[00:00] I find tha ubuntu is much more friendly to a newbie. [00:00] Vog-work: That's true. Ubuntu generally has a more friendly culture. [00:00] Or at least I did a few years ago. [00:00] Also we do try to do stuff to make it easier that's hard to do in Debian. [00:01] Because we team maintian it's easier for us to implement stuff that touches several packages. [00:01] in terms of stability there are no issues at all besides the inherent risk of using newer packages? is it totally world-class? [00:01] My experience has generally been good. [00:01] likewise [00:01] Newer also means more support for newer hardware too. [00:01] So has mine, Debian is really quite stable, but will be missing features available in ubuntu. [00:02] that said, lenny's quite current for a debian release [00:02] Yes, but when did Lenny freeze? [00:02] Deeps: I'll need to go look at lenny, it's been awhile. [00:02] whats a *big* feature ubuntu server has that debian lacks? it seems apt-get is 90% of what we need [00:02] ScottK: unsure, but it does pack a .26 kernel i believe [00:02] in terms of ease-of-use that is [00:03] As an example, even though Lenny just released and 8.10 released months ago, 8.10 has a newer kernel. [00:03] baz: out of box support for various raid controllers, even the crappy on board ones. [00:03] Deeps: Intrepid, that's 4 months old has .27. [00:03] ScottK: no doubt, i meant compared to debian releases of yore, it's quite current by debian release standards [00:03] Deeps: Certainly. [00:03] ScottK: i was in no way attempting to compare against ubuntu [00:04] I use Ubuntu for the LTS, and because it just works better than others [00:04] Certainty of support times is another difference. [00:04] ScottK, what do u mean by that [00:04] Ubuntu Server is supported for 5 years. Debian Stable is supported one year after the next release. [00:05] So how long is Lenny supported for? [00:05] baz: tasks, ubuntu has predefined tasks for webservers, mailservers, printservers, and so on [00:05] ScottK: Good call I always forget to point that out. [00:05] baz: which makes installation of such services relatively simple [00:05] Debian has tasksel too. In some ways not as refined however. [00:06] Deeps, i think I have seen that before - are u refering to checking the LAMP box during install and having it all built automatically for you? [00:06] baz: thats the one [00:06] Myself I like that Postfix is the MTA of choice, since that's my MTA of choice too. [00:06] yeah, thats too awesome [00:06] Baz: It's weird I always find things that happen automatically with ubuntu that I used to have to do manually. But I never remember them after thae fact. [00:06] baz: like ScottK says, debian has tasks too, but they're not as refined in some ways [00:07] That said, Debian does stop and spend a long period of time fixing bugs and stabilizing stuff, so it is likely to be more stable for some uses. [00:07] The user base for ubuntu is so much larger and the fourms are much more active. So when you get thet weird problem somoen else has probably already had it and figured it out. [00:08] Vog-work, big time, i often append "ubuntu" at the end of a google search even if my question is not ubuntu related and i get better aNSWRS [00:08] And theose bug fixes that ScottK mentions are subsequently incorporated into the Ubuntu code base most of the time. [00:08] Yes, but we get more new versions with new features in the meantime, so who knows. [00:08] one think I thanks is the MOTU effort to incluide packages in main [00:09] I find Ubuntu very stable for what I do. [00:09] Of course since I'm an Ubuntu developer and have fixed a lot of stuff that affected me, it's not random. [00:10] so it seems ubuntu is the way to go for me... I am not running huge clusters that require 99.99999999999% availability and i dont have a full team of geeks backing me up, so ease of use is important [00:10] and of course the biggest community [00:10] Baz: alot of us are in the same situation. Hence, why the community exists. [00:11] baz: that said, if you're looking for an impartial opinion, asking in #ubuntu-* whether ubuntu is better than X isn't likely to lead to many votes for X ;-) [00:11] hehe, true [00:12] but instead of votes i got logical facts [00:12] By the way it makes me proud to be part of it when I see a user come into the channel asking a question and then a semi clear answer can be hashed out. [00:12] and I think it is easier to contribute to ubuntu than debian, Ubuntu has launchpad, the whole infraestructure to work with an community operating system [00:13] On that noe I need to head home.... :) Good luck baz [00:13] thanks Vog-work! [00:13] indeed, best of luck [00:13] the only thing I dont like about ubuntu is the logo [00:13] lol [00:13] * jmedina still prefers slackware logo (not the new one) [00:13] bed time for me, work in 8 hours :/ [00:14] * Vog-work is a Tux purist [00:14] * jmedina is using a gentoo t-shit, so I look a linux hardcore user XD [00:14] * ScottK gives Vog-work a smack just to remind him of the good old days in #debian. [00:15] jmedina: Gentoo T-shirt says to me "way to much free time". [00:16] my last emerge world was the day I started to use ubuntu :) [00:16] ScottK, actually, using