[00:01] If I "ifdown eth0" does eth0:0 stay up? [00:03] Gargoyle: nope [00:03] Gargoyle: urr, dunno [00:03] I think it's a no! :) [00:07] What are the concequences of rebooting the server if it's eth0 ip address will knowingly conflict with another server? [00:07] will eth0:0 come up? [00:09] Ahh, balls to it... I'll just change them one at a time... worst case is that it sets off the alarms at the hosting co and some poor sys admin has to come and check what is going on! [00:09] it'll come up, cached arp records be damned [00:09] and your router will get confused if it doesn't have static arp mappings [00:10] and your hosts will hate you [00:10] :) [00:10] I am just gonna do em one at a time. so one address will disappear from the network for a min or two! [00:12] awww, poo [00:12] ssh has not come back up! [00:12] Don't you just hate remote managing servers! [00:13] dont tend to manage many locally [00:13] so i guess not, heh. [00:14] * Gargoyle wipes his brow... [00:14] It was just on a go slow! [00:20] andol: I don't know, I'm afraid - mathiaz might [00:21] DogWater: I'd very much like to work with you to debug your Kickstart problems, but I'm just about to go to bed. Could you please (a) file a bug on https://launchpad.net/ubuntu/+source/kickseed/+filebug with as much detail as you can (at least a copy of your Kickstart file, and the installer's syslog file if you can), and hop onto #ubuntu-installer sometime tomorrow? [00:21] DogWater: (I wrote our Kickstart implementation and if it broke it's probably my fault.) [01:13] I have created a directory called, "/vmfiles/vm" However, if I go into the vmfiles directory and run "ls" it does not display the vm directory. But I can still navigate to /vmfiles/vm. Any ideas why vm does not show up in ls? [01:26] New bug: #319843 in mysql-dfsg-5.1 (universe) "root password is not asked during default install" [High,Triaged] https://launchpad.net/bugs/319843 [03:18] hi, in ubuntu , how can i make a user owner of a directory and all its sub directories, i'm trying with chown [03:19] chown -R user:group /path/to/dir [03:20] thanx === MianoSM1 is now known as MianoSM [04:51] hi [04:52] can someone help me with an issue am having concernin ia31 lib please [06:30] is there a specific solution for slow response from 8.10 server?? there is like 2-5 seconds delay before it respond ... in ssh session, webpage, wordpress. i have exhausted the resources on looking into dns, resolv.conf, hdparm. i'm at loss to resolve it. [07:17] Does anyone uses nagios? I could really use a helping hand here. I tried to compile nagios-plugins-1.4.13 but the make command returns undefined reference to `np_net_ssl_read' [07:18] vertx: users of nagios usually don't compile their plugins by hand :) [07:20] I tried googling it. The only suggestion I could find is that I have to install ssl-dev package, for which I have already done using libssl-dev package. Did I do anything wrong in between? === MohammadBoozary is now known as Mohammad[B] [07:25] @Koon: thanks for your reply. But I can only see nagios2 available on the repo. I'm using Ubuntu Server 8.04, btw. What do you suggest? [07:35] Good morning from norway :) [07:35] I can't get this guide: https://help.ubuntu.com/community/PamCcredsHowto to work with intrepid. What am I missing? [07:36] when i try to use getent passwd, whe user is found, but when i try to log in, i'm getting authentication failed === frippz is now known as Fredrik === Fredrik is now known as frippz [08:03] vertx: the plugins are there somewhere, let me see [08:03] nagios-plugins (1.4.11-1ubuntu5) [universe] [08:04] splitted into nagios-plugins-basic nagios-plugins-extra and nagios-plugins-standard [08:04] Hello! [08:04] when i try to start my xen instance i get this error: Error: Device 51713 (vbd) could not be connected. Backend device not found. [08:04] with this config file: http://pastebin.com/m4692ce62 [08:05] any idea why? [08:06] spiekey: maybe #ubuntu-virt can help you. [08:06] thanks [08:34] moin [10:17] I've got a Dell PowerEdge 