[00:00] Quick question: my ssh session terminated when I was running aptitude upgrade on my server; it looks like it was in the beginning of reinstalling grub on the boot hard disks when the session died. [00:01] I'm reasonably confident it didn't actually begin to write any changes, so that's not a problem, but I've still got an aptitude process running and I do need to go back and tell it to run the installation. How do I do that? [00:16] trimeta, apt-get install -f [00:16] That will work even if there's apparently an aptitude process still running and holding the lockfile? [00:16] this will try to fix any package [00:16] And if not, is it safe to kill that process? [00:17] aptitude is still running? [00:17] Apparently; that's what pgrep -p aptitude says. [00:17] But I'm the only user logged in, and I'm not running it. [00:17] if it's stuck, i guess you have no choice but to kill it [00:17] OK, killed; now to run the command you suggested. [00:18] ...Lock file still used. [00:18] you have to remove the lock file [00:18] It's an empty file, so I just rm it? [00:18] yes [00:19] That got me "E: dpkg was interrupted, you must manually run 'sudo dpkg --configure -a' to correct the problem." [00:19] then do it [00:19] And when I ran that, I got "debconf: DbDriver "config": /var/cache/debconf/config.dat is locked by another process: Resource temporarily unavailable" [00:19] debconf? [00:19] try lsof | grep config.dat [00:19] Nothing. [00:20] sudo? [00:20] Yeah, that changed things: now I'm getting "frontend 7105 root 4rW REG 251,1 44378 655367 /var/cache/debconf/config.dat" [00:20] I have no idea what "frontend" is, though... [00:21] Aside from the fact that ps -l confirms that it's currently running. [00:21] you might want to kill this process [00:22] OK, now "sudo dpkg --configure -a" seems to be working. [00:22] It gave me the grub configuration screen again, and I proceeded as usual. [00:23] Running aptitude upgrade again produces no errors or unusual output. [00:23] I think this has probably worked...but sine it's about time I rebooted the server anyway, I might as well do so now and see if it actually installed properly. [00:24] hola, buenas noches [00:32] hello again. i am doing here the upgrade from interpid to jaunty but for some reason now the upgrade process hangs on depmod for a while and do nothing [00:32] http://paste-bin.com/view/7403fdd8 [00:33] its like it is waiting for something cpu usgae is very low and it does nothing do now since 10 to 15 minutes [00:33] really bad [00:33] hi [00:33] server at least is still availble and working [00:33] aragon_ hello [00:34] hello Translates expected my problem [00:35] I have problems with my lampp, I put different password security and the data base is not active [00:36] you mean mysql database ? [00:36] or what database ? [00:36] yes is lampp [00:36] mysql [00:36] lampp server [00:36] weird xperia [00:37] is disabled after giving my lampp security [00:37] aragon_ well do this here to know iw mysql first is working "sudo /etc/init.d/mysql status" [00:38] lenios: its just sit there and do nothing [00:38] now since 15 to 20 min [00:39] should i do a manul modprobe [00:39] command not found [00:39] he ? [00:39] thi must work [00:39] xperia, try something [00:40] lenios i am hearing [00:40] it must be stuck, not sure if modprobe can solve that [00:40] yeah i guess that too [00:40] ctrl + d / ctrl+c? [00:41] I have installed the qt4 prawns and ide [00:41] lenios looks like ssh got disconected hmm [00:41] Write failed: Broken pipe [00:41] i have still other ssh connection [00:42] alguien sabe español [00:42] what now looks like upgrade process stoped becouse of ssh connection closing [00:42] aragon_, try #ubuntu-es maybe [00:42] ok [00:43] xperia, you're not using screen? [00:43] thank [00:43] i have done it over ssh [00:43] you can recover it, then [00:43] yes ssh conection is no problem [00:43] i can reconect but how processing with upgrade further [00:44] last time i used "sudo do-release-upgrade" [00:44] it downloaded all new packages [00:45] and prelaced a lot of software with new version but then it stoped hmmmm [00:46] if you stop the process, you can run another do-release-upgrade [00:46] I had a machine reboot mid-upgrade for some reason [00:46] and was able to continue that way [00:47] qman:: okay will try again "sudo do-release-upgrade" hope just not that all packages will be redownloaded again [00:47] packages are cached in /var/cache/apt/archives [00:47] unless that directory has been wiped, it will not need to redownload anything [00:50] qman__ looks very good. it proccess again with the upgrade very fast. [00:51] it does every package from new check for reinstalling if it is old [00:55] New bug: #685028 in postfix (main) "package postfix 2.6.5-3 failed to install/upgrade:" [Undecided,New] https://launchpad.net/bugs/685028 [01:31] hey [01:32] how can i reduce the size of my ubuntu server installation by 230M ? [01:33] are there any huge packages i could remove? [01:34] or a way to show bigger packages? [01:39] I'm no expert (disclaimer here) but I believe you can install a minimal server with the alternative CD. Then just add the parts you want/need. [01:40] Anyone have any experience with NFS>? [01:40] could i just remove the 'locales' plugin or could this cause trouble [01:43] s/plugin/package/ - lool [01:49] How do you tell if nfsd is exporting a directory? [01:54] hi all... I'm running MM... I've install apache2, php5, mysql, and phpmyadmin... but I cant open localhost/phpmyadmin... any clue?? [02:04] New bug: #685014 in php5 (main) "Please merge php5 5.3.3-5 (main) from Debian unstable" [Wishlist,In progress] https://launchpad.net/bugs/685014 [02:51] is there a way to write all 0's to all unused space on my HD? [02:52] osmosis, zerofree :) [02:53] or the slow way, dd if=/dev/zero of=temp bs=1M === shennyg_ is now known as shennyg [02:56] patdk-lap, hmm...but my partition size is 500GB with 1.9GB used. qcow2 file is 80GB. creating the tmp file would fill up the whole 500GB, which seems a bit overboard. [02:57] you never said how you wanted to do it, you just asked how to do it :) [02:57] use zerofree [02:57] better question gets a better answer :) === shennyg_ is now known as shennyg === squishy is now known as SquishyNotHere [03:55] how do I view the changelog for a package before installing the update? [03:58] the last change? or all changes? [04:45] Hi [04:51] What would be best for logging the ip addresses of anyone who visits my server? [04:51] the website, to be exact? [04:55] your webserver :) [04:55] yes [04:55] I know, that is why I answered your question [04:57] patdk-lap: I mean, which piece of software would be best to run on the server to record ip addresses of visitors? [04:57] I don't plan on misusing this information or anything [04:57] just curious [04:57] your webserver [04:57] lol [04:57] dive deeper man! [04:57] your webserver is the best software to use to log the ip of peple using your website :) [04:58] patdk-lap: So it automatically does it? or are you trolling me? :p [04:58] you mean to parse the logs of ips and find out what people do? or something more? [04:58] cause your question to just log ip's, it does by default [04:59] nah, just record ip addresses of any connection to a file [04:59] "this ip conneted" [04:59] "this ip connected" [04:59] so on [04:59] what webserver do you use? [05:06] patdk-lap: ubuntu 10.10 [05:06] ok, so you use that os, but what webserver? :( [05:06] apache, lighttpd, ngix, .............. [05:07] i'm going to bet apache [05:07] $10 on apache! [05:07] dhttpd, thttpd, ...... [05:09] patdk-lap: apache [05:09] check the access logs in /var/log/apache2/ [05:10] ZacLnxNewb, $10 please :P [05:10] patdk-lap: spectacular. :D [05:10] arrrghhh: Psh, you probably could just look it up easily using who is or something [05:10] and my deluge torrent server won't save the port settings for some reason [05:11] haha i was just jokin [05:11] didn't do any pokin around tho. [05:11] i'd say apache is probably the most popular [05:11] apache is what is installed by default when a webserver is needed by something else [05:12] I use it as a print/web/sftp/file server [05:12] and at the moment, I'm sing my main page to auto-redirect traffic to the wikileaks ip address [05:13] lol [05:13] definitely lol [05:14] deluge web ui won't save the port settings [05:14] they don't stick [05:14] hrm. i use rtorrent... sorry. [05:14] [05:15] hm [05:16] it was a pain to learn, i'll admit it. but i freakin love it now that i took the time to learn it. [05:17] i used torrentflux before, and the webui felt very disconnected from the actual torrent applications running underneath... plus, 1 torrent 1 process is not cool. [05:17] arrrghhh: what do you think of my action towards wikileaks? [05:17] arrrghhh: yeah, multiple processes sounds like a pain in the ass [05:18] lol [05:18] I just accidently used the fword talking to my mother [05:18] classy [05:18] no comment dude, on all of it :P [05:20] arrrghhh: Well, I've been hanging out with the guys all night [05:20] she really doesn't care, it slipped and I generally don't use it [05:20] arrrghhh: no comment on wikileaks? that a negative connotation? [05:22] :p [05:23] eh you take it as you wish. [05:23] best part about no comment. you're not sure if i'm pleading the 5th or what ;) [05:24] I believe in openness and free speech. :D [05:24] I don't think it's an appropriate response that people are basically disabling their website [05:24] so [05:24] I'm trying to help. :D [06:57] hello [06:57] I am trying to use ubuntu 8.04 LTS [06:58] oh good [06:58] If I install gnome with apt-get install ubuntu-desktop , will there be any problem ? [06:59] becuase I have installed 8.04 LTS server in one machine [06:59] then installed gnome . some days everything was ok [07:00] later I am getting error that gnome is not automatically loaded [07:01] I have to give startx command to login into gnome desktop [07:01] I am getting error : Failed to initiaze HAL [07:02] after giving ok in that error box , I am logging into desktop [07:02] can anybody tell me , where is the problem ? [07:03] and what shall I do ? [07:04] 10.10 is the latest [07:04] hmm [07:04] btw good morning all [09:41] <[diablo]> morning [09:41] <[diablo]> anyone have any idea why the virtual kernel has no NFS support [09:56] Hi. Running Ubuntu-Server 10.04 (LTS) with Apache2/PHP5. PHP has suddenly stopped interpreting .php files and instead the source gets served by Apache. Believe server was running fine before last 'apt-get upgrade' (Yesterday), but can't be 100% sure since PHP isn't used everywhere in the server. PHP files are served with type: application/x-httpd-php, FastCGI of PHP fails also. Perl CGI works fine though. Anyone else experienc [09:56] ing similiar problems or have an idea? [09:58] Have made sure that libapache2-mod-php5 is installed and a2enmod'ed into Apache. Also Apache reports on start up that 'PHP' module is loaded. [10:00] Also have removed the entry from php5.conf also, tried running apt-get --reinstall on apache2, php module and php itself earlier as well. Just starting to run out of ideas now. [10:02] Actually nvm, not sure if I re-installed Apache itself. But PHP and the PHP modules at any rate. Apache itself is running just fine otherwise. [10:16] I'm getting no such device when I try to mount an NFS share. Any ideas? [10:28] Weyrling, check the checksums on your php binaries, or the package downloads in /var/cache/apt/archives [10:28] I've seen bad downloads cause this problem before [10:35] qman__: Thanks, will need to check that. Is there an easy way to do that or do I need to go hunting for the checksums, etc? [10:35] debsums? [10:35] well, the package's sum should be easy to find, but the binaries not so much [10:35] you'd have to find a known good system [10:48] Hmm, well the debsums program reports the Apache modules a-ok at least. I also cleared the cache and re-downloaded and re-installed php5, php5common, libapache2-mod-php5, libapache-mod-fastcgi but it doesn't seem to have helped. [10:55] Did the same for Apache itself as well now. It seems like that for some reason the PHP interpreter isn't invoked at all when a .php file is attempted accessed, but instead it just gets served as a normal file. The PHP executables themselves don't seem to error when I start them.. and the modules get loaded fine. It's as if there'd be something wrong with the settings but I hadn't touched them before the problem began. The PHP [10:55] module loads up fine and mod_fastcgi also... They just don't catch on, even though the file type seems set correctly too (application/x-http-php or so). [11:03] Ah, rewriting into the mod_php5 specific AddType/AddHandler values seems to have enabled it partially for the site now at least. Will see if it helps for the rest of it too.. [11:18] Well buh, something in the updates apparently had made it so that I had to do an explicit +ExecCGI instead of just ExecCGI on the FastCGI handler. Well, fine enough for me though, site seems to be working again. [11:42] Alright, good luck everyone else then, I'm off. [13:18] I am trying to set up ufw but it seems to block all connections and will not log anyting. I am sure I have got the rules set up right. ufw status works show 22/tcp alow in anywhere and I have set logging to full yet nothing is hapening. as soon as i disable the firewall i can ftp and ssh again (i have done a rule for ftp). I realy have no idear what is whrong. [13:23] hi guys any sugested firewall application for linux..? which is GUI- i can install directly. [13:25] ruben23: just use ufw if you want a simple firewall [13:26] how about application [13:27] ruben23: just learn ufw [13:28] I doubt it will take longer than 2 minutes to set it up [13:30] Cromulent: unless you are me, do you think you could help with my problem posted earlyer? [13:32] paddy__: what does "sudo ufw status verbose" tell you? [13:35] paddy__: can you telnet to the ports [13:35] that it is active, logging is on full, defaults are deny in allow out [13:35] and rules are as previously stated [13:35] what does new profiles: skip mean? [13:36] Cromulent: ikonia ^ [13:36] paddy__: paste the output [13:36] here or bin? [13:36] paddy__: deny out ? [13:36] bin [13:36] paddy__: you know your appllications need to respond too [13:36] allow out [13:36] deny in ? [13:37] ikonia: yes [13:37] paddy__: you need "allow in" to hit the port [13:37] paddy__: to ssh in - that is an incoming connection [13:38] ikonia: the defaults are deny with alow in for 22 and 21 [13:38] paddy__: your output should look something like this http://pastebin.com/RUUDyumr [13:38] paddy__: you know it's iptables, so the allow needs to be before the deny [13:39] ikonia: the output is the same [13:40] Cromulent: ^ [13:40] paddy__: output is the same as what ? [13:41] paddy__: can you telnet to port 22 ? [13:41] ikonia: it is the same is Cromulents output [13:41] paddy__: can you telnet to port 22 ? [13:41] ikonia: from local or remote? [13:41] paddy__: try both, see if they differ [13:41] they shouldn't [13:43] connecting to server failed [13:44] what server should i telnet to from server? [13:44] what ? [13:44] paddy__: show me the commands you're using, exactly [13:44] the bad server is remote [13:45] paddy__: show me the exact commands you're using [13:45] ikonia: cant go out [13:45] paddy__: show me the exact commands you're using [13:45] firewall is blocking everything [13:45] paddy__: show me the exact commands you're using [13:45] telnet google 80 [13:45] and that's being blocked ? [13:45] works on my machine [13:46] but not on server [13:46] looks abit incomplete to me [13:46] paddy__: on the server, to "iptables -L" and pastebin it please [13:47] fatal module ip_tables not found [13:47] ikonia: i guess thats bad [13:47] paddy__: you need to use sudo [13:47] errrr not sure how you have a working firewall [13:47] ahh yes [13:47] sorry, I take that for granted, sudo iptables -L [13:49] http://pastebin.com/Y8UzrFDt [13:51] paddy__: can you actually paste the output of sudo ufw status verbose too? because the output you just posted does not have any allow rules [13:51] paddy__: and if you stop the firewall, you can telnet to google.com on port 80 ? [13:53] http://pastebin.com/0E4JePeZ [13:53] ikonia: tenet works if i disable ufw [13:54] what was that last paste you sent me [13:54] starting ufw i just got ERROR: problem running ufw-init [13:54] http://pastebin.com/0E4JePeZ is status verbose [13:55] paddy__: ok, so that is the firewall active, but you can't do anything with it out going ? [13:55] paddy__: can you do: 'sudo ufw disable ; sudo /lib/ufw/ufw-init start' and paste the output? [13:56] Skip starting firewall: ufw (not enabled) [13:56] oh duh [13:56] paddy__: can you do: sudo ufw enable ; sudo /lib/ufw/ufw-init stop ; /lib/ufw/ufw-init start' and paste the output? [13:58] how do i pipe all of that to a file jdstrand [13:59] paddy__: I really only need the last command, so just add '> /tmp/out' to the end. [14:00] load of errors in prev commands [14:00] last one is problem running /etc/ufw/before.rules [14:00] problem running /etc/ufw after.rules [14:01] problem running 'lib/ufw/user.rules [14:02] paddy__: can you put "sudo sh -c '/lib/ufw/ufw-init stop ; /lib/ufw/ufw-init start' > /tmp/out" into a pastebin? [14:02] paddy__: well, the contents of /tmp/out that is [14:02] meh [14:02] you probably need 2>&1 at the end of that [14:04] still same [14:05] paddy__: can you paste the output? [14:08] jdstrand: 2> is http://pastebin.com/QQguzdm1 [14:08] 2>&1 prints everything to scren [14:10] paddy__: can you run: 'sudo /usr/share/ufw/check-requirements' and paste the output [14:13] http://pastebin.com/q3bG1as8 [14:14] irssi just saved me, i almost pasted the whole thing here instaed of the url, whoops [14:14] paddy__: that is not the output of check-requirements [14:16] http://pastebin.com/u0J8SFAB [14:16] paddy__: your kernel doesn't have the required configuration to run ufw [14:17] jdstrand: how fix? [14:17] paddy__: recompile your kernel or use a standard Ubuntu kernel [14:18] how do i change to a standard ubuntu kernel? [14:18] paddy__: do you own this machine or is it a hosted environment? [14:18] i am on a vps [14:19] jdstrand: but I have serial acces [14:19] jdstrand: which is why when i killed ssh connections i was ok [14:19] paddy__: you need to talk to your VPS and show them the output