ubuntu there is a lot of spare time [00:18] I have not used gentoo since I migrated to dapper in desktop and servers [00:18] but not these days, people in crisis wants to learn and migrato to linux, most here in mexico [00:19] this econimical crisis its good for linux and open source [00:20] I have to go [02:17] today i used aptidue to update ubuntu server 8.10 and get and error message about GPG (invalid signature) error on http://security.ubuntu.com and don't know how to fix it. ( screen captures here http://www20.tok2.com/home/naaikada/Ubuntu%20Server-2009-03-11-08-48-42.png ). can I resolve this problem? [02:20] lonelywind1403: What happens if you do sudo apt-get update? [02:21] vorian: All the Main rdepends for libdb4.2 are gone, so it's likely killable. I'm patching kolab right now, so maybe you could hunt through for any others that are left? [02:24] ScottK: still that error message (screen capture: http://www20.tok2.com/home/naaikada/Ubuntu%20Server-2009-03-11-09-23-16.png) [02:27] lonelywind1403: What does "dig a security.ubuntu.com +short" return? [02:30] ScottK: it return 91.189.88.37 [02:31] Well that's at least right. [02:34] lonelywind1403: is there a proxy between you and the internet? [02:36] lonelywind1403: https://bugs.launchpad.net/apt/+bug/117010/comments/3 give an example of how to manually add a key for apt. I'd try that with the key ID in the error message. JanC's question is also a good one. [02:36] Launchpad bug 117010 in apt "invalid GPG key - is this a big problem? (dup-of: 24234)" [Undecided,New] [02:36] Launchpad bug 24234 in apt "apt-get update failing with bad signature." [High,Confirmed] [02:37] no, this server is connect direct to internet [02:46] uvirtbot: trying sudo apt-get update -o Acquire::http::No-Cache=true and having same error message [02:46] lonelywind1403: Error: "trying" is not a valid command. [02:51] lonelywind1403: uvirtbot is not human ;) [02:56] JanC: yeah, but i'm wondering how can it answer my question like that. ( searching ??? ) :D [03:00] lonelywind1403: it saw Scott's mentioning of a Launchpad bug # [03:05] JanC: did you get same error message at this time ? [03:05] I don't have any apt key errors no [03:08] JanC: if not, what if U send me a pgp database ( /etc/apt/trusted.pgp ) [03:13] lonelywind1403: Did you try the manual key adding step in the bug comment I linked you to? [03:22] ScottK: ~$ gpg --keyserver subkeys.pgp.net --recv 40976EAF437D05B5 && gpg --export --armor 40976EAF437D05B5| sudo apt-key add - [03:22] That looks right. [03:22] ScottK: result is ( gpg: imported: 1) [03:22] Now try sudo apt-get update again. [03:22] but when run apt-get update [03:23] still same error :( [03:23] Odd. [03:23] lonelywind1403: You've got the correct key then. I don't know what to tell you. [03:28] ScottK: i'll try copy /etc/apt/trust* from other server, hope it work :( [03:29] Good luck. [03:33] ScottK: bad news , nothing changed :( [03:34] I suspect something else is broken [03:35] I think so, but I don't know what. [03:35] maybe check for a leftover lock file for the apt/gpg database? [03:35] or something like that [03:38] may be something outdate ( cause i'm just installed this system from yesterday ) [03:39] A lock file would give a different error. [03:41] ScottK: I would hope so, but I've seen too many errors being masked by apt and/or dpkg to trust that :-( [03:47] All: should I post this error on forum ? [04:01] When I'm calling a blocking function (like recv(), accept() or something similar) in a process which has just a single thread (that is, I use no threading library or such).. does the blocking call prevent the process from being scheduled for execution? The reason why I ask is this: I have a lot (>400) of small worker processes, 80% of them in a blocked state because they are waiting for data to be received. However, scheduling all 400 p [04:01] rocesses is just a nightmare for the platform. So it'd be good if only processes are scheduled, which are not in blocking mode. === rgreening_ is now known as rgreening === lonelywind1403 is now known as Guest26356 [05:50] ScottK , JanC: the problem is solved, :D ( flowing this post http://ubuntuforums.org/archive/index.php/t-821058.html ) [05:52] lonelywind1403: what exactly worked? [05:52] sudo apt-get clean [05:52] cd /var/lib/apt [05:52] sudo mv lists lists.old [05:52] sudo mkdir -p lists/partial [05:52] sudo apt-get clean [05:52] sudo apt-get update && sudo apt-get upgrade [07:10] hello [07:12] could you tell me please, that a basic Apache (on Ubuntu Server 8.10) which worker uses as default? worker or prefork? [07:12] sorry, prefork or mpm [07:14] Depends. dpkg -l will tell you [07:15] thank you, apache2-mpm-prefork is installed :) [07:15] I allways forget about basic commands, like dpkg -l :( [07:16] :) [07:16] I'm in the process of installing Ubuntu Server on my old Dell [07:17] I've wasted a whole day trying to bring up NetBSD on another old machine and...basically I give up. I need to get some work done and Ubuntu Server seems like the path of least resistance. [07:23] hello _coredump_ [07:23] <_coredump_> hello ball [07:29] What's a light Web server to run on Ubuntu Server ? [07:30] apache, lighty, cherokee, nginx, thttpd [07:31] could you please take a look at my Apache2 config? I'm trying to tuning it... http://pastebin.com/d775394dc [07:31] hads: thanks [07:32] the HW: AMD6000+ x2 64bit, 4x2 GB dual channel DDRII 800MHz, SataII winchesters in Raid array [07:32] Where can I find out more about local Ubuntu communities? [07:33] this a web server with a few domains, but I'd like to host more domains... in the evenings 500+ users are online [07:34] How do you host more than one domain on a Web server? Do they all share a single IP address? [07:34] yes [07:34] Depends [07:35] * ball sighs [07:35] I'm pretty sure I want to set up a loco [07:35] ..or at least a local user group [07:45] kirkland, With ecryptfs, will my folder names be encrypted as well? [07:46] kirkland, Will ecryptfs work on top of ext4? [07:46] kirkland, What kind of encryption are ecryptfs using as default? AES 128? Can I change this myself? [07:48] what's the command to type to update an Ubuntu Server host? [07:49] ball, You mean between ubuntu releases, like 8.04 to 8.10? [07:49] Blinkiz: no, just to make sure that security patches are applied etc. [07:49] sudo apt-get update && sudo apt-get dist-upgrade [07:49] ball, oki.. Yeah, like hads said [07:50] oh, not dist-update. Should be just "upgrade" [07:50] dist-upgrade [07:51] Thanks, that's running now [07:51] ball, :) [07:53] bye people [07:53] I need to decide what Web server to install [07:53] hello drdebian_ [07:55] If I install thttpd, will it automatically be configured to launch at boot time? [07:57] Likely [07:57] ball, But they way, if you want security updates to be installed automatic on your ubuntu server, you can do "sudo apt-get install unattended-upgrades". Config file will then be in /etc/apt/apt.conf.d/50unattended-upgrades" [07:59] Hmm... that might be a good idea. [08:00] hello BrunoX1ambert [08:11] How do I turn on sshd? [08:12] Install it [08:15] E: Couldn't find package sshd [08:15] I think I tried "openssh" too [08:15] should it be OpenSSH ? [08:16] Search [08:16] "apt-cache search" and grep are your friends. [08:18] ball: openssh-server. [08:18] That's great, thanks. I've written down te "apt-cache search" tool too. [08:18] ball: And most services installed will be launched at boot time, unless you change it with update-rc.d. [08:19] That's good, it'll save me a job [08:21] <^law1^> guys, if wan install tomcat which package to install ? there r so many package tomcat6, tomcat6-admin, tomcat6-common, tomcat6-docs, tomcat6-examples, [08:21] <^law1^> tomcat6-user [08:22] ^law1^: depends on what you want to get [08:22] ttx: Error: "law1^:" is not a valid command. [08:23] <^law1^> i want to get tomcat run just like in windows [08:23] Usually "tomcat6" is what you want (the system daemon) [08:23] "tomcat6-user" is a package that lets you deploy multiple private instances [08:24] anyway, the server guide is your friend here [08:24] http://doc.ubuntu.com/ubuntu/serverguide/C/tomcat.html [08:26] <^law1^> oh i'm opening it now :) [08:30] <^law1^> ttx, THX this is a great guide :D [08:40] #openvz is silent, but maybe someone here knows. How do I increase the maximum memory available to an OpenVZ VE? [08:40] It seems to default to 256MB, which isn't enough for fat ol' GHC. [08:44] What's a GHC? [08:47] hello, is there a server tuning guide? [08:47] moin [08:49] hello incorrect [08:49] hello ball [08:50] <^law1^> moin kraut :) [08:54] hello oly [08:54] hi ball, [08:54] * thefish feels the love [08:55] <^law1^> hi ALL :) [08:55] hello to you all :) [08:57] I'd be awfully tempted to install Ubuntu Server on the box behind me if I could get the installer to work. [08:57] where does it get to ? [08:59] oly: if I can get this rewired slightly, I'll try it again. [08:59] okay, just asking possible you need to provide a parameter to the kernel to get it to boot or something :p === mdz_ is now known as mdz === ScottK2 is now known as ScottK [10:10] what is the ulimit pipe size? === jussio1 is now known as jussi01 [10:34] incorrect: "ulimit -p" [10:34] i mean what is pipe size [11:00] guessing some one here might know the answer to this, can i copy files over ssh but use sudo / su to increase my permissions ? [11:03] could anyone test my upstream for me? [11:04] someone with at least a 300kb/s download rate [11:04] jserver, why don't you use http://www.speedtest.net/ [11:05] not accurate, im am on a 15Mb connection and it reports 5 [11:05] so [11:05] could well be accurate [11:06] umm [11:06] 1/3 of the bandwith alloted to me [11:06] yep [11:07] but my downstream im not worried about, I get so many mixed results for my upstream, I would like to have someone just max it out to see what it can do [11:10] of course I dont wanna bother anyone, just figured I would ask and see if anyone would. [11:14] do you have an iso image i can pull from you then? [11:14] url? [11:14] wget jserver isn't working [11:16] my server is locked down at the moment [11:16] well that isn't exactly helpful [11:16] is ddc reliable? [11:16] dcc [11:16] I will wget a file from you [11:17] 1 min [11:22] heres a 24mb file, should be enough to test the bandwidth [11:25] http://www.sites4christ.org/LimeWireLinux.deb [11:25] one of my client sites [11:28] jserver: 90KB/s [11:29] you