1950 that I'd like to read temps from. anyone who knows what software is supported for this? [10:18] what are temps!? [10:19] temperatures [10:20] not temporaries :) === MianoSM1 is now known as MianoSM [11:48] I need help - I tried to remove sendmail using apt-get and it somehow failed. Now anything I triy to do with apt-get or dpkg starts sendmail config ("Setting up sendmail-bin (8.14.3-4) ...") and then fails again with "E: Sub-process /usr/bin/dpkg returned an error code (1)" - what to do ? [12:03] anyone ? [12:06] barduck: apt-get -f install ?? [12:07] tried "apt-get -f install sendmail" and "apt-get -f remove sendmail", same result [12:07] dpkg -a configure [12:08] that is a syntax error [12:08] Hmmm. I cant remember what it's ment to be come to think of it. [12:08] barduck: is there any other message between "Setting up ..." and "E:"? [12:09] agentk: you meant dpkg --configure -a, but it won't help here [12:09] cjwatson: yeah, there is tons of stuff sendmail config trying to do until it finally gives some errors and quits [12:09] I dont want sendmail [12:09] barduck: try 'dpkg --remove sendmail-bin' [12:09] I want to completely get rid of it [12:09] cjwatson: same result [12:10] it says sendmail is not installed [12:10] then runs the sendmail config [12:10] surely not, it can't possibly begin with "Setting up sendmail-bin" for starters [12:10] barduck: could you put a transcript on paste.ubuntu.com, please? [12:10] cjwatson: of course [12:13] cjwatson: this is when I run dpkg --configure -a : http://paste.ubuntu.com/108205/ [12:14] it seems that /etc/init.d/sendmail has been removed but sendmail-bin still expects it to be there. Personally I would create /etc/init.d/sendmail that just reads like this: [12:15] #! /bin/sh [12:15] exit 0 [12:15] chmod +x that file, and then try again [12:15] I will try that [12:15] you can remove that file later [12:15] this is of course a bug that we'd appreciate you reporting on LP [12:16] I think I also removed /var/cache/apt when I tried to fix the above problem so now I get "E: Could not open lock file /var/cache/apt/archives/lock" from apt-get [12:16] should I just create the dir ? [12:20] yes [12:21] cjwatson: ok, that fixed the dpkg --configure . But now I have sendmail in a zombie state. The package is not installed but I have traces of it all over my system. Any way to completely remove it [12:22] ? [12:23] you removed /var/cache/apt ?? [12:23] what else did you mess about with? ;-) [12:23] barduck: put the output of dpkg -l "sendmail*" on paste.ubuntu.com please? [12:24] cjwatson: alot ! :) sendmail is stuck all over my system and it completely gets on my nerves...and I am not known for my patience :) [12:24] cjwatson: coming right up [12:24] you may have interfered with dpkg's own cleanup and made your problems worse [12:24] it couldn't have been worse [12:24] or could it ? [12:26] cjwatson: http://paste.ubuntu.com/108206/ [12:26] frippz: dmidecode might get you the info you need [12:26] frippz: see http://www.thegeekstuff.com/2008/11/how-to-get-hardware-information-on-linux-using-dmidecode-command/ [12:26] Koon: thanks. will look into it [12:27] barduck: ok, 'dpkg --purge sendmail-base sendmail-bin sendmail-cf' [12:29] holy cows ! It is gone! I am free at last! [12:29] cjwatson: thanks a lot [12:29] and I know better now what to do if this happens again [12:30] * barduck runs to install postfix [12:30] barduck: (and remember to remove the /etc/init.d/sendmail hack you put in place) [12:30] cjwatson: it is gone... [12:30] from the purge, I guess [12:31] ok [12:32] barduck: apt-get purge "sendmail*" would probably have done it in the first place [12:32] or use aptitude to mark all sendmail-related packages as purged, and then tell it to go [12:32] cjwatson: what I initially did that screwed it all is "apt-get purse sendmail" [12:33] *purge [12:33] I suspect sendmail*'s dependencies are subtly broken in some way that affects purge, and thus you need to get the order right :-/ [12:33] hence suggesting purging them all