. [16:42] ScottK: Total H9 [16:42] Integrating via public documented interfaces is a pretty fundamental software engineering priciniple. [16:42] great- for now I'll just say that - I think page deletion should get some wiki review [16:43] nealmcb: To go with it, the Postfix upstream strongly recommends against using Mailscanner with Postfix. It's not just us. [16:43] "dear crackheads. die. kthx" [16:44] ok, I'll admit that Wietse would never say it in quite those words. [16:45] nealmcb: If you aren't frightened enough, the MailScanner developers describe this as a "Political" problem. [16:46] nealmcb: If they believe that, I'm not using their software. [16:46] * nealmcb nods [16:47] ScottK: rotfl [16:56] leonel: You can stop working on Dapper clamav debdiffs. pitti just copied clamav 0.92.1 from dapper-backports to dapper-updates. [16:56] jdstrand: ^^^ Thanks for looking into it. [16:57] \o/ [16:57] ScottK: thanks for doing it :) [16:58] ScottK: stopped since wednesday when we talked about it yesterday I was out all day and today afternoon I resume the gutsy / feisty patches [16:58] Great. [17:12] jdstrand or kees: Now that pitti has copied clamav from dapper-backports to dapper-updates, would one of you do dapper-security (there are security fixes in there)? [17:12] I'd hate to see someone who has security, but not updates not get the security fixes. [17:12] ScottK, kees: I'll do it [17:12] jdstrand: Thanks. [17:15] jdstrand: One lingering clamav concern I have is that you can't currently build clamav on hppa for Gutsy. Apparently sendmail existed for a while and then vanished, so clamav will always FTBFS on hppa in Gutsy. Not sure what the best thing to do is to deal with the security issues in that package. [17:19] ScottK: ok-- I'll talk to the archive admin and get some more info [17:19] jdstrand: Thanks. [17:21] missing hppa shouldn't block us from updating the other archs. [17:22] Agreed, but the lack of libmilter-dev in hppa means you can't build a fixed version for hppa. [17:22] Personally I'd suggest removing all the clamav binaries for hppa in gutsy, but that's just me. [17:23] Actually, now that I think more about it ... [17:24] kees: You still build security updates on a separate dak instance, right? [17:24] ScottK: unfortunately [17:24] kees: Would you check and see if perhaps dak magically still has sendmail for gutsy/hppa? [17:25] We've done gutsy security updates before and IIRC they built for hppa. [17:26] kees: Nevermind. The security updates FTBFS too. [17:27] * ScottK sticks with the remove it it can't be supported vote. [17:56] Does anyone have a Directadmin lifetime licence for sale? [18:45] ok I don't know if to put this in #ubuntu or here. But I am doing an installfest and I have to have a server as a router. [18:45] New bug: #225818 in php5 (main) "PHP/5.2.4-2ubuntu5 + xdebug" [Undecided,New] https://launchpad.net/bugs/225818 [18:47] DBAmethyst: Server setup questions are on topic here. [18:52] ok [18:53] well just asking for help. need some dns and some dhcp help [18:53] as I have never done this to a box. normally use a wrt54gl [18:56] ScottK: I played around with clamtk early this morning, and was able to build it after setting (>=0.93) in the deps === bmckee is now known as furicle === furicle is now known as bmckee [18:57] Great. sommer: See the mail I just sent to the motu ML. === bmckee is now known as furicle [18:58] yep, I wanted to double check that is the correct way to update the package? [18:58] is it possible to run a ssh -X session but have the sound play locally? [18:59] sommer: Did you add the PPA to your sources.list so it would have clamav 0.93 to build against? [18:59] ScottK: yeppers, even built and scanned a virus successfully [19:00] Yes. Then that'll do it. You didn't actually need to bump the depends. Just build it and run it with 0.93. [19:00] heh, gotcha should I change it back? [19:01] gregbrady, read up about pulseaudio TCP forwarding ... its pretty easy to set up [19:01] No need. What that says is that we don't need to do anything special with clamtk. [19:01] sommer: Just update the wiki. [19:01] ogra, thanks. [19:01] cool, will do === `6og is now known as Kamping_Kaiser [19:05] sommer: FYI, you can upload stuff to the PPA since you are a team member. [19:05] ScottK: ya, I was wondering about