got to be kidding me [11:30] cant run a server on that kinda shit [11:30] * jserver smacks his ISP in the ear [11:31] suppose to be a whole lot more than that [12:11] Anyone know of a way I can monitor a folder for changes? [12:13] nietchzie [12:13] my bad, wrong window [12:14] forgive the spelling too, hehe [12:22] heath|work, inotify-tools ? [12:25] I will look into that thanks [12:31] i dont know for sure, but i know inotify can notify you of filesystem changes i have never tried the command line tools though [12:31] only ever used it in python :p [12:49] funny you mention python, that's what I am reading now [12:50] hmm.. check out dnotify [12:51] i used it to detect when a ftp script had down loaded a file so that i could act on it [12:51] works really well [12:53] oly, does it run all the time, or do you cron it? [12:54] well the ftp script is a cron job, but the python script is a daemon thats always running in background [13:28] a quic question: Is rc.local runned after the network is up? [13:29] i have to wget some configuration files everythime a computer starts [13:29] yes [13:29] unless network is dhcp [13:30] if you want to run something after bringing up network interface, /etc/network/if-up.d/ is much better place [13:35] The network is using DHCP. does if-up.d have root access? [13:35] scripts in that directory are executed with root privileges, yes [13:36] ok, and all i have to to is to add it and make i executable? [13:36] yes [13:37] tanks, you are an angel, ivoks [13:38] Scix: Note, though, that it gets executed for *every* interface. Even the loopback interface. [13:39] Scix: ... You can check the $IFACE environment variable for the right interface. [13:40] soren, tanks for the note :) [13:40] Sure :) [13:40] soren: you will never be a better angel than I am! :p [13:40] :D [13:41] * soren bows to ivoks [15:12] compiling openldap-2.4.15..when running make test port 9001 appears already in use?? any workaround? [15:16] ahh.. [15:16] cliebow: lsof -i:9001 should tell you whats currently using it [15:17] a n opensuse bug says pull localhost from ::1 line in /etc/hosts.. [15:17] that got things going.. [15:20] Hopefully you don't use anything ipV6 anywhere [15:20] nope.. [15:20] ill set it back if the tests are successful.. [15:47] I am looking for a smarter web load generating tool, siege is great, but it doesn't spider the side or pull any of the links in the page [15:47] does anyone here know of a tool to do this? [15:50] incorrect: what some people do is use a script that "replays" real traffic based on webserver log files, but of course you need the site to already be in production for that [15:51] I have a set script i want to simulate [15:52] incorrect: wget? [15:53] i could write something, but i don't have massive amounts of time [15:59] incorrect, does this help? http://www.joedog.org/index/siege-faq#a21 [15:59] looks like you can build a url text file [16:00] heath|work, siege isn't very smart [16:00] i am using siege right now [16:00] i think apache's flood might do the trick [16:14] ScottK: fyi, I should post a FFe for bug 323601 by tomorrow. It should end up in your plate, being server-related. [16:14] Launchpad bug 323601 in likewise-open "Upgrade to likewise-open 5 for Jaunty" [Wishlist,In progress] https://launchpad.net/bugs/323601 [16:16] ttx: likewise-open is in Main, so that needs ubuntu-release approval, not motu-release. [16:17] ScottK: It would be done as a separate, new package, since you can't update from likewise-open 4 to the 5 version. [16:17] so it should hit universe... [16:18] Ah. [16:18] and conflict with the version in main. [16:18] ScottK: the Ffe will detail that aspect. [16:18] ScottK: just wanted you to know about it before I post it. [16:18] Also make sure you cover if this relates to any spec. [16:27] ScottK: the main reason for doing like this is to protect users. likewise did not create a clean upgrade path (IMO) from 4.x to 5.x. === zul_ is now known as zul [16:27] ScottK: it requires users to rejoin the domain. [16:27] dendrobates: Makes sense. My major question would be should this just wait for Karmic and get an early backport at this point? [16:28] I'll have a look at it when it's ready. [16:29] ScottK: we have been discussing that, but there is a huge customer demand, 5.x was actually released during the intrepid cycle and we held off then. [16:29] I see. Grumble. [16:29] ScottK: so users are doing the upgrade themselves, in not so good ways. [16:30] Well if Canonical is up to doing the archive admin work on it and there's a ready package there isn't a lot of reason to not put it in. [16:30] ScottK: but look at it and let me know what you think. [16:30] Also if we get it in now, we could backport to Intrepid/Hardy soonish and reduce non-standard installs. [16:31] Where does bug 341030 belong? [16:31] Launchpad bug 341030 in ubuntu "Local mail delivery missing in default server install" [Undecided,New] https://launchpad.net/bugs/341030 [16:32] * ScottK looks [16:33] bdmurray: That behavior is by design. [17:12] bdmurray: I commented in the bug. [17:13] ScottK: great, thanks for doing that! [17:14] The reporter won't be happy, but at least they have an answer. [17:24] how would i add a