of the check-requirements command. it is possible they will tell you 'no' at which point you must use standard iptables with simple rules [14:19] *access [14:20] jdstrand: can I not change the kernel? [14:20] paddy__: if you must go to standard iptables, please use 'sudo ufw disable' first [14:20] paddy__: I don't know. you need to talk to your VPS [14:20] hi guys where i can find the iptables file of ubuntu server..? [14:20] jdstrand: is there no easy way to do it? [14:21] paddy__: you have to be able to reboot the machine and select a kernel. the VPS provides the kernel. you need to talk to them [14:21] jdstrand: I can do that, in grub? [14:23] paddy__: possibly. but aiui most VPS control the kernels in use for security reasons [14:24] paddy__: what kind of virtualization is being used? [14:25] guy can i manipulate directl from iptables on ubntu server..? [14:25] paddy__: if it's OS-level virtualization (e. g. openvz, virtuozzo, linux-vserver, ...) you can't change the kernel or load custom modules [14:49] joschi: openvz [14:49] I cancled my vps [14:50] whoa [14:50] that's a big step to take [14:55] ikonia: it was only a toy ;) [15:01] hello to all. somehow i am not able to do "sudo do-release-upgrade" it breaks allways with error message [15:01] /usr/bin/dpkg returned a error code [15:01] Your system is unusable ? can somebody help me with this ? [15:03] xperia: was it up to date before you did do-release-upgrade [15:03] xperia: what process have you followed for the upgrade [15:05] ikonia: yes it was full up to date and it worked 100% but i have did a system clean with the ubucleaner script that remove some stuff from the server like not used linux kernels and got then before the upgrade allways the error message that smething with the linux kernel thing is not right when i performed a sudo apt-get upgrade [15:06] New bug: #685201 in squid (main) "shutdown_lifetime setting behavior does not match doc" [Undecided,New] https://launchpad.net/bugs/685201 [15:06] and now the release upgrade also breaks at same problem it say [15:06] ubucleaner ? where di you get that [15:06] "did" [15:06] from the net it is like ccleaner but for ubuntu. [15:07] that's not an ubuntu package, and you don't know what it does [15:07] why did you not just clean the machine yourself [15:07] make a search on google ubucleaner think i got it from opendesktop [15:07] I have no idea why you would run untrusted scripts like that [15:07] well needed somethign automatic and did not know that something like this exist in ubuntu [15:08] "needed" ? what makes you think y ou needed to "clean" you ubuntu machine [15:09] well the release upgrade breaks becouse of ubuntu-linux-modules-2.6.24-19-server [15:09] breaks in what way [15:09] looks like package manager has problems like before the upgrade [15:10] about this modules that were probably removed [15:10] the package manager had a problem before the upgrade ? [15:10] so you knew the package manager had a problem, and you still did a distribution upgrade ? [15:11] ikonia yes it right it complained about the same thing. thinked with a upgrade it will solve the problem but it does not. [15:11] needed free space on the disk [15:11] how can i fix this package manager problem becouse of the deleted modules probably ? [15:11] sorry, that's just a crazy approach, the core component that controls package installation is having a problem, I'll use the problematic package to upgrade it's self [15:12] xperia: re-install the package that deleted the modules [15:12] that contained the deleted modules [15:12] so i would need to install the linux server kernel 2.6.24-19 probably ? [15:13] okay let me looks if i can find it [15:13] if that contains the modules you deleted, then yes [15:31] ikonia_ i get this error here [15:31] FATAL: Could not open '/boot/System.map-2.6.24-19-server': No such file or directory [15:31] what does this mean ? this is my file list of /boot http://paste-bin.com/view/db4b29e5 [15:32] xperia: it means the file is not there [15:32] any possibility to fix that easy so dpkg dont search anymore for it [15:42] xperia: remove the package, and re-add it [15:44] ikonia: just have fixed the problem now. needed only to create this missing file with touch and now everything works like it should. thank you a lot for your very friendfull help [15:44] http://ubuntuforums.org/showthread.php?t=831658 [15:44] xperia: well done [15:44] nice trick on dpkg === SquishyNotHere is now known as squishy [17:00] http://pastebin.com/tYpKYbuY [17:02] RoyK: what do you want me to do with that ? [17:11] * RoyK is just grinning [17:19] ikonia: I expect he's bragging about the available space [17:20] oh, a pointless post [17:38] with kvm, im getting the error "Could not initialize SDL - exiting". What does that mean? [18:01] how can I go about putting nfs in fstab, without having mountall hang on boot when the network cable isn't plugged in? [18:24] patdk-lap: autofs? [18:27] noauto? [18:28] there's also an additional option for nfs in fstab, iirc [18:28] _netdev [18:29] (used to prevent the system from attempting to mount these [18:29] filesystems until the network has been enabled on the system). [18:29] royk, no, autofs isn't a solution [18:29] Hi Guys anyone know how I would install Mcrypt for php? [18:30] garymc: apt-get install php5-mcrypt [18:30] ivoks, but _netdev isn't a flag I'm suppost to give it, it's something that mount does internally for network filesystems [18:30] ivoks and is that it? [18:30] garymc: yes [18:30] do i need to configure anything? [18:31] nope [18:31] cool. My php programmer has asked for it. Whats it for? [18:31] patdk-lap: that option is explicity for that as far as i understand [18:31] I can't find _netdev in the man pages [18:31] man mount [18:32] garymc: http://www.php.net/manual/en/intro.mcrypt.php [18:33] ivoks that didnt seem to work [18:34] garymc: it works [18:34] garymc: but your question wasn't complete [18:34] i typed "apt-get install php5-mcrypt and it didnt install anything [18:34] garymc: it should've been 'what do i have to do to enable mcrypt module in apache's mod-php' [18:34] oh.....what do i have to do to enable mcrypt module in apache's mod-php [18:34] then it's already installed [18:34] ? [18:35] how do I know its already installed? [18:35] dpkg -l | grep php5-mcrypt [18:35] Ahh it says so [18:35] ivoks, still hangs at mountall with _netdev [18:35] when you install php plugin [18:36] ok so do I need to enable it or something? [18:36] it's available for php, but you have to remember that apache needs to reload the php module to pick up all the new plugins [18:36] so, restart apache [18:36] patdk-lap: are you mounting /home maybe? [18:36] nope [18:36] I'm mounting /var/www/virtual [18:38] then that's a bug [18:38] using lucid 10.04 installed today :) all updated [18:39] still a bug [18:39] it should obey _netdev [18:41] _netdev is irrelevant on cifs and nfs shares. And mountall understands _netdev. [18:41] https://bugs.launchpad.net/ubuntu/+source/mountall/+bug/504224 [18:41] Launchpad bug 504224 in mountall "NFS mounts at boot time prevent boot or print spurious errors" [Medium,Fix released] [18:42] Note that you can get around the recovery shell by using the "nobootwait" option it /etc/fstab for /home. [18:43] I tried nobootwait, optional, bootwait, _netdev [18:43] none fixed it [18:43] and this is for /var/www/virtual, not /home [18:44] ask in #upstart :) [18:45] heh :) [18:46] note that it's saturday [18:47] but there are 23 users :) [18:47] and I did read that bug report and comments like 3 times before I asked :) [19:04] ive got a question about setting up raid and lvm :) [19:04] ive got a 2TB drive and a 1TB drive. can i parition the 2TB into 1TB sections and use raid5? [19:12] resno: I'm not that into raid, but it would miss the point [19:13] you can almost do anything with software raid [19:13] (I could be wrong however) [19:16] cokegen: i know its mising the point, but i dont really want to backup stuff, i just want to protect my data from a failed drive when using lvm [19:22] resno, that won't protect anything :) [19:22] if the 1tb goes down, your protected [19:23] if the 2tb goes down, your screwed [19:23] so using the 1tb is pointless, and using raid5 is just going slow down the 2tb drive [19:23] and cause the 2tb drive to seek like nuts [19:24] hi , anyone can private help ? [19:24] patdk-lap: so whats the solution to my problem? [19:24] don't do it :) [19:25] what should i do? [19:25] make a 1tb partition on the 2tb drive, use raid1 on that 1tb drive and the 1tb partition on the 2tb drive [19:25] you have redundant 1tb now [19:25] and you can do whatever yo uwant with the extra 1tb on the 2tb drive [19:25] i dont really care about data redundancy though [19:26] i need the space more so then the redundancy [19:26] " i just want to protect my data from a failed drive when using lvm" [19:26] that is redundancy [19:26] the whole thing of having 2 drives that could fail and ruin it all is scaring me [19:27] well, everything you said, will loose all your data if either fail [19:27] well, the 2tb is ok, you would be ok [19:27] but if the 2tb fails, you loose all [19:27] and you don't gain any space [19:28] so it's moot [19:28] i