that... I upload the .changes file correct? [19:06] Yes. [19:06] cool, just need to configure dput :) [19:06] I don't think uploading clamtk is needed, but if you get a package that needs changing, feel free. [19:07] hmmm, I can't seem to find anything about forwarding on the pulseaudio website [19:08] * delcoyote hi [19:11] gregbrady, google for puseaudio and module-native-protocol-tcp or module-esound-protocol-tcp [19:13] ogra, thanks but the returns are way above my level of understanding. [19:16] ScottK: hrmm... fedora doesn't seem to have a pyclamav package [19:17] ScottK: I also attempted to build that package, but it kept wanting to use clamav-0.92 [19:17] Odd. [19:18] sommer: They may call it something different. [19:18] Other distros may have patches too. [19:24] From what I can tell KVM is Ubuntu's 'preferred' virtualization choice - right? I see lots of people talking about Xen, is anybody *using* KVM in production? === gouki_ is now known as gouki [19:26] ogra, I ended up searching by "ssh" and "local sound" and it according to the returns, it does not appear to be all that easy and requires that each program be setup individually. [19:27] gregbrady, essentially you only need to set pulse locally to listen to tcp ... and then set PULSE_SERVER or ESPEAKER in your ssh session pointing to the local pulse [19:28] and by production I mean people get fired if it dies too often :-) [19:32] * ScottK notes http://archives.neohapsis.com/archives/postfix/2008-05/0080.html and suggests that packaging the additional clamav signatures might be a useful thing for someone interested in getting involved in packaging for Ubuntu to do. [19:33] is there a good write up some where about raid performance tuning, as far as best practices go? [19:33] ScottK: sure, I'll give it a try :) [19:35] furicle: you can also try #ubuntu-virt [19:36] Ah - ok - didn't know about that one.... I'll take a wander. [19:36] furicle: what sort of use case are you looking at? production means different thing [19:36] *things [19:43] ogra, thanks for the tips but it appears as though I have a ton of research to do before I understand this stuff. [20:09] does anyone know anything about raid performance tuning? or best practice methods for bottleneck tracking? [20:14] hi guys, has anyone got a problem with ssh frozen in hardy ? [20:15] i would connect to the remote box, do things for about 10 secs, and it would be frozen [20:16] ssh only or the entire box? [20:16] jackfruit: Wired or wireless network? [20:16] jackfruit: from laptop to wired remote box [20:16] faulkes-_: what do you mean ? [20:17] faulkes-_: the only thing i run on the box now is ssh [20:17] I mean if you try to reconnect, does it allow you [20:17] yes it does [20:17] jackfruit: I experience periodic networking freezes on my laptop that sound like what you are experiencing. [20:17] and i'm in, i can run commands [20:17] 10 secs, boom, frozen [20:18] Does it come back if you wait? [20:18] seany: make sure your think about blocksize / etc when creating filesystem on raid [20:18] ScottK: i waited for about 2 -3 ' w/o anything so i break the connection [20:18] jiqiren: in so far as matching up raid blocks with FS blocks in relation to the files being created? [20:19] ScottK: weird thing is i've used ssh on my laptop before w/o any problems [20:19] That's not the same problem I'm having. [20:20] wireless [20:21] mathiaz: does the open-iscsi bug about sysfs deserve an SRU? im thinking about it and no [20:21] seany: what i mean is when you create your filesystem, make sure the stripe size (raid) and stride size (ext3 or whatever) match [20:22] jiqiren: right [20:23] ScottK: it's bugging the crap out of me, since i cannot do anything w/o ssh [20:24] seany: so if you have raid with 64k stripe, you'd pass this to make a fs (ext3): -E stride=16 [20:24] I'd suggest trying plugging the laptop into a wired network and see if that makes a difference. [20:24] (assuming your blocksize is 4096) [20:24] jackfruit: ^^ [20:24] Do anyone knows why wvdial is not on ubuntu 8.04 server? or am I doing something wrong? [20:25] seany: are you using ext3 or something else? [20:25] ScottK: i mean i'm talking on here right, and irc time out is what, 120 sec ? [20:25] jiqiren: right now i'm just playing around :). ext3 and NTFS over iscsi [20:25] hmm