user to sudoers list [17:26] orudie: you either add the user to the admin group, or you can explicitly add the user in the sudoers file: https://help.ubuntu.com/community/Sudoers [17:57] is it a good idea to match the hostname with a domain name /? [17:57] if the primary use of the vps server is a website for this domain name [18:03] how can you match hostname with domainame? [18:03] hostname.domainame [18:03] server.server? :) [18:05] ivoks, hi friend ? [18:05] :) [18:05] ivoks, we are about to do the transition of the email accounts to the host you helped me with a few weeks ago [18:08] ok [18:10] ivoks, can you help me with some stuff ? [18:10] Anyone knowledgeable about LVM? After completing a 8.10 install I added a vg and created my lv. The three lvs are mounted as /tmp /home and /var. /var fails to mount on boot but the other two automount ok. The /var lv can be mounted after boot so I know it is ok. Any idea where to start fixing this one? [18:10] sure [18:11] jcrow: /etc/fstab [18:11] ivoks: /etc/fstab is configured for /var partition [18:12] how is configured? [18:13] You may have hit the nail on the head. I went to look at the fstab and noticed that the line for var has default vice defaults for the mount options. I did not notice that before. I am rebooting now. This is a server so it takes a minute or five. [18:17] ivoks: /etc/fstab was the problem. I had been digging through the Ubuntu boot process for the last hour to find where it might be failing. All I needed was a kick in the right direction. Thanks! [18:17] jcrow: np [18:25] ivoks, how reliable is zoneedit ? [18:26] what's that? [18:27] ah... dns service [18:27] never used, i have my own dns servers [18:28] I've had success with dnsmadeeasy [18:28] they're quite diversified, and unbelievably cheap for what they offer [18:29] http://www.dnsmadeeasy.com/s0306/price/dns.html [18:34] ivoks, so if i want my email clients to be set up as pop, i cat use mail.mydomain.com as incoming mail server, it should be pop.mydomain.com instead? [18:38] err? [18:38] i don't understand [18:38] you can use single hostname [18:38] you don't have to set up different hostnames for different services [18:39] Hello there [18:39] Any good guys here ? [18:40] no, only bad guys [18:40] :P [18:40] I've finally shutdown the server after 2 months or more.. [18:40] not because id' wann't to but because I had to, they had to do some electricity maintance.. [18:41] Somehow something is going wrong [18:42] I got message that filesystem check went wrong or smth and to look into /var/log/fsck/checkfs http://pastebin.ca/1358382 [18:42] What exactly does this error mean ? [18:42] Does it has to do something with my external USB hard drive ? [18:42] did fstab it wrong ? [18:42] did I fstab it worng ? [18:43] Any ideas ? [18:43] you don't know what UUIDs are? [18:43] ivoks: id of hard didsk ? [18:43] disk [18:44] right [18:44] uniq id of the partition [18:44] ivoks: and something is crossing over [18:44] no... [18:44] you are using -A switch [18:44] and fsck can't read lines with UUID in /etc/fstab [18:45] does this happen during the boot or you are manually running fsck? [18:45] ivoks: no during the boot :) as I said the first time shutdwon and start today: http://pastebin.ca/1358385 [18:46] this is my /etc/fstab [18:46] not sure if it's right [18:46] not during the boot OR no, during the boot? [18:46] i see [18:47] you configured your fstab wrong [18:47] # [18:47] UUID=B4B0049FB00469E6 /mnt/p2p ext3 relatime 0 2 [18:47] that last '2' at the end means 'stop everything if you find an error on this partition' [18:47] '0' would mean 'just ignore everything' [18:48] for partitions that aren't crucial for server's operations, it should be 0 or 1 [18:48] ivoks: funny part is I can't see my 1tb external HDD if I do fdisk -l, just 200gb /mnt/p2p and /home, / 160gb: http://pastebin.ca/1358388 [18:48] 2 will stop booting in case of an error, as you can see in your case [18:48] ivoks: so there are errors :S [18:48] my guess is that you don't have that disk in server [18:48] or your IDE controller is broken, or something else [18:49] anyway, replace that '2' with '0' [18:49] reboot [18:49] and then fix the partition problem [18:49] ivoks: about IDE's, I'll stop server now, check the cables [18:49] frojnd: don't [18:49] frojnd: fix fstab, boot the system, let it work [18:50] and then check if there are driver problems or something like that [18:50] so just change 2 into 0 [18:50] maybe it isn't hardware problem [18:50] I see [18:50] her, 2 into 0 only for /mnt/p2p [18:50] /sher/yes/ [18:50] grr... [18:50] :) [18:51] restarting [18:51] let me see the monitor [18:51] that UUID is wrong [18:51] you added it yourself? [18:52] it's too short [18:52] for ext3 [18:53] ivoks: I've changed it into 0 and it still says that UUID is worng :) [18:53] it doesn't exist [18:53] cause it is wrong [18:53] ivoks: for /mnt/p2p [18:53] but the server will boot now [18:53] ivoks: yes [18:53] ivoks: before you help me further... [18:53] ivoks: I have to plugin CD-ROm [18:53] frojnd: so, again, you added that uuid yorself? [18:54] yourself [18:54] ivoks: I think I foud some how to identify UUIDS.. [18:54] ok, well, this one is wrong :) [18:54] ivoks: If I