wouldnt gain the extra space on that? [19:28] on what? [19:28] how does using raid5 over 3 1tb drives (2tb usable) give you more space than 2tb drive? [19:28] paritioning each drive to 1TB and raiding it? [19:28] oh ok [19:29] your solution is to use each drive as it's own drive [19:29] if you want all space, and don't caare about redundancy [19:29] then whatever drive fails you loose, but not the other [19:30] so not using lvm? [19:30] not sure [19:30] I normally just put lvm on each drive [19:30] I think if you lvm both drives together into one lvm, it makes a raid0 out of it, givin you the issue again [19:30] but I have never done that, so [19:31] so lets just say i dump the lvm [19:31] and install each drive alone... [19:31] how do i get data to start using it? [19:31] get data? [19:31] wget? [19:32] how i instruct the server to use that space [19:32] fstab [19:32] mount [19:32] if you ask me, I'd prefeer doing 2 single backups on those different disks [19:32] anything you normally would do with a drive [19:32] ya, I would do a raid1 if 1tb over them [19:32] ive never done anything like this... so this is all new [19:33] then use the 1tb unprotected for junk [19:33] you hav enever installed a harddrive? usb stick? cdrom? :) [19:33] never did software raid [19:33] never did software raid [19:33] we aren't talking about software raid at all [19:33] and this is my first time messing with adding anew drive since the fresh install [19:33] cause everytime I give you the options, you say, no [19:34] if I were you I'd just backup things twice [19:34] * patdk-lap wonders back to ospf land [19:34] ive already just about tapped out the 1TB drive i have running. so im looking to increase my data storage [19:35] he wants what every mythtv user wants, but I haven't found one that is reliable [19:35] seperate harddrives with seperate fs's on them [19:36] but mount them as one large drive [19:36] drbd is another option [19:36] no raid, if one drive dies, you loose that info, but not everything [19:36] yes, thats exactly what i want... [19:36] sorry, drdb would be: http://www.drbd.org/ [19:36] have some friends using it on production servers and work like a charm [19:36] drbd is more data redundancy [19:37] yeah, on another machine, I know [19:37] resno said he didn't want redundancy [19:37] zfs over fuse ? [19:37] if you want to trade off some performance [19:37] never thought zfs was reliable [19:38] it is, just that you loose a good deal of performance [19:38] there are like 3 different fuse things to do this, each of them failed horrible for me [19:38] im looking for a solution to maximize the space on the new drive and previous one [19:38] just format and mount them, done :) [19:38] resno: mind that there's no perfect solution to every problem [19:39] cokegen: ive learned that [19:39] why not 2 copies with rsync over the 2 disks ? [19:39] separately [19:40] I mean, if you just need backups [19:40] if you need to actively use that space it wont work [19:41] rsync? backups? [19:41] well, he isn't 100% clear on what he want [19:41] he wants to have 3tb of drive space [19:41] he doesn't want backups [19:41] just giving him some ideas [19:42] he wants as many files as possible to be working, if one fails [19:42] hmmmm [19:42] i like lvm, but dont like the danger it presents [19:42] lvm in itself is not a danger [19:43] but trying to have 3 tb on those 2 disks ONLY is [19:43] just taht [19:43] want security, use zfs, or raid 1 [19:43] (some level of security) [19:43] the mythtv people worked around it by just mountingeach drive as it's own drive [19:43] think on a failing power source too as a point of failure [19:44] then having the myth software write to whatever drive has the most free space [19:44] is what I told him [19:44] just mount each drive separately [19:44] patdk-lap: ill look at what myth pepole setup then [19:45] no FS will save you if the power source fails real bad (and it happens) [19:45] you guys server admins? [19:45] I kinda am [19:46] that's why I'm telling you, there's always more than one way to do things [19:46] in that department at least [19:47] yea, lvm seems like a great solution... [19:47] I use it [19:47] it's cool to grow [19:47] but i dont like the concept of any one driving failing and everything crashing [19:47] with reiserfs you can even grow your fs on the fly without unmounting [19:48] can't you have another machine and go backing up once a day ? [19:48] I found THAT to be the most reliable solution on small setups, really [19:49] that way you only use one disk and the cost is low [19:49] in the worst case you loose a day of work or less [19:49] if raid fails you need to recover [19:49] anyone familiar with linux vps + ubuntu ? [19:49] it's a pain in the ass [19:50] cokegen: i do have another machine i can back up to but it doesnt have that much space [19:50] add there another disk [19:50] problem solved [19:50] you can't have security without spending some money [19:50] not with only 2 disks [19:50] heh? everyhing I think supports growing the fs on the fly while mounted, ext2, ext3, ext4, xfs [19:51] no, ext3 doesn't [19:51] you have to unmount first [19:51] not sure ext3 [19:51] and xfs [19:51] I grow ext3 all the time mounted [19:51] I'm 99.5% sure you can't do that [19:52] I'll google it [19:52] that's one of the good points of reiser btw [19:52] anyone can help with linux vps + ubuntu karmic [19:52] man resize2fs [19:52] "The resize2fs program will resize ext2, ext3, or ext4 file systems. It can be used to enlarge or shrink an unmounted file system [19:52] located on device. If the filesystem is mounted, it can be used to expand the size of the mounted filesystem, assuming the kernel supports on-line resizing." [19:53] been using it for like 4 years now [19:54] I don't know if it's safe [19:54] karmic? building on a version that's only supported for another few months? [19:54] but it's a good thing to know if you say it's working [19:58] I read somewhere that online resizing with extX was not safe [19:58] did you happen to notice the date it said that? [19:58] but it seems that it works [19:58] joe: what kinda help ? [19:58] I know years and years ago, it wasn't [19:59] if I do sudo rm file , then realize a made a mistake, ..is there a way to undo? [19:59] yeah, probably years ago [19:59] 2 to 3 years [19:59] osmosis: i think its gone when you erase it [19:59] again, good thing to know [20:01] patdk-lap: http://tldp.org/HOWTO/LVM-HOWTO/extendlv.html [20:01] that's were I read it :P [20:01] old docs [20:01] and it was dangerous on etch which was the distro I was using at the time [20:02] ya, that was pre-2006 [20:02] in 2006 resize2fs online resizing was added [20:03] osmosis: unmount as that partition as quick as you can and try some utils in the system rescue cd @ http://www.sysresccd.org/Main_Page [20:03] in reality, I'm going back to linux [20:03] so there are things I need to "refresh" [20:05] was tired with the lack of hardware support on the desktop and well, the other day tried a maverick livecd and everything was working [20:05] seems the perfect time for a comeback [20:05] heh, I've been using ubuntu since feisty [20:05] and using slackware since '95 [20:05] I was started in '95 too [20:06] with an infomagic cd set [20:06] I have one slackware system still running [20:06] had slackware, debian and red hat plus a mirror of some ftp's [20:06] and in 2000 or 2001 was sick of always having to do stuff to normally work [20:07] so I went to windows on the desktop, never for servers, but to work I only needed putty and that was it [20:07] and well, the ubuntu pple did a great job polishing stuff [20:07] is really almost there [20:08] the only thing that kept me off unix for desktop, was the ungodly amount of screen space X used [20:08] the flash player is the only thing that I can't make it work 100% ok [20:08] use noflash :) [20:08] yes, I know [20:08] it's more of a security concern these days [20:09] anyway, when I go to youtube it doesn't work ok [20:09] fullscreen hangs, and other problems [20:09] anyway, it's a proprietary piece of shit [20:09] I'll not blame linux this time [20:10] adobe should work on an open source player ... it's good for everyone, and no one but them will be able to reproduce something like flash to work on [20:11] but it will open the door to proper implementations on different OS's and that kind of stuff [20:11] sucks always having that one little problem that's too big to do the switch [20:11] I do web development so I care about flash ... have to [20:13] did you try with a different browser? flash was unusable in firefox for me, but in chromium it's ~98% [20:13] stutters sometimes but watching videos works fine [20:15] have to test chromium [20:15] is that I just installed maverick on my other machine [20:15] was trying everything on a liveusb with persistence [20:15] to actually see if it was worth the shot [20:15] worth the try [20:16] (my english is not that good, sorry ;-) [20:17] I really like chromium/chrome but there's nothing like firefox/firebug to do web development [20:18] neither opera nor chrome are good in that regard [20:20] Evening all , I need 2 dhcp servers on the same subnet , which are physically 2 