is there a timeout value in ssh [20:25] jackfruit: Yes, but ssh is a different protocol with different sensitivities. I'd suggest try it. [20:26] jiqiren: i've got 9 250g disks in a raid6 vg at the moment [20:28] seany: in that case you might want to make sure the vg/lv layer is also using correct matching sizes (I assume you are using software raid in linux then lvm on top?) [20:29] jiqiren, yes [20:33] jiqiren: are there any good docs about this online some where? [20:33] jiqiren: there's lots of stuff about raid config, but not alot about performance [20:33] ScottK: i'm wired up, still not letting me connect [20:33] now it's not letting me connect instead of letting me in for 10 sec [20:34] :)) [20:34] OK. Not sure what to tell you then. [20:34] seany: I don't think so... in the end default values are not that bad - software is getting better about already taking care of things like this [20:34] this is weird as crap [20:34] jackfruit: any chance of duplicate IP numbers on that LAN? [20:35] furicle: on my local one or remote one ? [20:35] jackfruit_: your laptop.... [20:36] nope, i'm pluged into directly to the modem [20:36] so much for that brainwave :-( [20:37] jiqiren: have you heard anything about iscsi performance numbers? vs nfs/local ? [20:38] jiqiren: back when I was cursed with taking care of exchange, iscsi sucked - there would always be timeouts long enough to force a failover with an active/standby config of exchange [20:39] hehe... i mean seany [20:39] seany: typically happened when everyone in the office checked their mail at the same time [20:39] jiqiren: what was your storage sitting on? [20:39] seany: but this was MS, I've never used iscsi on anything else [20:40] it was a dediated iscsi nas [20:40] <_ruben> MS's iscsi initiator is actually quite decent these days [20:40] <_ruben> "too bad" its even better than vmware esx's initiator [20:40] _ruben: i havn't had too many problems with it [20:41] the ms one, haven't had a chance to play with iscsi as a back end for esx [20:41] seems like using NFS is "all the rage" :-p [20:43] :( my last bonnie test segfaulted [20:46] ahh it is my laptop connection after all [20:46] this is weird [20:47] i've had had problem with it before [20:47] <_ruben> i kinda prefer iscsi over nfs (block device versus directory sharing) .. then again, dont have much experience with either, yet [20:57] _ruben: well you have obvious issues pushing block devices around when you're trying to do something shared (clusting) [20:58] hence the plug for NFS in some instances [20:59] <_ruben> seany: true [21:00] hello, how cant i upgrade my ubuntu 7.10 to 8.04 ??? [21:03] hollman: do-release-upgrade [21:04] Nafallo: thnaks, now... they say me that it's not recomendable upgrade in a session of ssh [21:05] oh? well. I thought it had magic for that stuff. [21:05] dunno then. [21:05] hollman: it's not, you should connect to the machine locally and/or using an IP KVM / terminal server [21:07] Deeps: thanks [21:07] upgrading... :P [21:11] zul: which open-iscsi bug ? [21:15] * faulkes-_ is loving iscsi [21:21] hi all [21:22] I have just installed Ubuntu 8.04 (desktop) on a 4GB RAM based computer. Do you guys know how can I install (apt-get) a "server" kernel for PAE support ? [21:23] /ping $m [21:23] /ping $me [21:23] No one? [21:26] so, goodbay [21:49] damn - cupsd apparmor problem: audit(1209761208.193:16): type=1502 operation="inode_permission" requested_mask="rw" denied_mask="rw" name="/dev/tty" pid=6667 profile="/usr/sbin/cupsd" [21:50] no files in /var/log/apparmor ... [21:56] nealmcb: http://www.mail-archive.com/cooker@linux-mandrake.com/msg107191.html [21:56] happens every time I try to print a particular document from evince, but not a different one [21:56] nealmcb: webmin is mentionned [21:57] nealmcb: the first paragraph is a good summary [21:57] nealmcb: The reason they all suck worse than vi is because they all aim far too [21:57] nealmcb: low [21:58] nealmcb: that goes back to the argument of webmin gives to much power to the targeted users [21:59] well, it is still less power than a shell, which they know less about and can do more damage with. does ebox keep a change history? [21:59] * nealmcb continues to read [21:59] nealmcb: not yet - there is a save change button - but you cannot review the change [22:00] nealmcb: this is the thing I like about xwrt [22:00] nealmcb: I can review the changes to be made before applying them. [22:00] * nealmcb just puts all of /etc under bzr [22:01] nealmcb: I've downloaded the latest version of webmin fron webmin.com and poked around [22:01] does bzr have a perl api ( !! ) [22:01] mathiaz: in a virtual machine I bet :-) [22:01] nealmcb: I only use virtual machines ;) [22:02] nealmcb: For example, I went to the ssh server module, didn't make any change and saved [22:02] nealmcb: the result was that sshd_config was modified [22:02] did it ask or warn? [22:02] nealmcb: since webmin deals with config files, that could you lead to configuration prompts during a package upgrade [22:03] nealmcb: no [22:03] nealmcb: considering the target audience of webmin it isn't the best thing to do [22:04] nealmcb: if you use webmin, that you'd be ask for configuration prompts on package upgrades [22:05] nealmcb: the other issue with that is that there are now 2 package on the system that have different ideas about what at configuration file should look like [22:05] that is a good point. will we get any advice from ebox before doing an upgrade (say yes to xyz, and no to def) [22:05] nealmcb: the package version and the webmin version [22:06] nealmcb: in the ssh example, webmin thinks that 'IgnoreUserKnownHosts no' should be part of the default sshd_config [22:06] nealmcb: whereas the default sshd_config doesn't have the option set (which default to no) [22:07] nealmcb: so although webmin reads the configuration files, it adds default values in the process, which are written back to the configuration file [22:07] nealmcb: you still need to have defaults set somewhere, but they shouldn't be imposed on the configuration file [22:08] !webmin [22:08] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. See !ebox instead. [22:08] guys, regarding the change history in eBox, we have almost implemented the "log admin" module, which will tell you before saving all the changes you have done [22:08] you will see a list of the modules changed and what has been changed [22:08] nealmcb: this may be the reason behind the statement: It is not compatible with the way that Ubuntu packages handle configuration files, [22:09] stuff like: module firewall -> added rule to output table, module network -> change configuration in interface eth0 from static to dhcp [22:09] foolano: great - so you'll be able to review which changes are going to be applied ? [22:09] exactly [22:10] and that's something we want to add by default in the framework, so the module developer doesn't have to worry about it [22:10] foolano: how about keeping actual diffs? helpful at upgrade time or for invoking help from better admins [22:11] foolano: agreed. that should be part of the framework. [22:11] nealmcb: you mean something like rollback option ? [22:12] nealmcb: what do you mean exactly by keeping diffs [22:12] ? [22:12] hm.. rollback is probably more complicated than just a diff [22:12] well rollback would be nice also, but detailed info on what it did is a prerequisite for any sort of troubleshooting or getting assistance [22:12] foolano: if you're updating a list of files, you could keep a diff between the file before and after the update [22:12] I want ebox to keep a log of all changes to any config file (especially those which aren't internal to ebox) [22:12] foolano: you know which files are going to be rewritten, so you can generate the diff [22:13] at a level of detail that allows rollback at least outside ebox if not within it [22:13] i c, that would be very easy to implement if you find it useful [22:13] nealmcb: right [22:13] * nealmcb is on the phone now [22:13] foolano: yeah - nealmcb use cases is much better than a rollback [22:14] foolano: knowing what happened to your system makes troubleshooting much easier [22:14] foolano: having a log of which action has been done on the system is great [22:14] * nealmcb ... and will be for 30 min or so.... [22:15] foolano: having a diff of the actual changes in the configuration file is even better :) [22:15] coming up for the next release :) [22:16] !ebox [22:16] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/eBox [22:16] the "admin log" was scheduled for release 0.12. We already had some parts of eBox using it, but