at once [12:34] that won't happen again... [12:34] but thats only because sendmail got on my nerves and I will never install it again [12:34] :) [12:36] ok, I think my system is in serenity again... [12:36] many thanks, cjwatson [12:37] no problem [12:39] So is anyone working on fixing the kickstart system in 8.10? [12:48] what's wrong with it? [12:50] DogWater: I replied to you last night, but maybe you missed it? [12:50] 00:21 DogWater: I'd very much like to work with you to debug your Kickstart problems, but I'm just about to go to bed. Could you please (a) file a bug on [12:50] https://launchpad.net/ubuntu/+source/kickseed/+filebug with as much detail as you can (at least a copy of your Kickstart file, and the installer's syslog file if you can), and hop onto [12:50] #ubuntu-installer sometime tomorrow? [12:50] 00:21 DogWater: (I wrote our Kickstart implementation and if it broke it's probably my fault.) [14:20] Is anyone using rdiff for backups? [14:22] hi all [14:22] how to i can resolve it --> http://paste.ubuntu.com/108240/ ? please help me :( very important [14:27] cjwatson: sure I can post it for you, it is just kind of strange really the 8.04 install with the same ks file works fine (except it still asks me to partition which i need to eliminate) but the same file on 8.10 is just insaneoflex [14:30] DogWater: kickseed is sometimes sensitive to other changes in the installer, so it can happen that we change something else and then if we forget to update kickseed it all breaks [14:30] DogWater: (remember to remove passwords from the Kickstart file before posting it, obviously!) [14:31] hello all [14:33] anyone here taken the LPI tests? [14:36] New bug: #292293 in openvpn (universe) "udev rules for tun device have wrong permisisons" [Undecided,Invalid] https://launchpad.net/bugs/292293 [15:20] How did I set a new IP for eth0:X? [15:23] What is the proper way to manage the /etc/inetd.conf file? I keep deleting the ## string from the services I need but something keeps putting it back [15:33] soren, do you happen to be around ? [15:34] * ogra needs some qemu advice [15:35] does anyone know by what means Landscape receives temperature info from a server? [15:35] frippz: lmsensors? [15:35] henkjan: you sure? [15:36] frippz: no, just a guess [15:36] more likely acpi through sysfs [15:36] or /proc [15:37] I'm on a trial here and just noticed that the temperature info was empty. would like to fix that :) [15:37] or it installs a little dwarf inside your server case that put his fingertips at the neuralgic temperature points [15:37] ogra: :P [15:37] better be a damn small dwarf for a 1U server :D [15:38] yeah, dont forget to feed him little sandwiches ... else he wont measure :) [15:39] "gnome inside" [15:39] Authority: update-inetd [15:40] cjwatson: thanks, i'll give that a try [15:42] cjwatson: does it make sense that if I'd previously just used vim to edit the file that something else would be disabling the service again though? [15:51] Authority: it sounds odd, but it's a while since I looked into update-inetd [15:51] Authority: you might want to file a bug [15:52] cjwatson: well I'm going to watch the file for a few days and see if it happens again before I do something like that. might be good to check though to see if someone else has already filed a similar bug. thanks for the assistance [15:53] well, if you have i.e. openbsd-inetd installed and its set to start standalone in /etc/default, and you run apt-get upgrade, debconf might listen to the /etc/default setting and take it out on inetd.conf [15:53] err [15:54] s/openbsd-inetd/tftpd-hpa/ [15:54] s/on/of/ [15:54] s/debconf/some maintainer script that uses debconf/ [15:54] right [15:54] however, any such action is a bug [15:54] but in this case /etc/default counts ... [15:54] and should be reported and fixed [15:55] if you're really crazy enough to have two configuration files with potentially conflicting files and try to sync one to the other, you should at least check which is newer. Better still, don't do