have onyl ONE IDE cable and HD and CDROM [18:54] log in to your server [18:54] can I put jumpers to Cable select ? [18:54] don't touch hardware [18:55] so CDROM and HD would be rocognized [18:55] just boot your server [18:55] ivoks: yes yes.. [18:55] ivoks: this one isn't the ubuntu issue :) [18:55] before I'll set any UUIDS I'd also like to plugin CDROM [18:55] now I only have HD [18:55] ok [18:56] I jsut have to put pins on Cable select ? [18:56] yes [18:56] on bouth CDROM and HDD ? if I have one cable for both ? [18:56] ok [18:56] brb [18:56] Thanx for now, and please don't go away :) [18:56] I'll need some more help from you :) === JanC_ is now known as JanC [19:05] ivoks: here ? [19:06] how do you view the uid of a user? [19:06] yes [19:06] ivoks: now I only see 160GB DISK :D [19:06] however cdrom works [19:07] ivoks: this has something to do with pins I've select right ? [19:07] Hey i need some information on importing and exporting mysql thru phpmyadmin. [19:07] I've put pins on hdd to cable select, also on cderom but only cdrom is recognized [19:07] frojnd: i still don't understand why you touched hardware in the first place [19:07] :S [19:08] ivoks: I'd like cdrom to be there... [19:08] your system was fine [19:08] ivoks: but without cdrom [19:08] you really need it? [19:08] I cant remember my passwords to my databases but if im exporting them to a plaintext file that shouldnt matter once i've imported them to the new system? [19:08] heath|work: grep $USERNAME /etc/passwd [19:09] thanks jbernard [19:09] chamunks: mysqldump --default-character-set=utf8 -u root -p --all-databases &> mysql_dump_all_databases.sql [19:09] Im pretty sure that my assumption is correct but i need someone who's done this before to tell me that im okay to format my old server. [19:09] chamunks: then, on import, again --default-character-set=utf8 [19:10] heath|work: more specifically: grep ^$USERNAME /etc/passwd | cut -d: -f3 [19:10] ivoks: ocasionally :\ [19:10] ivoks, thanks so much. since i dont like using root for my web apps and only like assigning separate users for separate processes will this be easy to do on the new system? [19:11] ivoks: not important as we speak... I'll play with cables some other time, How can I figure iut ouw UUIDS for both disks ? the one that's on IDE cable and the one that's plugged int via USB ? [19:11] ivoks, im not mysql savvy I just know that its required to keep settings etc. [19:11] ok, too much lines with 'ivoks' on the screen [19:12] :) [19:12] frojnd: sudo vol_id /dev/sdXY [19:12] kirkland: do I have to update the kernel on hardy if I want to test/run kvm 84 from the ubuntu-virt PPA ? [19:12] oops :P ill keep those lines you hooked me up [19:12] y i need some help becouse im new whit linx ,i instal verlihub but i canot run any lua scripts [19:12] mathiaz: imo, yes; i don't see how it would work without kernel module of same version [19:13] chamunks: different mysql servers/clients have different character sets conffigured [19:13] while we are at it [19:13] Im going ubuntu to ubuntu im just going through some growing pains. [19:14] mathiaz: why do we predefine latin1-sweeden as default character set in mysql? :) [19:14] mathiaz: why not... utf-8? [19:14] oooh [19:14] Sweden! :-D [19:14] sweden, right :) [19:14] ivoks: probably because this is what upstream does... [19:14] it's a compile time option [19:14] iirc [19:14] ivoks: I agree we should use utf-8 - I think 5.1 may use it though [19:15] didn't try it yet :/ [19:15] chamunks: well, however, it's wise to add --default-character-set=utf8 during export and import [19:16] ivoks: ok, which one is correct ? ID_FS_UUID=c3652bfc-ad3b-479f-96d0-fbac77abf314 or ID_FS_UUID_ENC=c3652bfc-ad3b-479f-96d0-fbac77abf314 ? [19:16] standards rock [19:16] frojnd: notice that they are the same :) [19:17] ivoks: right :) [19:17] ivoks: thanx [19:17] I just have to replace the previous one with this one now [19:18] yes [19:19] ivoks: what about the USB one ? If I can't even see it on fdisk -l [19:20] mathiaz: i just reinstalled my hardware with hardy server [19:20] mathiaz: let me do some harder kvm-84 testing on hardy [19:20] mathiaz: previously, i just sniff-tested it on my mythtv box [19:20] mathiaz: but i can't risk downtime on that machine--my wife kills me when mythtv goes down :-) [19:20] frojnd: usb/scsi/lvm/ide/whatever everything's the same [19:20] "mission critical server" in the kirkland household [19:22] ivoks: after what command? [19:22] ivoks, what if for some reason the databases "collation" before export is latin1_sweedish_ci [19:23] For some strange reason. [19:23] :) [19:23] that's default in ubuntu [19:23] chamunks: then you can import without --default-character-set [19:24] so that will convert to utf8 with no hicks? [19:24] nope [19:24] it will stay like that [19:25] sorry i meant if i used the set=utf8 switch you sent do you think it would be okay? [19:25] no, don't use switch [19:25] i told you to export and import with that switch [19:25] if you exported with another charset (and you did), then don't use that switch [19:25] I don't think I udnerstand what u meant with usb/scsi/lvm/ide/whatever everything's the same ivoks [19:25] right my wording is bad [19:26] I havent