we disabled it for 0.11.99 because there were some modules which needed to be slighltly modify in order to use it [22:17] modified [22:17] * [22:19] the only thing is there are some actions that, with the current design, take action immediataly, things like changing the password for a user within ldap is done immdiataly [22:19] foolano: the ebox factoids states "It is designed to work with Ubuntu/Debian style configuration management" - could you elaborate on that ? [22:20] foolano: I know you haven't written the ebox factoids [22:20] i think what that means is we have tried to stick to the debian policy when it comes to modifying configuration files [22:21] stuff like asking the user beforing overwriting a conf file if it has been modified by him and so on [22:21] before* [22:22] and there's also a software module to install system packages which is based on apt [22:22] foolano: right - I've seen that in webmin too. [22:22] foolano: how do you handle debconf questions ? [22:23] right now, we don't manage them, we install packages with DEBIAN_FRONTEND=noninteractive [22:24] foolano: ok [22:26] by the way, do you guys think that openldap in ubuntu will ever use the new configuration schema? [22:27] foolano: are you refering to the config backend 2.4 ? [22:27] mathiaz: exactly [22:27] foolano: probably - that should be discussed at UDS :) [22:27] we sent this email: https://lists.ubuntu.com/archives/ubuntu-server/2008-February/001141.html [22:28] mathiaz: hehehe, alright. It would be nice to have that by default, although i dunno how many things it could break :) [22:29] i mean, how-to's and stuff which expects the old-fashioned config backend [22:29] foolano: we'll probably have a discussion about openldap during UDS - your request is not the only one [22:29] foolano: having a way to insert schemas in a ldap tree could be useful to any application [22:30] foolano: re debian policy and configuration files, I'm not sure how ebox isn't compliant [22:31] IIUC the sections about configuration files in the debian policy define how packages and maintainer scripts should behave [22:31] it doesn't apply to application [22:31] mathiaz: i thought so too [22:32] otherwise any text editor would be subjected to the same policy [22:32] yeah, exactly what it states is you cant mess around other config file in maintainer scripts [22:32] but it doesn't say anything about upstream itself [22:32] foolano: as the purpose of ebox is to modify the configuration files, it can do it whithout being non-policy compliant [22:33] mathiaz: i agree on that [22:33] but i also think it's nicer to warn the user if it's going to overwrite stuff modified by him [22:34] foolano: yes - I totally agree with that [22:34] foolano: there are some ideas that can be taken from the debian policy and tried to adapted to ebox [22:34] mathiaz: i hope that the time i spent modifiying that is worh it :P [22:35] i would be pleased to discuss them :) [22:35] foolano: are you coming to UDS ? [22:35] yeah, but just a couple of days [22:35] foolano: great - we'll meet there and discuss it then. [22:36] foolano: when will you be there ? [22:36] i haven't booked my ticket yet, i was waiting to see when it would be better if you guys wanna talk about ebox [22:39] foolano: how long do you plan to be in Prague ? one day ? [22:39] mathiaz: two days [22:40] plus one for my journey [22:41] foolano: as you may know, there are two events: one is FOSSCAMP and then UDS [22:41] foolano: FOOSCAMP is more for upstream talking to us [22:41] foolano: UDS is more contrained in terms of time available [22:42] foolano: we have one hour sessions to talk about topics [22:42] yep, i attended the last FOSSCAMP [22:42] foolano: ok - did you see parts of UDS ? [22:43] nope, i couldn't cuz i had to catch my plane on monday [22:43] foolano: right - I remember now [22:43] my bloody plane was rescheduled and i had to leave earlier [22:43] foolano: so FOSSCAMP has less time constraint [22:44] foolano: we're busier during UDS as we focus on what we want to achieve for intrepid [22:44] i c [22:44] foolano: did you like the session you had at FOSSCAMP ? [22:44] so it's better if i go during the fosscamp? [22:44] mathiaz: yeah, it was useful [22:45] foolano: yeah - it may be better for you to target FOSSCAMP then. [22:45] foolano: If