such a silly and confusing thing. [15:56] (where "you" = package maintainer) [15:56] well, its how tftd-hpa comes from debian ... [15:56] i always found that behavior odd [16:04] When I run apt-get upgrade, new kernels are being held back. What's the best way for me to tell what security fixes (if any) are included in the new kernels? [16:04] colonelqubit: reading the changelog? [16:05] I like apticron because it sends a email with that info, the security update, the changelog, and even the CVE report [16:05] jmedina: okay, but 'linux-server' is a metapackage, so while dpkg -l tells me what kernel rev I'm on now, how can I tell what version I will be upgraded to? [16:08] colonelqubit: dist-upgrade would install the new kernels [16:08] mmm I think you never upgrade to a major kernel version, only some patch to your current kernel [16:08] (see the apt-get manual page for the difference between upgrade and dist-upgrade) [16:08] someting like [16:08] * Kernel ABI bump for linux-source-2.6.15 version 2.6.15-52. [16:08] Security update. [16:09] colonelqubit: if you install apt-listchanges and configure it appropriately, it'll show you all the changelogs for everything you're upgrading, and prompt you before going ahead and upgrading [16:09] it sounds like what you want [16:10] cjwatson: hmm. I'd prefer to not have to install any new packages, but that does look promising. [16:10] in fact, apticron uses apt-listchanges [16:11] colonelqubit: you will have to install new packages any time the kernel ABI changes, I'm afraid, and that is sometimes necessary [16:11] you can always say no to apt-listchanges and then be more selective, if you like [16:12] cjwatson: oh, I just meant that I wish I didn't have to install the 'apt-listchanges' package, hoping that apt-get or dpkg already had enouch functionality... [16:12] oh, right [16:12] cjwatson: what about 'apt-get install linux-server -s' ? [16:12] you can in principle pick the changelog out of a package with just dpkg but it's annoying. [16:12] (maybe the -s goes before the package name) [16:13] colonelqubit: that will tell you what new packages will be installed as a result, but will not show the changelog [16:13] nor anything about security fixes, etc. [16:13] cjwatson: sure, but if I can get the version # of the kernel, I can then just look online at the changelog [16:13] in that case why not just 'apt-cache show linux-server', look at the highest version number paragraph there, and check the dependency [16:14] or indeed, why not just https://launchpad.net/ubuntu/+source/linux and look for the highest one targeted at your release :-) [16:15] cjwatson: good point -- I should be resonably certain that upgrading linux-server will give me the latest and greatest in the package repository? [16:18] generally, though sometimes linux-meta updates lag a little bit behind [16:18] I always recommend to suscribe to the USN list [16:22] jmedina: oh, yes, I subscribe to USN, but I usually only glance at it. The current sysadmin is out with a new baby, so I'm taking over duties temporarily... :-) [17:41] Does anyone know if there is a way to force the settings of a ufw command? I'm trying to run a script to enable ssh commands across several hundred machines and each time it asks for a y|n confirmation because my setting "may disrupt ssh communications" [17:43] kaje: no. the version of ufw in jaunty will have preseeding/debconf ability though [17:43] kaje: for your current situation, you can do: [17:43] ufw allow ... [17:44] sed the /etc/ufw/ufw.conf file [17:44] /etc/init.d/ufw start [17:44] right, but when I do ufw enable it asks me if I'm sure I want to do that... I'm not looking forward to hitting y hundreds of times [17:44] kaje: notice we bypassed 'ufw enable' there [17:44] ahh [17:45] so I could just copy the ufw.conf file over and then start it? [17:45] kaje: sure. that also goes for /var/lib/ufw/*rules files too [17:45] kaje: and really anything in /etc/ufw/* [17:46] kaje: but you can add rules (eg ufw allow 22/tcp) without it being