networks connected by a "TUN" . Is there a comprehensive document somewhere i can read through so i don't overlook anything . A dhcp on either side of the tun needs to assign ip's from diffrerent pools to machines on their respective sides. I would need to block packets with requests , to pass over the tun i would think , probably more to it .. Atm i have it set up funct [20:22] i said tun didn't I .. i meant TAP [20:22] Dibbler__: what about just blocking 68 from trespassing the tun ? [20:22] bc each dhcp should reply only requests from each side [20:23] so, I say just do some iptables magic and block 68 (or whatever port dhcp is running, can't remember) from going from one side to the other [20:23] Dibbler__: sounds like it would be easier to use subnetting and routing. are there a reason that's not an option? [20:23] thats how it is running now [20:24] of course, if the machines are in the same subnet that is bc the machines have to work together [20:24] subnets have nothing to do with reachability [20:24] no [20:24] they are communicating fine :) [20:24] Dibbler__, what about blocking 68 from one side to the other ? [20:24] it's flippin windows [20:25] i don't want to change all the firewall rules on all the computers to include other subnets for filesharing [20:25] Dibbler__: so... if it's working now, I have to ask what you're trying to achieve? [20:25] you don't have to [20:25] like i said [20:25] oh. [20:25] you have to just block 68 from one side to the other [20:25] is that possible ? [20:26] the fact it's reachable and routing nicely doen't mean disocovery works etc [20:26] it's not , [20:26] i cant block traffic from one side to the other [20:26] well i could [20:26] but that would only block replies [20:27] i can block 68 .. but to where [20:27] they don't have an ip atthat point [20:27] Dibbler__: block on the interface level [20:27] not ip [20:27] what links that "TUN" ? [20:27] i could block outgoing 68 yes [20:27] 2 linux boxes ? [20:28] yes 2 ununtu servers [20:28] ubuntu [20:28] well, do an iptables rule to block 68 from one side to the other [20:28] problem solved [20:28] i could block 68 from both mac's to their respective other sides [20:28] you have one dhcp server on each side [20:28] well side would mean machine in this case [20:28] since there are no "sides" [20:29] dince it's tun [20:29] one or more machines [20:29] and therefore bridged [20:29] and 1 interface [20:29] aha [20:29] you're right [20:30] so on some level i would need to deny packets .. lvl 3 with a specific header from crossing over [20:30] that's not firewall stuff [20:31] thats above [20:31] can't you configure the dchp server on each side to assign a different pool range ? [20:31] yes [20:31] that way you can completely block 68 from each side and no ip address would overlap [20:32] oh yes i can [20:32] machines still communicating etc etc [20:32] but imagine a machine with no fixed address [20:32] being connected [20:32] why would any of the dhcp's think it's his to give an ip to [20:33] don't understand ... [20:33] take a switch , connect 2 dhcp servers [20:33] connect a client [20:33] which one will give the ip [20:33] theres no difference [20:33] when using tap [20:33] you have two ubuntu machines with tap or tun interfaces [20:34] i want to use tap [20:34] you simply need to block packets on 67:68 from going through those interfaces [20:34] i am using tun [20:34] you need some physical separation [20:34] there is only 1 interface [20:34] vlan [20:34] the bridged one [20:34] iptables -A OUTPUT -o tun0 --sport 67:68 -j DROP [20:34] something like that [20:35] thee is no traffic [20:35] there [20:35] but if you connnect everything to a switch then there are no "sides" [20:35] exactly [20:35] the tap is a bridge [20:35] there are no sides [20:35] Dibbler__, you apply the rule to your bridging interfaec [20:35] it still works [20:36] you just create the rule based on the interface, not the IP [20:36] but output from where [20:36] the input and output are both tap [20:36] it doesn't matter [20:36] there is no firewall involvement [20:36] yes, there is [20:36] it is an interface, the firewall applies [20:37] it doesn't matter what kind of magic it's performing to bridge the networks, there are still two tap interfaces [20:37] if you put a 68 67 frop on eth0 [20:37] and you can still filter traffic on those tap interfaces [20:37] what is going to prevent 2 computers from using it [20:37] no [20:37] you don't use eth0 [20:37] you can't [20:37] you use tap0 [20:37] yes, you can [20:38] I use a tap interface with openvpn, the firewall still applies [20:38] there is no difference between it being eth0 or tap [20:38] i would need to filter on level 3 [20:38] the actual requests [20:38] it's not as complicated as you're making it [20:38] why not redirect all traffic through separate interfaces ? [20:38] then i would need diff subnets [20:38] and thats how it works now [20:39] just because the networks are bridged doesn't mean the interfaces go away [20:39] you can still filter traffic [20:39] no, bc you can block just dhcp [20:39] I'm having trouble with a borked package installed from a ppa. aptitude purge problem package list is intended to remove data and configure files as well as uninstall, right? [20:39] there is only 1 interface qman [20:39] rules need a source and a dest [20:39] no, they don't [20:40] you can simply drop all traffic from anywhere to anywhere on the DHCP ports [20:40] how not qman__ ? [20:40] on your VPN interface [20:40] see [20:40] anywhere IS a source or a dest [20:40] anywhere to anywhere [20:40] and that solves the problem [20:40] there is no to [20:40] it doesn't matter whether you're using bridging or tunneling, it'll work [20:41] if you make a bridge there are no filters between the interfaces on an ip level [20:42] enquora, purge removes all configuration files as well as the package [20:42] Dibbler__, yes there are [20:42] Dibbler__ is right [20:42] just because it's not routed doesn't mean you can't filter it [20:42] a bridge is like connecting those 2 int to a switch [20:42] directly [20:42] physical separation is not possible [20:42] the traffic still passes through the interface, and netfilter still processes it [20:43] netfilter does ipx now does it .. [20:43] who said anything about ipx? [20:43] me [20:43] just now [20:43] becaus ethat will pass on a bridge [20:45] wait. what are you trying to achieve again? [20:46] 1 subnet on 2 physical locatiuons connected through a tap [20:46] i need a dhcp on both sides [20:46] even if netfilter doesn't process the ipx traffic, the traffic is still going through an interface [20:46] and netfilter can still filter it [20:46] and why do you need to block dhcp at all? [20:46] I have a configuration this way [20:46] it works [20:46] just make each dhcp server serve half the subnet? [20:47] yeah, that's what I said Nafallo [20:47] with dynamic ip's .. how would i prevent one or the other from giving it [20:47] the thing is that the interface is bridged and he has no "sides" [20:47] Dibbler__: why do you need to prevent it? [20:47] discover will be sent [20:47] there are still two sides to a bridge [20:47] to have a "side" he'll need physical separation [20:47] there is no ip at that point [20:47] both dhcp wil reply [20:48] just let whatever dhcp answers faster give the ip? [20:48] then what if the tap disconnects [20:48] ip doesn't mean that if you connect that to a switch the traffic isn's available [20:48] to me it seems you need physical separation [20:48] like a vlan or something like that [20:49] Dibbler__: you're on the same subnet... why does it matter? [20:49] or just connect either side to different network interfaces [20:49] but yes, Nafallo is right, you're trying to do that on the same subnet [20:49] there's no point on that [20:49] the client won't throw away the IP just because it can't reach the DHCP server. at least not until after the dhcp timeout. [20:49] N: it matters becaus ethey are in different locations [20:50] but i supose i could just do with the 1 thne [20:50] then [20:50] and hope for the best [20:50] the only reason it would matter is the default gateway setting [20:50] yes that too [20:50] when the bridge dips those with addresses from the other side will not be able to reach the internet [20:50] yes [20:50] they both have different def gateways [20:51] so yes [20:51] thats why i need 2 lol [20:51] and thats why i asked [20:51] meh. no one said anything about reaching the Internet :-P [20:51] but like I said, a simple firewall rule blocking DHCP solves this [20:51] it's not that complicated [20:51] believe me Qman [20:51] you can't block a bridge if it goes nowhere but on both sides of it [20:52] there's no such thing as a bridge that goes nowhere [20:52] Dibbler__: qman__ didn't tell you to block the bridge... [20:52] you are bridging two networks together [20:52] it goes from one side of the bridge to the other [20:52] and you can't filter trafic on that [20:52] on an ip level [20:52] yes, you can, because there are still two tap interfaces [20:53] yes, you can [20:53] ignore the bridge for a bit. it's just an interface on top of other interfaces. [20:53] what about a virtual interface to help