exported yet im looking at the phpmyadmin stuff to give myself a visual. [19:26] frojnd: it will work with usb [19:26] ivoks: what exactly will work with usb [19:26] 20:19 < frojnd> ivoks: what about the USB one ? If I can't even see it on fdisk [19:27] ivoks: I can't see the 1tb sotarage on fdisk -l :P [19:27] which is connected via usb [19:27] ill use the default charset i wont complicate things thanks for the advice. [19:28] frojnd: check dmesg [19:28] ivoks: on what should I be throwing my eye ? [19:28] ivoks, thanks again take care. [19:32] ivoks: at the end some usb device is using adresses from to a lot :D [19:32] now I can't see the whole dmesg [19:33] just the last lines [19:33] ivoks: at the beginning of the boot process there is a message that this ID..something should use another chip, perhaps another driver... dunno what exactly it say [19:35] ivoks: here is the dmesg output, right after the boot: http://pastebin.ca/1358427 [19:36] take care people... [19:36] ivoks: don't go :S [19:37] frojnd: something's wrong with your usb disk [19:37] try disconnecting and connect it again [19:38] ivoks: ok [19:38] ivoks: wait [19:38] when in OFF mode right ? [19:38] maybe usb port isn't recognized :S [19:38] for the love of god... [19:39] i better disconnect than continue... [19:39] bye [19:39] hehe :) [19:39] I scared him [20:00] kirkland: ok - I was planning to update to kvm 84 on my hardy server once I've finished with this round of iso testing [20:00] kirkland: and then re-rerun the iso tests with kvm 84 - that should be a good test case [20:01] kirkland: as far as stability is concerned. [20:01] mathiaz: i would definitely appreciate your help testing kvm-84 [20:01] mathiaz: i'm still working out the kernel bits [20:02] kirkland: ok - as far as I can back out the changes easily I'm glad to help [20:02] kirkland: but for now I cannot use kvm 84? I should wait for the kernel bits? [20:02] mathiaz: i think you can use the userspace for kvm84 [20:03] mathiaz: but you won't get all of 84's goodness until you use the kernel module too [20:03] kirkland: hm ok. [20:03] kirkland: and the plan is to try to push both userspace and kernel ? [20:03] mathiaz: the userspace bits will get you a much newer, better, shinier qemu [20:04] mathiaz: the kernel space will get you a better kernel module, for handling the virtualization acceleration stuff [20:04] mathiaz: there's important magic in both places [20:04] mathiaz: bugs fixed by each [20:04] kirkland: my question is actually whether I should upgrade to userspace 84 now and if that would be helpfull to you [20:04] mathiaz: and potential regressions introduced by each :-) [20:04] kirkland: or wait for you to sort out the kernel bits [20:04] mathiaz: i think it would be interesting to test the userspace [20:05] mathiaz: knowing that existing kvm-62 kernel issues won't be solved by a kvm-84 userspace [20:05] mathiaz: however, kvm-62 userspace issues (qemu-related) may well be solved by kvm-84 userspace [20:05] kirkland: ok - so I'll give 84 a shot once I've finished this round of iso testing [20:05] mathiaz: thanks. [20:09] New bug: #341278 in tomcat6 (universe) "CVE-2009-0781: XSS in tomcat6 and tomcat5.5" [Undecided,Confirmed] https://launchpad.net/bugs/341278 [21:24] wereHamster: Looking at your script move finished passwords, it has no clear support for adding in authorization. Where to stick in a user/pass? [21:25] I tried a few things with no luck [21:27] lemme start again, this script: http://pastie.org/357738 gives the error: ** Exception: Problem with http://localhost:9091/transmission/rpc [21:54] I just installed Denyhosts on my server, and now I can ssh into my box. The /etc/hosts.deny is adding my IP -- how do I change that? [21:54] do I need to add myself to the /etc/hosts.allow? [21:58] hrmm [22:04] does anyone know if there is a CLI prefered method to upgrade to a new release? i know upgrade-manger -d work in GUI land, but has anything been made to work for CLI? i usually just change my apt sources to newer version but that may not be the safest i am thinking. thanks. [22:13] hansin: do-release-upgrade [22:15] how about sudo apt-get dist-upgrade? [22:16] ajmitch: thanks, that looks like CLI style update-manager equivalent. [22:17] goofey: that is how I usually do it, and have never had issues (after updating sources list), but I think there can be a small chance if there are some major infrastructure adjustments that something could be missed. i think "dummy/transitional" packages are sometimes used to help alleviate this. [22:18] hansin: ahh, I wasn't sure that was the "official" way to do it, just one way I happened to know of [22:20] goofey: that is how i have always done it actually, but wanted to find a more "official" way. i'm not saying this is it, but i bet it is considered more appropriate. maybe it doesn't matter, but i'm going to try this way when jaunty gets released. [22:21] hansin: my guess is there will be "upgrade your server to jaunty" webpages after the release too [22:25] hi [22:25] anyone here? [22:25] sure, why not. [22:26] I was wondering if you guys know good website for a beginner to set up chtoor [22:26] chroot* [22:27] MK-BB: take a look at this; there seems to be quite