enabled [17:46] I see [17:46] kaje: and then just turn it on at the end eith 'enable' or with what I suggested a moment ago [17:47] kaje: if you are scripting then use expect to automatically answers... [17:47] expect, I'll take a look at that [17:47] Thanks for all the help guys =) [17:49] * jdstrand considers adding 'force-enable' [17:55] hi guys, i have a server that is unable to ping fqdn (ie google.com)... it's setup as DHCP and /etc/resolv.conf looks the same as other ubuntu boxes that are OK. Any ideas? [17:59] ziggles: I presume you meant it is setup as a DHCP server right? could it ping any other IIPs (private or pulic)? [18:03] hi vertx. It's actually a client at this point. I had it set on a manual/static IP but then noticed this problem. I then switched it to DHCP to make sure i was not screwing up any configs [18:06] would anyone have info on making wifi better, as in VNC on my lan is beyond slow how do i improve it [18:06] ziggles: try pinging upstream, to your modem/router's ip address, then do a tracepath to google or something, see what happens. Does ping to google returns with an ip address? [18:07] vertx: what seems to be ziggles problem [18:08] Faust-C: he could not ping FQDN (ie google) [18:08] Faust-C: i cannot ping fqdn from one host on my lan. [18:08] oh oops :X thanks vertx [18:09] vertx: i can ping the router, and tracepath results in No Reply after the router [18:10] i wonder if its the router that's blocking me out... but it's just a cheap dlink home router... no real configs. strangely enough this host was able to ping out last night [18:10] New bug: #319848 in mysql-dfsg-5.1 (universe) "upgrading from mysql-server-5.0 to mysql-server-5.1 doesn't work" [Undecided,Confirmed] https://launchpad.net/bugs/319848 [18:12] ziggles: i just had that issue [18:12] ziggles: do you have DNS on your lan [18:13] i do not [18:13] ziggles: ill give you a example of my setup [18:13] this really isnt a fancy setup lol... i just have a few ubuntu boxes on a home network one being a server. I'm trying to config this box before i bring it out [18:14] Faust-C: awesome, thank you. [18:14] att wwan (internet) <- (NAT using arno-iptables) -> eth0 (lan) [18:14] eth0 runs DNS and DHCP [18:15] once i installed bind inet for wireless clients works [18:16] was it working at all before you installed bind? [18:16] only for local wired clients [18:16] but laptop wasnt getting inet [18:16] ziggles: logs have any info? [18:17] the only error i have really is this: Jan 22 10:17:01 lucca console-kit-daemon[5161]: CRITICAL: cannot initialize libpolkit [18:17] but i dont think it has anything to do with the network [18:18] ziggles: what about the following logs [18:19] daemon, syslog, [18:19] abd debug [18:19] and* [18:22] Faust-C: i THINK it looks ok? http://pastie.org/367939 [18:22] i have two bridges [18:23] sec lemme look, ill make my setup more clear too [18:23] bridging eth0 and eth1 to br0 and br1... (they appear in the logs) [18:23] I've also tried setting them to manual and no luck [18:23] thanks Faust-C [18:23] i dont think you can make 2 bridges out of 2 devices [18:24] but lets see [18:25] * Faust-C gets his config [18:26] really? [18:26] my goal was to have virtual machines running through different nics to avoid a bottle neck [18:27] and it was working... for a few hours lol [18:28] ziggles: heh im doing the same thing [18:28] but ill show you a diff way [18:29] awesome thanky ou [18:31] Faust-C: I have a question. How can one determine that bind cache(s) domain name look-ups? Where should I look for them? Isn't it supposed to be cached in /var/cache/bind? That directory seems always to be empty, and streams of name look-up keeps being requested directly to my ISP. [18:31] vertx: it depends on your setup [18:32] for instance w/ mine it uses *itself* when it actually uses ISP [18:32] so its a forwarding DNS server [18:32] ziggles: http://pastie.org/367943 [18:33] vertx: you can also set it up to be a caching server [18:33] which i need to do considering my inet is just