a bit [20:53] taps are virtual interfaces [20:53] not needed [20:53] block the traffic and then do the tap to the other [20:53] also, you don't want to block it on an ip level. you want to block it on an interface level. (phy) [20:53] right [20:53] just block all DHCP traffic in and out of the tap interfaces [20:54] stop digging up the stack :-P [20:54] it's a virtual interface, but it's still an interface [20:54] and you can still create rules that use it [20:54] but there is nothing going in and out the tap interfaces [20:54] yes, there is [20:54] wait sec [20:54] 1 sec [20:54] all traffic that goes between the two sides of the bridge goes through the tap interface [20:55] http://www.shorewall.net/OPENVPN.html#id36132361 [20:55] so everything is clear [20:55] this is what im talking about [20:55] bottom bit , bridging 2 networks [20:56] yes [20:56] i get 1 bridge on each firewall [20:56] this is also what I'm talking about [20:56] my server is handing php a wrong timezone of 01/01/1970 anyone know how I fix this? [20:56] that is a bridge of the phys interface and the tap interface [20:56] the way openVPN works is by creating virtual interfaces [20:56] all traffic that goes across the bridge goes through those virtual interfaces [20:56] and you can apply firewall rules to those interfaces [20:57] br0 [20:57] on both firewalls [20:57] have a look [20:57] garymc: php.ini [20:57] give me a sec [20:57] yeah but what have I got to set? [20:58] then you simply block it on br0 as well [20:58] it's not hard [20:58] ive uncommented date.timezone = [20:58] and added UTC on the end [20:58] quote "The Shorewall configuration is just a Simple Bridge." [20:59] but its not fixing it [20:59] you're digging too deep [20:59] LOL [20:59] have you even tried to set the rules? [20:59] you're confusing theory and simplified explanations with how it actually works [20:59] at the moment it's 2 subnets [21:00] seriously i'm not [21:00] phpinfo() says that php.ini is actually loaded ? [21:00] it's not magic, all traffic passes through interfaces [21:00] http://www.php.net/manual/en/datetime.configuration.php#ini.date.timezone [21:00] and all interfaces can have rules applied [21:00] it doesn't [21:00] that's how it works [21:00] no [21:00] yes [21:00] cokegen : yes [21:00] you can only filter level 5 with a firewall [21:01] bridges are level 3 [21:01] HOw can I test my time settings are correct [21:01] I have date.timezone = UTC [21:01] and it HAS to work [21:01] there is no ip level traffic [21:01] believe me [21:01] New bug: #685284 in linux-meta (main) "Deadlock in n_tty_read()" [Undecided,New] https://launchpad.net/bugs/685284 [21:02] yes i have added mine to that. It was commented out. [21:02] Once I change I ran /etc/init.d/apache2 restart [21:02] if there are no ip traffic, what is it then? [21:02] not working [21:02] are you trying to do dhcp over something else than ip? [21:02] your dhcpdiscover is not ip level [21:03] garymc: phpinfo() should say something like Loaded Configuration File => /srv/php/etc/php.ini [21:03] with the php.ini at the end [21:04] also, comment other date.something stuff in php.ini [21:05] and of course restart apache or whatever webserver you have (if you actually have one) [21:06] Dibbler__, dhcp happens on IP [21:06] yes [21:06] I can't explain it any clearer [21:06] yes it does [21:06] all traffic passes through interfaces [21:06] source address 0.0.0.0 [21:06] firewall rules apply to interfaces [21:06] http://en.wikipedia.org/wiki/Dhcp#DHCP_discovery [21:06] dest 255.255.255.255 [21:07] i know [21:07] your VPN/bridge/whatever does not change that fact [21:07] no no it doesnt [21:07] you just need to find which interface your traffic is passing through, and filter it [21:07] and you can't filter traffic between branches of a bridge [21:07] yes, you can [21:07] on an ip level [21:07] you can filter on every interface that exists [21:07] even if no routing occurs [21:08] php 5.2.10 [21:08] even if they don't have IPs assigned [21:09] i think i found what i need [21:09] ebtables does layer 3 filtering [21:09] ill have a look at that [21:09] thank you for trying guys [21:10] garymc: worked ? [21:10] Dibbler__: you are telling me iptables -A INPUT -i tap0 -j DROP wouldn't cause any disruption in your tunnel traffic? [21:10] br0 [21:10] in this case [21:10] not tap [21:10] Dibbler__: br0 is on top of tap0 and ethX. [21:11] yes [21:11] Dibbler__: you are climbing to far in teh OSI stack again. [21:12] i will try [21:14] but once you make a bridge [21:14] your tap and ethX iterfaces get no Ip velel address or anything [21:14] really guys [21:14] just stop thinking about the damn bridge for a bit :-P [21:14] :-) [21:14] Nafallo: oh hai [21:15] hi lifeless :-) [21:15] once it's there [21:15] the interfaces connected to it have no lvl 5 traffic [21:15] you can't make rules [21:15] and yoy din't seem to believe me [21:15] fine. you can't make rules :-) [21:15] I give up [21:15] again i am sorry [21:16] i hope you never have to do this [21:16] I'm running a bridge with 7 mixed virtual and physical interfaces in production. [21:16] level5 traffic? your blocking http cookies? :) [21:17] http://forum.soft32.com/linux2/Ebtables-stop-DHCP-ARP-ftopict40834.html [21:17] * Nafallo shrugs [21:18] HOw do I update php5.2.10 to 5.3? or the lates version? [21:18] garymc: worked ? [21:18] no I got disconnected [21:18] I think I need to update php [21:18] ebtables? [21:18] come on man :) [21:18] iptables -m physdev --physdev-in eth1 [21:18] iptables -m physdev --physdev-in eth0 [21:18] Nafallo , again i'm sorry to have bothered you [21:18] kthxbye [21:19] hi ivoks :-) [21:19] i'm fine now , i have found what i needed , [21:19] ok whats the best way to update php withour busting anything? [21:19] Dibbler__: good. :-) [21:19] garymc: always better to have the lastest php but you don't need it to fix that [21:19] garymc: apt-get upgrade [21:19] no? [21:19] upgrading php will always bust php :) [21:19] I dont wanna do that [21:19] so manu people want php 5.3 [21:20] just to realize their app doesn't work [21:20] and i hate web developers [21:20] ok so what do you suggest? [21:20] and I hate you [21:20] :D [21:20] kill webdevs [21:20] :) [21:20] why do you need php 5.3? [21:20] 90% of the php code out there sucks [21:20] I'm pretty happy with php 5.3 :) [21:20] just because of the number or what? [21:20] didn't have anything broke from the upgrade [21:20] well my php developer is running a test version on his server which is working and hes running 5.3 [21:21] and php in itself sucks, just like any other language for the web [21:21] my server is running 5.2.10 [21:21] there's very little that doesn't work from 5.2 to 5.3 [21:21] garymc: tell your web developer that isn't how work is done [21:21] it's very well documented [21:21] you have production and then you copy prouction env to create staging [21:21] http://devzone.zend.com/article/11701? [21:21] ok im lost on what to do here [21:21] or you just make your program compatable :) [21:22] you don't let your web dev to come up with some random stack that's impossible to maintain [21:22] http://www.php.net/manual/en/migration53.incompatible.php [21:22] garymc, you using karmic? [21:22] cause lucid comes with 5.3, and lucid is lts, so [21:22] errrm cant remebr if I upgraded or not [21:22] garymc: what problem do you actually have [21:22] garymc: have your web dev read the migration docs, and get him to code compatibly [21:22] cokegen: his problem are web devs :) [21:22] I don't think upgrading php will fix it [21:23] ivoks: so I'm the problem ? :D [21:23] hes writing a script that shows a date , but on my system its showing 01/01/1970 [21:23] :) [21:23] with what function exactly ? [21:23] date() ? [21:23] date was there before i was born :) [21:23] garymc: and on your system [21:24] yes, but I think the problem is the data he's providing to the date() function [21:24] garymc: if you run 'date' in a shell, what does it show? [21:24] if you provide date() with null you get 01/01/1970 [21:24] here is the string hes using echo date('d/m/Y', strtotime('next friday')) [21:24] so the problem could be something else [21:24] echo strtotime('next friday'); [21:26] ok yes possibly and Im oblivious to what it is [21:26] Nafallo: hi :) [21:28] lifeless : Sat Dec 4 21:27:38 GMT 2010 [21:28] and echo strtotime('next friday'); ? [21:28] should give you a timestamp (that's php code of course) [21:28] not in a shell [21:31] garymc: also, is date/time support enabled in a phpinfo() ? [21:31] ill look now [21:34] yes its enabled [21:34] you need to see if echo strtotime('next friday'); is actually returning an int (timestamp) [21:34] php 'echo strtotime('next friday');' [21:34] will that work? [21:35] nope :) === coke_ is now known as cokegen [21:36] php -r 'echo strtotime("next friday");' [21:38] where do I put that? in terminal window? [21:38] on code [21:40] on code? [21:40] just do a new php file and put [21:41] see what that gives you [21:41] should give you an int [21:41] 1291939200 [21:42] what the hell is that ? ^ [21:43] php -r "print date('d/m/Y', strtotime('next friday'));" [21:43] that's date [21:43] that's Fri, 10 Dec 2010 00:00:00 GMT [21:43] well, then that