a few options though don't know much about how to chroot: http://www.google.com/search?hl=en&q=ubuntu+chroot+howto&aq=0&oq=ubuntu+chroot+how [22:29] hansin, I was trying to setup jailed ssh but google is not seem to help me [22:31] I just installed Denyhosts on my server, and now I can ssh into my box. The /etc/hosts.deny is adding my IP -- how do I change that? [22:31] do I need to add myself to the /etc/hosts.allow? [22:52] kirkland: http://people.ubuntu.com/~mathiaz/jaunty-raid1-degraded-boot-fail.png <- is this the RAID1 failure you saw? [22:54] mathiaz: http://launchpadlibrarian.net/23151208/Screenshot.png [22:54] mathiaz: that's my error [22:54] mathiaz: which looks to be the same [22:54] mathiaz: https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/334994 [22:54] Launchpad bug 334994 in linux "Degraded RAID boot fails: kobject_add_internal failed for dev-sda1 with -EEXIST, don't try to register things with the same name in the same directory" [High,In progress] [22:55] * mathiaz nods [22:55] so ehrm... once it works. there will be a way for me to tell the machine to always start in degraded mode without me having to run around to the data centres and press buttons, right? [22:56] New bug: #341432 in krb5 (main) "package libkrb5-dev 1.6.dfsg.4~beta1-3 failed to install/upgrade: failed to delete `/usr/lib/libkrb5support.so.dpkg-tmp': Read-only file system" [Undecided,New] https://launchpad.net/bugs/341432 [22:56] Nafallo: yes, i added that in intrepid [22:56] Nafallo: and backported it to hardy [22:56] Nafallo: you can preseed the debconf question [22:56] aha! it should already work then :-P [22:56] Nafallo: or configure with 'dpkg-reconfigure mdadm' [22:57] kirkland: was just about to ask. thanks. [22:57] Nafallo: there's a question about 'do you want to boot your raid if it becomes degraded' [22:57] * Nafallo fixes [22:58] until I have money to buy an E200 and a hotswap kit :-) [23:07] <_dbd_l12> just had the hardest (well - most tries) install of ubuntu server 8.04 on an old machine (P III) with an AOpen DVD drive. The iso I md5summed and burned to CD kept getting random read errors during install... Tried the usual suspects (slow burning, different cd-types, bootflags (all_generic_id nodma etc.) but in the end i could only resolve it by burning the iso to a dvd instead and booting from dvd. The installation went flawless from [23:07] <_dbd_l12> there. Should i report a bug with details concerning hardware for this? [23:10] did you do md5sum on CD? [23:12] <_dbd_l12> i did the md5sum on the iso before burning (using a different machine). I still have the cd so i could still md5sum them... [23:12] try that [23:12] so you could eliminate 'burning errors' [23:13] <_dbd_l12> ok, i'll see what md5summing the disc itself gives [23:20] <_dbd_l12> damn, its seems like i have already thrown the used cd's away so it seems like i'll have to reburn / check if the problem persists and then md5sum the disc itself. if the md5 is ok i'll come back but that will prob. be tomorrow or in 2 days. thanks for the idea ivoks [23:21] np [23:21] good luck [23:33] kirkland: http://testcases.qa.ubuntu.com/Install/ServerRAID1 <- does that make sense? [23:33] kirkland: does it cover all of the RAID1 use case? [23:34] I'd suggest swap to be on RAID0 to be a more real world example but... *shrugs* [23:34] good enough for a test case I guess. [23:35] Nafallo: why raid0? if you loose a disk you may be screwed [23:35] Why would you put swap on a raid0? [23:36] mathiaz: how often do you need swap afterwards? it's better to keep that as fast as possible in case it ACTUALLY is needed. [23:36] mathiaz: raid0 is important for non-mission-critical data that you want to serve very fast [23:36] Surely two separate swap partitions would be better? [23:36] Save the raid overhead [23:36] mathiaz: my archive mirror, and iso mirror are on a RAID0 [23:36] maxb: not sure it's such a huge overhead? [23:36] mathiaz: if i lose a disk there, i gotta resync with the archive, but meh ... [23:36] mathiaz: i'd rather have the more space, and faster read/writes [23:37] kirkland: oh yes - I agree with this use case [23:37] kirkland: however if you setup a system with RAID1 it means you wanna protect yourself from disk failure [23:37] kirkland: if you want it to be fast you must cache your archive/mirror in RAM ;) [23:37] mathiaz: that description looks good [23:37] raid0 can't compete with memory performance, but gets hell a lot closer than raid1 :-P [23:37] kirkland: and by putting swap on raid0 you're not doing it correclty [23:37] mathiaz: to be exhaustive, you should try booting from each disk in the array [23:38] just create two swap partitions [23:38] raid0 for swap is pointless [23:38] swap sucks altogether, i don't use it at all [23:38] right - but raid1 for swap makes sense IMO [23:38] 8GB of ram is enough :-) [23:38] mathiaz: i agree with you [23:39] i use raid1 for swap [23:39] mathiaz: i have exhaustive test instructions at https://wiki.ubuntu.com/BootDegradedRaid#Implementation [23:46] kirkland: wiki page updated with a link to your instructions. Thanks for the review! [23:46] mathiaz: you bet; thanks for putting this together [23:47] mathiaz: i hate to see regressions in this :-)