you want to be there to have influence over what's planned for Intrepid, you ought to be there for at least part of UDS. [22:46] foolano: well - you'd only be interested in one topic [22:46] foolano: or session [22:46] foolano: OTOH if you plan to base your work on ubuntu, it's better if you can be at UDS [22:47] ok, understood [22:47] * nealmcb returns [22:48] foolano: also, FOSSCAMP is friday and saturday - UDS is monday to friday the following week [22:48] Hi, I instaled ubuntu server 8.04 on virtual pc. But I don't know how to configure internet. [22:48] mathiaz: i'll check if i can extend a little bit my days in Prague [22:49] i can't i guess i'll attend FOSSCAMP for sure [22:49] if* [22:50] foolano: there will probably a session about system administration from a GUI/web interface [22:50] foolano: during UDS - that would be relevant to ebox [22:50] do you know when the schedule will be published? [22:50] foolano: it's the number one feature asked on brainstorm.ubuntu.com [22:50] foolano: nope [22:51] foolano: however since you're relevant to the topic, we could make sure that the session is schedule while you're available [22:51] that would be perfect [22:53] I have an Ubuntu 8.04 Server, is it possible to take the hard drive out and move it to a different server? [22:53] yes [22:53] Thanks [22:54] mathiaz: i think I could be in Prague from saturday to monday, so we could have a session in FOSSCAMP, and I could attend UDS on monday [22:55] mathiaz: anyway, i'll see if i can attend more days so you dont have to reschedule based on my attendance [22:58] foolano: I think saturday to monday would be the best option [22:59] cool, i'll try that [23:21] Hey, is it possible to set up software raid on a device without booting off the CD? [23:21] in other words, on a live system [23:22] Vlet: I'd think so as long as you're not wanting to make / part of the raid [23:23] sommer: nope, just storage... how would I go about doing so? what do I look into? [23:23] sommer: I've only done it from the installer, so I'm not sure what to look into :) [23:23] google, linux raid howto [23:24] duh :) [23:24] sommer: thankya [23:24] also, you might try raid+lvm linux [23:24] Vlet: np [23:31] dendrobates: Congratulations. [23:33] Hello all [23:34] ScottK: thanks. [23:34] trying to install 8.04 64 bit, and finding the install is unable to mount the CDROM [23:34] any suggestions? [23:34] This is a T-105 Dell Server [23:35] has anyone reported problems with this, I seemed to read something about the 64bit having this, and the 32 not [23:36] can anyone verify? [23:37] * nealmcb cheers for dendrobates [23:37] cpare: does it boot from the cdrom? [23:38] dendrobates: congrats from me as well :) [23:38] ? [23:38] sommer - it does boot from the cd without issue [23:38] only one CD-Rom drive [23:38] dendrobates: what did you do now? :-) [23:39] I tried to check the disk, but it needs to mount there as well [23:39] Nafallo: finally became MOTU. [23:39] cpare: is there any errors [23:39] dendrobates: ah! good to have you onboard :-) [23:39] if I hit the space bar rapidly, I can see soemthing about Invalid Parameter [23:40] cpare: if you hit alt+f2 you'll get a console, and alt+f4 is the installer output, is there anything there? [23:40] cpare: also, have you tried a different CD? maybe it's just a bad disc? [23:43] soren 1, iptables 0. [23:43] \o/ [23:44] does anyone know about setting up mysql server with jdbc? [23:44] heh, did you play table tennis agains iptables? [23:44] sommer: I'm not sure what kind of games iptables thought we were playing. [23:44] I am getting ready to burn the Desktop 64bit, then convert to server [23:44] I won. That's what matters :) [23:45] and see if that helps [23:45] heh, iptables plays for keeps! [23:45] it could be a bad disk [23:45] cpare: ya, I'd give it a try at any rate [23:45] there were threads about that online, but it seemed bunk [23:45] will do, thanks! [23:45] cpare: welcome [23:56] does anyone know about setting up mysql server with jdbc? [23:56] matt____: I've done it once... if I remember correctly [23:57] matt____: are you getting a specific error? [23:57] sommer: if by specific, here one is... [23:57] sommer: *recreates error [23:59] when using openoffice to try and connect to it, i get "unexpected exception encountered during query"