date line should never give you 01/01/1970 [21:43] you're looking somewhere else man [21:44] i bet he doesn't have connection to sql :) [21:44] date is independent from whatever config you have providing the 2nd paramenter [21:44] ok it must be the php code my pdev is using [21:44] php -r "print date('d/m/Y', '0');" [21:44] that's 1/1/70 [21:45] ivoks, THAT should give you 01/01/1970 [21:45] yep [21:45] unix epoch [21:45] but not with 1291939200 [21:45] I say impossible [21:45] of course [21:45] that's what i'm saying [21:45] he's 'dating' 0, instead of meaningfull time [21:45] maybe he placed the code somewhere else etc etc [21:46] he's feeding 0 to date() [21:47] would updateing be harmful? [21:48] well, it seems php is NOT the problem [21:48] so I'd recommend against upgrading in this case [21:49] problem is in web dev, as always :D [21:49] make that 'php wana be web dev' [21:50] am i gonna mess up php if i try upgradE? [21:51] garymc: we can't tell you if upgrading php will help or not [21:51] garymc: cause date() function isn't the problem [21:51] garymc: you don't know where it is, and it's very hard for us to find it for you [21:56] fuck ... x-chat dies on me [21:57] * cokegen is starting with all the linux idiosyncrasies [22:14] ok i need to upgrade php [22:15] why's that ? [22:15] Apparently there are other reasons I need to upgrade php [22:16] does "apt-get upgrade" upgrade everything? [22:16] for security, not much from the functionality perspective [22:16] not just php? [22:16] apt-get upgrade php [22:16] (I think) [22:16] you think i will break php? [22:16] you should update whole distrbution [22:16] I agree with ivoks [22:16] will it mess any of my current setup? [22:16] if you have php 5.2, then you are in pre-lucid [22:17] I got a working web server on there etc [22:17] it an LTSP server [22:17] Linux Terminal Server Project [22:17] do you even know which version of ubuntu you have? [22:17] I have upgraded from jaunty to karmic [22:18] I think im on karmic [22:18] ubuntu 10 something [22:18] karmic is 9.10 [22:18] what command tells me my version? [22:18] lsb_release -a [22:19] Karmic 9.10 [22:19] it also says "No LSB modules are available" [22:19] that will be EOL in 5 months [22:19] So If I upgrade will it delete my SQL server or anything? [22:20] no, it will upgrade [22:20] ok so whats the command to upgrade the whole system? [22:20] apt-get upgrade [22:20] you are maintaining that server since jaunty? [22:20] yes [22:20] and you don't know basic functions of apt? [22:21] well, im in and out of use, ive got a bad memory used to smoke alot of pot [22:21] you need to update to current karmic patches and fixres [22:21] apt-get dist-upgrade is used for that [22:21] can i not do that with one command? [22:21] do-release-upgrade does all that surely? [22:22] Nafallo: i don't think it updates current release [22:22] Ok so could you tell me the exact command I need to use? [22:22] not sure if it did back in karmic, but pretty damn sure it did in 10.04. [22:22] i did, a minute ago [22:22] apt-get dist-upgrade [22:22] ^ yes [22:22] garymc: Error: "yes" is not a valid command. [22:23] yes [22:23] ok im going to do it [22:23] you should do that reguallry [22:23] if you didn't, then your system is full of security holes [22:23] ok its running now [22:25] i guess its gonna take a while [22:25] Nafallo: i didn't do do-release-upgrade since hardy :) [22:26] ivoks [22:26] ivoks: your systems wouldn't be covered by support then ;-) [22:26] it failed [22:26] Nafallo: eh? [22:26] Nafallo: from hardy to lucid [22:26] garymc: hit the panic button!!! [22:27] ivoks: pretty sure our only supported options would be to use the upgrade-manager these days. I could be wrong though. [22:27] * Nafallo shrugs [22:27] garymc: since you never did security updates, your system probably doesn't know where to find them [22:27] it says to try apt-get update [22:27] that's why you should run apt-get update first [22:28] ok lol [22:28] Nafallo: you didn't understand me [22:28] Nafallo: i haven't run do-release-upgrade on anything newer than hardy; all my servers are LTS [22:28] ivoks: ah. sorry. yes. makes sense now :-) [22:29] ivoks: I'm obviously a bit disappointed that your laptop isn't running the latest and greatest, but that's not my concern ;-) [22:29] it's maverick :p [22:29] oh. and you would have done it through the gui... never mind. [22:29] * Nafallo blames the party last night :-P [22:34] Ivocs : I ran upgrade but it says im still 9.10 [22:35] of course you are [22:35] don't you read? [22:36] yep I ran apt-get dist-upgrade and it hasnt upgraded me but looked like it went smooth [22:36] 22:21 < ivoks> you need to update to current karmic patches and fixres [22:36] now you are on up to date karmic [22:36] to upgrade to ubuntu 10.04 [22:36] Ohhh right, so what do I do? [22:36] you should run do-release-upgrade [22:36] so [22:37] apt-get release-upgrade [22:37] khm [22:37] khm? [22:37] 22:36 < ivoks> you should run do-release-upgrade [22:38] so in terminla type "do-release-upgrade" ? [22:38] ok its away [22:39] ok it says I f I continue an additional ssh daemon will be started at port '9004' [22:40] we know how it goes [22:41] we've done it... lost of times, if not a hunderd :) [22:41] OK, I dont mean to be a pain [22:41] i updated 10.4 this morning now my network interfaces are missing [22:41] talking about perfect timing :) [22:41] you gotta be kidding me!! [22:41] can anyone help me either undo the update or walk me through claming the network interfaces [22:42] cap_00: dmesg | grep eth [22:42] ARRRRHHHHH!!!! Dont be saying this shit [22:42] i don't get anything under that command.... nada [22:42] cap_00: do you know what network cards you have? [22:42] my screen is just sitting at Calculating changes.. is this normal? [22:42] lshw -C network shows both my interfaces as UNCLAIMED [22:42] intel, built into the mobo [22:43] cap_00: uname -a says what? [22:43] wait... unclaimed? [22:44] uname -r [22:46] ya, first one says linux server 2.6.32-26-server yada yada [22:46] lspci | grep Ethernet [22:46] paste it on the pastebin [22:47] pastebin? [22:47] garymc: please control the language [22:47] pastebin.ubuntu.com [22:47] oh [22:47] sorry [22:48] k [22:48] i can't ssh, obviously so i'm writting everything out by hand [22:49] oh,right [22:49] then just type the model [22:49] anyone here familiar with cyrus ? [22:49] Intel Corporation XXXXXXX Gigabit Network Connection [22:49] that XXXXXXX [22:49] type it here :) [22:49] ivoks: do "sudo ifconfig -a" do you see an eth device for it [22:50] 82573L and 82573E [22:51] those are supported [22:51] hmm, I have an 82574L, e1000e driver [22:51] sorry, that was for cap_00 [22:51] looking at 82575EB [22:51] ya, everything was working fine until this morning when i updated using apt-get update [22:51] cap_00: lsmod | grep e1000 [22:52] then lsmod | grep igb [22:52] ifconfig only shows lo even though i name eth0 and 1 in /etc/networking/interfaces [22:52] k jas [22:52] sounds like the modules are not loaded [22:53] ya [22:53] i don't understand you cap_00 [22:53] :) [22:53] ah, Just A Second [22:54] what exactly do you have in interfaces? :) [22:54] neither lsmod did anything [22:54] interfaces aren't the problem, at least not yet [22:54] modprobe e1000e [22:54] hhm [22:54] modprobe igb [22:54] not sure which one is for that model, but one of those must be [22:55] FATAL: Module e1001e not found [22:55] e1000e [22:56] same [22:56] you don't have e1000e module?! [22:56] says... [22:56] /lib/modules/2.6.32-26-server/kernel/drivers/net/e1000e/e1000e.ko [22:57] do you have this file? [22:57] all config files need .conf: /etc/modprobe.d/aliases, it will be ignored in a future release then the fatal part [22:57] ah [22:57] mv /etc/modprobe.d/aliases /etc/modprobe.d/custom-aliases.conf [22:57] depmod [22:58] modprobe e1000e [23:00] did mv and not found [23:00] kernel module not found? [23:00] i think so [23:01] FATAL: module e1000e not found [23:01] ls /lib/modules/2.6.32-26-server/kernel/drivers/net/e1000e/e1000e.ko [23:01] does that ^ file exist? [23:02] it's called e1000.ko on my box [23:02] in the /e1000/ dir [23:02] there are both e1000 and e1000e [23:02] i asked for e1000e [23:03] i only have the e1000 when i do ls /lib/modules/2.6.32-26-server/kernel/drivers/net/ [23:03] how's that possible? [23:03] did you run out of disk space? [23:03] good question [23:03] nope [23:03] maybe his fs got corrupted [23:03] 50gb free [23:04] then just install the package again [23:04] or apt didn't finish unpacking the kernel? [23:04] apt-get install --reinstall linux-server [23:04] how, i'm not on the net anymore [23:04] actually, need the real package [23:04] dpkg -i /var/cache/apt/archives/linux-image-2.6.32-26-server*/deb [23:04] cap_00, doesn't matter, unless you did an apt-get clean [23:04] bah [23:04] dpkg -i /var/cache/apt/archives/linux-image-2.6.32-26-server*.deb [23:05] it won't overwrite any server files will it? like samba? [23:05] no [23:05] just broken kernel [23:06] Patrickdk: that would just reinstall meta package ;) [23:07] I realized that after I type it :) [23:07] no such file or directory [23:07] :( [23:07] boot an older kernel [23:07] usb stick time :) [23:07] or