above dial-up speed [18:33] Faust-C: the clients are set to request to local dns server, The DNS server is setup as bot forwarding and caching. Is that possible? [18:33] yep, thats exactly what im doing [18:35] sometimes i dont understand how i get stuff to work ..... [18:35] But how do I know whether the server is caching those domains or not. Would it keep those caches as a file? [18:36] vertx: you can enable logging [18:36] or try [18:36] sudo rndc querylog [18:36] and watch your logs [18:36] that will log every query request [18:37] you can disable it againt with rndc querylog [18:38] jmedina: thanks I'll try that. where does the log ends-up in? syslog? [18:38] yeap [18:38] jmedina: great. thanks again. [18:39] vertx: is that what you want? [18:39] vertx: and you can also use dig externaldomain.tld and then check the response time [18:39] you will notice that the second request gets faster because is already cached [18:41] jmedina: isn't there a physical evidence of those chaches? [18:41] *cache === apachelogger is now known as apachelogger_ [18:44] jmedina: I mean shouldn't be there one file or a series of files that shows the name look-up caches? [18:45] vertx: the cache is in memory, you can dump that cache to a file with "rndc dumpdb" [18:45] by default stored in /var/cache/bind/named_dump.db [18:45] XD [18:50] jmedina: apt-get rndc returns with "Couldn't find package rndc" :( [18:50] rndc is part of bind9 [18:50] dpkg -S `which rndc` [18:50] bind9: /usr/sbin/rndc [18:53] jmedina: i see. thanks. [18:56] damn faust-c is gone and i didnt get a chance to thank him. [18:59] does anyone have successfully compiled nagios-plugins-1.4.13 on ubuntu server 8.04? or is there an alternative, like a pre-compiled .deb package? I can only find a nagios2 on the repo. [19:01] nagios-plugins-1.4.13 compiles perfectly on my ubuntu 8.04 client, but not on the server. any ideas? [19:02] whats the compile failing on? [19:02] not that ive ever tried compiling nagios ;P [19:03] personally am using zabbix, coz its a bit easier to work with [19:04] adonm: is zabbix good? is it also free? [19:04] vertx: yeah its pretty nice [19:04] vertx: takes a bit to setup but for monitoring and email alerts id say its my favourite free tool [19:05] it does snmp monitoring + agent monitoring [19:05] and the agent is available for windows & linux & solaris [19:05] adonm: that sound great. i'll look it up :) [19:06] i think 1.4something is in the ubuntu repos, and 1.61 is available in a ppa somewhere (zabbix that is) [19:07] adonm: is there a major difference between the 1.4 and 1.6 version? [19:08] adonm: i mean performance or feature wise? [19:13] i think the maps stuff in 1.6 is a fair bit better [19:14] thats the network diagram style dispaly of faults across the network [19:17] adonm: their screenshots looks great. thanks. downloading now :) [19:19] login as: mianosm [19:19] mianosm@miano.us's password: [19:19] Keep it gangster, and let me know what's up. [19:19] Last login: Thu Jan 22 14:13:27 2009 from tbgw10.mybrighthouse.com [19:19] mianosm@miano:~$ sudo htop [19:19] mianosm@miano:~$ clear [19:19] mianosm@miano:~$ [19:19] Is there a fix for this issue yet? [19:21] what issue? [19:21] No request for a password as a sudo user? [19:21] Is it cool to just run sudo without prompting for a password? [19:21] did you log in before? probably password is cached [19:22] his log has the motd [19:22] fauxhawk: wat [19:22] yeap, but we are all over the world with different time zones :D [19:23] jmedina: it is when he logs in [19:23] MianoSM: "sudo -K" - resets the sudo timeout [19:23] It should be clearing the cache on logout then I'd imagine.... [19:23] echo "sudo -k" >> ~/.bash_logout [19:23] Big K that is. [19:29] jmedina: any idea why this isn't the default? [19:29] fauxhawk: I dont know, have you read the manpage? [19:29] you can set the timeout in sudoers === apachelogger_ is now known as apachelogger [19:30] vertx: apt-get install nagios-plugins ? [19:36] Koon: Thanks. I'll try that later. Right now the