that :) [23:07] so, someone was tampering with your files [23:08] if the cache is empty, someone deleted it [23:08] i was in the middle of trying to get KVM to bridge to my lan [23:08] and i find it... strange that only modules missing are those you need :) [23:09] other then that, all i did this morning was apt-get update and apt-get upgrade [23:09] dist-upgrade i hope [23:09] oh no!!!!!!!!!1 [23:09] just upgrade [23:09] upgrade wouldn't install new kernel [23:10] so, you were broken before you did apt-get [23:10] wow [23:10] ok, we'll lets fix it [23:10] [17:48] Dec 4 17:44:39 ubuntu cyrus/imap[12980]: executed [23:10] [17:48] Dec 4 17:44:39 ubuntu cyrus/imap[12980]: accepted connection [23:10] [17:48] Dec 4 17:44:39 ubuntu cyrus/imap[12980]: badlogin: [192.168.1.114] plaintext databits SASL(-13): authentication failure: checkpass failed [23:10] so i need 10.4 LTS image on a bootable usb? [23:10] or will a CD do? [23:10] no [23:10] does anyone have any idea what might be causing this issue ? [23:10] you need to boot and older kernel [23:11] hold down shift while you boot [23:11] use down arrow [23:11] if you don't have, copy the deb to the usb stick [23:11] is there a live 10.4 command line server usb ? i thought i tried and failed [23:11] or just install to a usb drive? [23:11] you just need usb stcik [23:11] got one [23:12] it is 386 or amd64? [23:12] oooooh copy that file from the cd? [23:12] 64 [23:12] http://hr.archive.ubuntu.com/ubuntu/pool/main/l/linux/linux-image-2.6.32-26-server_2.6.32-26.48_amd64.deb [23:12] download this [23:12] copy to stick [23:12] install on server [23:13] of course, if you have needed usb modules :) [23:13] lol lets hope [23:14] the backup drive is usb and it worked last week... [23:14] your system broke between last reboot and now [23:15] oh right... [23:15] fingers crossed [23:15] it would be smart to find out when, who and why [23:15] i'm the only one with root priveledge.... ??? [23:16] if $who is someone you don't know or $why just sounds stupid, you'll need to reinstall the system :) [23:16] really? [23:16] then i guess you know who $who is :) [23:16] lol [23:16] unless you have some disorder [23:17] i hope not [23:17] http://en.wikipedia.org/wiki/Dissociative_identity_disorder [23:17] like this one [23:17] lol [23:17] let me check the raid 1 array to see if something is up there [23:18] or... of course... maybe you just think that you're the only root [23:19] while reality is somewhat different [23:19] raid looks fine [23:20] it's not raid [23:20] if it would be raid, you won't be able to boot [23:20] it user [23:20] just wanted to make sure it wasn't something hardware.......... ish [23:20] it's user [23:20] how do i see a full list of priveledged users? [23:20] that's what i'm saying... if it's hardware problem on raid, then your FS wouldn't know who it is and refused to work [23:21] members of admin group [23:21] /etc/sudoers [23:21] beautiful! if my server gets too high a load average...all my virtual guests go crazy with CPU Stuck issues. wait a go kvm! [23:21] and, what lots of people forget, /root/.ssh/authorized_keys [23:22] osmosis: or... you could ask first [23:22] shuttleworth says it right here, ubuntu is not a server distro. http://www.markshuttleworth.com/archives/517 [23:23] i'm the only IT guy here, and certainly the only one that knows linux [23:23] nice job promoting your own distro to business when you are supposed to want to compete with Redhat [23:23] not that i'm much past a newb with a book [23:25] I don't see anything in that post that says it's not a server distro [23:25] ikonia: ? [23:25] ivoks: shuttworth's blog post [23:25] ikonia: what about it? [23:26] ikonia: have you read it? [23:26] osmosis, what version are you using? [23:26] ivoks: pretty much [23:26] it mentions server 4 times [23:26] once, where it says that server is ahead of desktop [23:27] yann2, LTS 10.04. Its been out for 7 months, and still not stable for virtualization [23:27] and other three times in links [23:27] osmosis CPU stuck you say? on the guest or the host? [23:27] osmosis: i do high cpu work on VM, and other VMs don't have a problem [23:27] ivoks, shuttleworth says that his focus is on improving the desktop [23:27] yann2, guest [23:27] osmosis, what are the guests? [23:27] "it just wasn’t being delivered in a way that would take it beyond the server, or to the general public." [23:27] 10.4 too? [23:27] forget about it. ive been trying to get it fixed for months. [23:27] yann2, also 10.04 LTS [23:28] have been running 10.4 hosts and vms for a while, havent had any issues yet, but relatively low load. I've got CPU stuck issues on a specific sun server and with hardy though :) [23:28] I think that it's safe to agree that linux is doing better in the server space that on the general public's desktop. he doesn't negate the work on -server anywhere in that. [23:29] kvm-pxe is broken in lucid though, but easily circumvented by using maverick's [23:29] oh well, dont worry about it. im just frustrated. sorry to vent. bye [23:29] or, even easier [23:29] yann2: all the files are there, just in the wrong place [23:30] simple cp solves the issue [23:30] usb works [23:30] ivoks, kvm-pxe? maybe, not sure, there was a symlink story, but I think it didnt work even with the symlinks (as far as I remember) [23:30] but can't install file [23:30] cap_00: how come? [23:30] i can install it right from /mnt/usb/ right? [23:30] I think there was another bug behind the one you mention :) [23:31] Wow this takes a long time [23:31] yann2: i've been using pxe boot since lucid was released [23:31] yann2: i've pxe booted lots of machines on lucid [23:32] not sure what you are talking about [23:32] says no such file or directory :( [23:32] cap_00: did you mount the drive? [23:33] unless i'm missing a typo.... sudo dpkg -i /mnt/usb/linux-image-2.6.32-26-server_26.32-26.48_amd64.deb [23:33] ya i can ls [23:33] cap_00: did you mount the drive? [23:33] ok [23:33] does installation starts? [23:33] or it tells that /mnt/usb/linux-image-2.6.32-26-server_26.32-26.48_amd64.deb doesn't exist? [23:34] doesn't exist [23:34] is that the only .deb file there? [23:34] yup [23:35] try sudo dpkg -i /mnt/usb/linux*deb [23:35] ok [23:35] that way making a typo is much harder [23:36] error processing /mnt/usb/...... conflicting packages - not installing linux-image2.6.32-26-server [23:38] that seems very bad :( [23:39] it also says -virtual conflicts with -server [23:39] virtual? [23:40] i installed kvm libvirt-bin ubuntu-vm-builder [23:40] is that virtualized server or normal server? [23:40] and linux-image-virtual for kvm [23:40] no man [23:40] :) [23:40] linux-image-virtual is for virtualized machines [23:40] woops.... [23:40] for guests [23:40] why would i do that [23:40] remove? [23:41] yes [23:41] reboot? [23:41] no [23:41] you have half-installed kernel now [23:42] 32.8kb of pain gone [23:42] the one you installed from usb stcik [23:42] stick [23:42] first, remove all virtual kernels [23:42] it still says conflicts [23:42] how? [23:42] apt-get remove linux-.*virtual.* [23:43] couple of damaged link warnings [23:43] says may need to re-run boot loader [23:44] did it finish? [23:44] which i did this morning after the udate......... all cause i installed the virutal kernel on my host? [23:44] i think it did [23:44] did it finish? [23:44] whell, do you have shell or not? [23:44] :) [23:44] yes [23:44] install the kernel from usb stick [23:45] k [23:45] YAY! [23:45] now reboot [23:45] *bows down* [23:46] http://en.wikipedia.org/wiki/Dissociative_identity_disorder [23:46] ^^ [23:46] ivoks: Error: "^" is not a valid command. [23:46] lol [23:47] they're up now [23:48] should i do a apt-get upgrade to clean anything up? [23:48] dist-upgrade [23:48] k [23:48] that'll keep me on 10.4 lts right? [23:48] yes [23:48] y [23:49] do-release-upgrade moves you to newer version [23:49] perfect command name [23:49] lol [23:49] i'm serious [23:49] :) [23:49] warning: linux-server linux-image-server cannot be authenticated [23:50] update [23:50] y/n? [23:50] k [23:50] i did linux-image-virtual for kvm like a week ago lol and just rebooted this morning [23:51] it's stuck at 0% connecting to us.archive.ubuntu.com [23:51] ah... us [23:51] can't do right... anything :) [23:52] lol [23:52] it's actually cname for UK mirror, iirc [23:52] or not [23:52] may need to tweak the network connection i can't ping google.com [23:53] i can only ping machine's on my lan [23:53] well, that's another story now [23:53] and since it's almost 1AM over here [23:54] i'll pass this one :) [23:54] lol ok [23:54] thanks for all your help, i should be able to peice this together [23:54] np, good night [23:54] night [23:54] im still upgrading :S [23:55] should I reboot server once relaese upgrade is completE? [23:55] ouch [23:55] or does it do it automaticly [23:56] cap_00 what you mean? [23:56] am i about to get my fingers well and truly burnt? [23:57] i was joking [23:57] :S phew [23:57] gota go pick missus up now too [23:57] do-release-upgrade will guide you [23:58] still upgrading... it goes on on and on like Ariston [23:58] sleep typing ;) [23:58] no, last smoke and last beer [23:58] ok ill be back in 20 mins if your still here