update manager is downloading 36 updates. Darn, it takes such a long time :( BTW, what version is it? [19:36] 1.4.11 [19:39] Koon: Nice, that should do it. Thanks. Sorry, my inet connection was such a bummer earlier today, so I was repeatedly disconnected automagically, and could not thank you for your help before. [19:50] vertx: you're welcome ;) [20:05] Anyone got any recommendations for a simple rrdtool setup that can give me cpu/mem/net graphs? I was looking at cacti, but it seems a bit much for 3 simple graphs [20:10] hello guys, I have an entry for my host of 127.0.1.1 in /etc/hosts is that normal?? [20:10] shouldn't it be 127.0.0.1 ? [20:18] P4C0: Well, actually the entire 127.0.0.0/8 points towards the loopback interface. [20:22] ok [20:22] but it comes like that by default? or i modified it? ... can't remember :( [20:38] P4C0: yes, it's normal and intentional [20:40] ok :) thanks cjwatson [20:40] P4C0: the point is to have each IP address having only one canonical hostname, to have each hostname having only one IP address, and to have both 'localhost' and the hostname chosen for your box defined as distinct hostnames [20:40] if your box doesn't have a static IP, then the best way to do that is to give it another IP address in 127/8 [20:41] see comment 30 on bug 8980 [20:41] Launchpad bug 8980 in network-manager "hostname -f does not return a proper FQDN" [Medium,Confirmed] https://launchpad.net/bugs/8980 [21:05] New bug: #320212 in samba (main) "cannot access samba share after mount.cifs" [Undecided,New] https://launchpad.net/bugs/320212 === maxbaldwin is now known as Guest36843 === Guest36843 is now known as maxbaldwin [21:44] dang.... i was in here earlier but still cant figure out why one of my hosts is unable to reach fqdns past my router [21:44] anyone have a pointer as to where i can start to debug? === maxbaldwin is now known as Guest89666 === Guest89666 is now known as maxbaldwin [21:48] i have a problem, i did chown -R 'username' www , now i cant change anything in my drupal admin web interface :( [21:50] orudie: not sure if this helps, but does the group that username is have an impact? [21:51] orudie: could you show the output from : ls -ld www [21:51] jmedina, hold on [21:52] paul@linode:/var$ ls -ld www [21:52] drwxr-xr-x 12 paul root 4096 Jan 22 16:46 www [21:52] paul@linode:/var$ [21:54] jmedina, ^ [21:55] well you need to give www-data write permisions to some files,which insecure in my opinion [21:56] FUCK ! [21:56] oops sorry [21:56] everything was running smooth [21:56] before [21:57] fuck where are the backups..... [21:57] :D [21:57] !ohmy [21:57] Please watch your language, attitude, and topic to help keep this channel friendly and helpful. Remember, there are kids here! [21:57] sorry [21:57] dude i had no idea lol [21:57] kids administering servers? [21:57] why would i backup before chmod ? [21:57] that is new [21:57] i mean [21:57] chown [21:58] jmedina: they gotta learn sometime... [21:59] i always use chown command, but never did it with a www dir. , so i guess this will be something that i had to find out sooner or later [21:59] orudie: orudie why wont you chown to the last owner? [21:59] jmedina, can you help me with that ? [22:00] orudie: you remember the last owner? [22:00] jmedina, no, like i said i never messed with the www dir. [22:00] jmedina, isnt it apache2 server ? [22:00] or [22:00] orudie: scroll up in your terminal, probably you can see the last pems [22:01] but i havent changed em [22:01] this is the first time [22:02] maybe there is an undo command for the chown -R ? [22:02] jmedina, ^ [22:02] What is the prefered way to share disk space from a ubuntu server to many ubuntu clients? [22:02] normally www-data owns /var/www/ [22:02] Deeps, so should i try chown -R ww-data /var/www ? [22:03] w [22:03] Deeps: yeap the problem is that drupal needs to write in some files, and if ehe chown -R I think is insecure [22:03] orudie: you should read the drupal manual and check the perms you need [22:03] orudie: could try, yep [22:03]