[00:14] Aison, what's your setup [00:16] well, i'm using an 8port adaptec raid controller with 8x1TB in raid6 mode [00:16] so i've got 6TB [00:17] the lan adapter is a dualport intel pci-express server nic (running in bonding mode with the core switch) [00:17] but I guess the bottleneck is this samba [00:17] ;) [00:21] damn, it's late, i have got go now, i'm back tomorrow ;) [00:21] alright [00:21] try doing a file copy with `iostat -k 2` running and watch the kb/s === chuck_ is now known as zul [00:31] SpamapS: ok, cool - i think it would be best if you'd briefly comment on how you think they should be handled, and that you will do it when you have time if noone else does first, then unassign, that would be ideal [00:40] hi === dendrobates is now known as dendro-afk [01:23] Hello, I'm having a weird (packaging?) issue with munin-node on 10.04 [01:23] for some unknow reason the config file /etc/munin/munin-node.conf is missing [01:23] even though dpkg -c /var/cache/.../munin-node.deb indicate that this file is in the package [01:24] If a conffile is deleted, easiest way to get it back is to *purge* and re-install. [01:24] and dpkg --info indicate the same [01:24] I have tried to purge and reinstall without success [01:24] removing or reinstalling won't get it back because it assumes you deleted it for a reason [01:24] munin-node.conf is still missing [01:24] dsalvetti: in that case I don't know. [01:24] FWIW I recently switched from munin to collectd, and it's much nicer IMO [01:24] how can dpkg know that I have deleted the file? (vs it's a fresh install) [01:25] dsalvetti: because the file isn't there [01:25] twb: thanks for the tip I'll check it out [01:26] but if I purge apache (for instance) it will remove apache.conf [01:26] and apache.conf will come back is I reinstall apache [01:26] Yes, which is why I said purging should work [01:26] how is that different in the munin-node case? [01:26] purging removes conffiles, remove leaves them as-is [01:26] purging does not wor [01:26] *work [01:26] dsalvetti: and as I said, I don't know why that is [01:26] oh sorry I misread you before [01:27] No problem [01:27] I can easily fix my problem by copying munin-node.conf from somewhere else [01:27] but I would like to see if there is a way to debug the issue [01:28] anyone has any suggestion of what I could do? [01:28] http://paste.debian.net/111774/ <-- my notes for collectd [01:28] Re debugging, I have no ideas [01:30] twb: thanks for the link, what do you prefer in collectd over munin? [02:10] Good evening... is anyone here familiar with the Canonical Ubuntu images on EC2? [02:12] !anyone [02:12] A large amount of the first questions asked in this channel start with "Does anyone/anybody..." Why not ask your next question (the real one) and find out? See also !details, !gq, and !poll. [02:15] Hah, thanks for the reply... I rebooted a instance based on AMI ami-2c57a545, and when the instance came back up my /etc/passwd was reset and the user I'd previously created was gone. How can I presist users across reboots? [02:17] I don't know anything about EC2, but an obvious way would be to store auth details on another server, i.e. LDAP, kerberos or NIS [02:18] Wow, thanks for wasting my time twb... that's why I asked if you knew EC2. I have no intention of doubling my costs on EC2 by running another instance just to store auth info. [02:18] Plonk. [02:19] CompDan: I would back up your data and then try dist-upgrade [02:20] 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. [02:21] I'm guessing this issue has something to do with cloud-init, but I know nothing about cloud-init and the documentation seems nebulous at best... :-/ [02:21] CompDan: are you using a EBS volume based ami? [02:21] http://uec-images.ubuntu.com/releases/10.04/release/ [02:21] Yes, ami-2c57a545 - ebs/ubuntu-images-testing/ubuntu-lucid-daily-i386-server-20110322 === Theravadan|afk is now known as Theravadan [02:29] Hrmm.... now I can't recreate it... great. [02:30] Well, I guess nevermind. If it recurs, I'll be back. :-/ [02:31] CompDan: yea I couldn't reproduce it either with the same ami....I would use the ones from this list http://uec-images.ubuntu.com/releases/10.04/release/ [02:43] Going to rebuild with one of the release AMIs and test this again... Thanks for letting me bend your ear (so to speak) thesheff17. [02:57] smoser: 9.04 to 10.04 isn't a supported upgrade path (re your blog post) [02:58] where did you see that ? [02:58] it was probably a typo [03:00] ScottK, ^ [03:17] smoser: In your update note on http://ubuntu-smoser.blogspot.com/2010/04/upgrading-ebs-instance.html [03:18] smoser: Except on review I see I misread it. [03:18] ah. yeah. [03:18] No. No I didn't. [03:18] that part wasnt an update, but could hav been more clear, yes. [03:18] This same basic process should also allow you to upgrade across a release, perhaps from a 9.04 Alestic instance to Ubuntu 10.04 LTS. [03:18] you'd have to go 9.04 -> 9.10 -> 10.04 [03:19] Yes. [03:19] You and I know that, but non-developers wouldn't. Not a big deal since it's not part of the update. [03:27] I am using an EC2 instance to, among other things, generate an image. I am ssh'ed into the EC2 instance. What's the easiest way to display that image on my local screen? 'display' is an ImageMagick program and that's crazy huge for showing a png to the screen. What else should I use? [03:29] apache and a browser? :) [03:30] [apache and a browser] That's my backup plan. [03:32] Sigh, the micro instances are pretty darned slow. :-) [03:47] Just out of curiousity, is it possible to force a package install via apt-get (like mysql-server) that has interactive dialogs to be non-interactive? [03:51] CompDan: look at this last post http://ubuntuforums.org/showthread.php?t=867747 [03:52] Hee. I first had to do that with Solaris and Java. Bah. [03:55] ah, thanks. I feel like such a n00b, all my experience is on non-Linux UNIX platforms, lol. [03:55] CompDan:http://padwasabimasala.posterous.com/non-interactive-scripted-mysql-install-on-ubu [04:02] Hi, I'm trying to install some packages by using the apt-get install command, but everytime I do that it says: The following packages have unmet dependencies: libwxgtk2.8-0: Depends: (tons of other packages listed) [04:02] Everytime I use it, that package pops up. [04:04] Tommy_K: did you do apt-get update && apt-get upgrade i'm pretty sure it will just install it. [04:04] Nope, tried it. [04:04] Earlier one of my team members had to install that package, which was a .deb [04:05] Tommy_K: well I see it in the repo libwxgtk2.8-0 - wxWidgets Cross-platform C++ GUI toolkit (GTK+ runtime) [04:05] Well, I did as well.... usually I check the repository, but this time it wouldn't take it [04:06] And my team member said he tried to use apt locally on that .deb he downloaded [04:06] Could that have caused it? [04:06] usually it install dependencies. wonder if there is a problem with multiple version on one machine. [04:07] you can't just do apt-get install libwxgtk2.8-0 [04:07] Tommy_K: also I would make sure you used the right repo for the right version. [04:07] of ubuntu. [04:07] It migth be worth apt-get remove libwxgtk2.8-0 followed by apt-get install libwxgtk2.8-0 [04:08] Tommy_K: run "yes y | aptitude --simulate install" and pastebin the output. [04:09] That worked [04:09] @ Mathuin [04:09] Yay! I win! :-) [04:09] Never mind, then :-) [04:09] :) [04:09] Thanks much! [04:10] You are welcome. [04:57] hallyn: What I've been doing is tagging such bugs with the tag 'upstart' .. which jhunt and I review every thursday. [05:16] is it possible to ssh -X to view a virtualbox window on my server? [05:22] CoolestGeek: Some versions of VirtualBox support RDP, so you could tunnel that over SSH and use rdesktop to see it, I think. [05:23] yea I know that the virtualbox i'm using supports RDP, as I rdesktop into a headless windows XP [05:24] however, dumbass here deleted the xp.vdi (the virtualbox image) [05:25] and want to set it back up, it can be done via the cli, but it takes ages compared to the GUI mode of setting up a virtualbox machine, so I would really like to set it up via a GUI, but it's not loading via -X. It did originally, as that's how I set it up to begin with.... [05:29] CoolestGeek: OK, so it is the virtualbox Manager window that you want to use remotely, not a virtual screen of a VM. I've never tried it, but it seems like it "should" work like any other X application. === dendro-afk is now known as dendrobates [05:30] yea, it's the manager window I need, the Virtualbox did work it before [05:30] but it was about a year ago I set it up [05:30] i deleted it totally by accident today :S [05:31] CoolestGeek: Ok, so check you can ssh -X other GUI apps from that machine, and then try it with the virtualbox manager, and if it gives you an error, paste the error for us to see. [05:31] when I run virtualbox once i've connected via ssh and used -X it just waits.... [05:31] it does work with other apps [05:32] i've got transmission working now [05:32] it doesn't give an error [05:32] xlogo is a good test program for ssh -X [05:33] vbox is a GTK2 program IIRC, which will be noticably laggy even over a 10mbps line. [05:33] hmm well the connection here is shit (australia) [05:34] CoolestGeek: It might be easier to script the VM setup using the command line interface, rather than fix the issue with remote use of the virtualbox GUI :) [05:34] so maybe I wait till I'm on a faster line and retry [05:34] hmmm yes, but i've not done it like that before, so am a little weary [05:34] but I suppose i will learn things. [05:34] CoolestGeek: oh, you can use it, it'll just be slow and you need to be accordingly patient [05:35] yes, I remember from the 1st time, I was in the UK then and had a considerably faster connection [05:35] so... as we're on the subject of this, does it post a security threat having a virtualbox accessible? [05:36] the XP I had required a login, but is it safe? [05:36] IMO Windows machines are not safe on ANY network, end of story. [05:37] The company has openly said that they do not consider security important for their Windows line [05:37] yea, but I need it on the server to run various crapware not available to linux [05:38] anyway... so you recommend setting up via the cli [05:38] IMO sysadmins should not need nor use a GUI for their sysadmin role [05:38] it's odd, as transmission is running over X now, slow, but still running [05:38] (If they have one on their desktop, I guess that's OK.) [05:39] virtualbox just waits.... [05:57] CoolestGeek: another way to do it is to run a standalone vnc server on your virtualbox host and then tunnel in a vnc client over ssh. [05:58] and then run the virtualbox GUI manager within the vnc session. [05:59] hmmm [05:59] so... I install vnc onto the server [06:00] then from my local machine i use vnc to connect to that? and set it up? [06:00] yep. tunnel the vnc session over ssh. [07:02] Can someone tell me why so many people use Alestic AMI's rather than those found on http://uec-images.ubuntu.com/ ? [07:04] Alestic used to support the ami...now they are supported by ubuntu devs. use http://uec-images.ubuntu.com/ [07:04] Oh, ok, that makes sense. Thank you! [07:07] itsTodd: Alestic is a great resource though as well. [08:05] Morning everyone [08:10] evening for me, but g'morning to you :) [09:12] Hi. I need to avoid mysql to be started on boot. Am new to this startup jobs. /etc/init/mysql.conf exist. If I comment the first start lines, mysql is not started. But then "/etc/init.d/mysql start" does not seems to work. [09:19] Blinkiz: you probably only want to comment the 'start on' lines, nothing else. Then start mysql with '$ sudo start mysql' [09:19] TeTeT, thanks for the suggestion. It was a apparmor problem in the end.. [09:23] hi ! http://paste.ubuntu.com/584698/ <= I would like to know why the pam module "pam_script.so" is not called? [09:39] Fricass, maybe because you met one of the previous filters and they're set as sufficient? [09:40] yep indeed raphink but I would like to leave them as sufficient and called in all cases pam_script.so [09:40] don't know if it's possible.. [09:40] sure [09:40] you'll have to put it as required on top of the stack [09:40] this way, login will fail unless pam_script.so succeeds [09:41] Yep But pam_script MUST be called after the operations done by pam_unix or pam_perso .. [09:41] That's the point :( [09:41] hehe [09:41] let's see [09:41] then pam_unix.so and pam_perso.so are not sufficient [09:41] they're required [09:42] Fricass, so you would like to make sure one of the two (pam_unix | pam_perso) are matched [09:42] and THEN make sure pam_script is also met [09:42] right? [09:42] exactly [09:43] as in [09:43] if pam_unix fails, pam_perso doesn't fail, and pam_script doesn't fail, it's ok [09:43] if pam_unix doesn't fail, pam_perso fails, pam_script doesn't fail, it's ok [09:44] yep [09:44] and what if both pam_unix and pam_perso fail? [09:44] can it still succeed if pam_script succeeds? [09:44] no [09:44] i can't [09:44] it can't [09:44] alright [09:44] a truth matrix can help to clarify the problem [09:45] in my opinion it's not possible :/ [09:47] hmmm [09:51] Fricass, what do pam_perso and pam_script do? [09:52] pam_script uses a file from pam_perso or use the authentication token from pam_unix [09:52] one of my idea is to remove pam_script and called a script at the session opening raphink [09:53] and leave pam_perso and pam_unix as sufficient [09:54] I'm guessing you wrote both pam_script and pam_perso [09:54] since they don't seem to be standard [09:54] :) [09:54] right [09:55] hmmm [09:55] how about [09:55] making pam_unix.so and pam_perso.so optional [09:55] instead of sufficient [09:55] this way they don't allow to success or fail, they're just run [09:56] optional - the success or [09:56] failure of this module is only important if it is the only module in the stack associated with this service+type. [09:56] so they will be run [09:56] but eventually, it's pam_script that will be required and decide whether auth succeeds or not [09:56] no pam_script just runs commands it doesn't decide [09:57] only pam_perso or unix decide [09:57] that's not possible [09:57] pam_script is required, that means it must return a status [09:57] the goal of the stack is not to run commands, it's to authenticate === xampart_ is now known as xampart [09:58] so indeed if you only want to run commands, maybe using sufficient for pam_unix and pam_perso and running pam_script appart would be better [09:58] Indeed you're right [09:59] But I need to write a shell script which is called after the auth process and that can take the $userid and $password .. [10:00] And I don't see how to do that with a simple shell script (easier with a pam module : PAM_AUTHTOK, PAM_USER ..) [10:00] may I ask what is your need? [10:02] I need to take the $user and $passwd ; then depending on $user I use $passwd to do some stuff with partitions. That's the goal of my pam_script. [10:02] If I can write a shell script that can take $user and $passwd it would be great I think [10:03] sorry for insisting, but is there a way you could do without the password? [10:03] No [10:03] like using sudo... [10:05] No because the password provided is not the standard one, it's used with smartcard [10:05] ok [10:07] I'm stuck raphink ! [10:08] you need the password to manipulate the partitions? [10:09] yep I need to get it [10:11] I think you're stuck indeed ;-) [10:11] I'd really try to get rid of the passwd restriction [10:12] :( [10:12] once the user is authenticated in the system, there must be a way to use another auth token for the parititions, no? [10:17] no I need the one used in authentication [10:21] New bug: #741589 in tomcat6 (main) "tomcat cron deletes app log files after 14 days" [Undecided,New] https://launchpad.net/bugs/741589 [10:21] Fricass, I'm out of ideas for you ;-) [10:22] np raphink I thank you for your time [10:23] raphink: once logged, you don't know if it's possible to get the password provided on auth system? [10:23] I would hope not ;-) [10:23] that would be quite an issue [10:24] imagine if you could get my password once I'm logged, you could use my sudo rights... [10:26] yep [10:30] what kind of partition stuff do you need that requires a password? [10:37] Fricass, hi there. Notice some questions how a script can ask for a password. It's easy in batch. Just use something like "read -s passwd". -s means it will not be echoed in stout. === jjohansen is now known as jj-afk [10:37] batch = bash [10:38] Fricass, Then use can use that variable to put into something else. Like it do: echo "$passwd | cryptsetup luksOpen /dev/disk/by-uuid/fe9acab3-1434-43a6-9c8a-73f7d1d0e2bd 500G" [10:39] Blinkiz, Fricass doesn't want an interactive script, he wants to grab the password from pam [10:39] so users only have to authenticate once [10:39] raphink, aha, gave it a try anyway :) [10:39] I hope not pam can tell my password to a script. Kinda security risk [10:40] even root? [10:40] even root [10:41] even pam_unix uses shadow, which contains encrypted passwords, not clear text [10:41] root shouldn't need my password. he can use my privs without it [10:41] (unless when shadow is not activated, but the passwords are still encrypted anyway) [10:42] Fricass, clear text passwords are not supposed to be accessible anywhere [10:42] raphink: I got an idea ; I create a pam_check which is required and checks some results [10:43] then pam_unix, pam_perso are optional, pam_script required (always OK) and pam_check required [10:43] then pam_check will result the auth process [10:43] where do you put pam_check? [10:43] on top of the stack? [10:43] bottom [10:43] I do something similar currently to mix pam_unix (including nss-pgsql) and pam_ldap [10:43] since my users in ldap all have a uid > 60000 and my users in pgsql are < 60000 [10:44] hm [10:44] I think it's gonna work [10:44] I put a pam_succeed.so uid > 60000 [10:44] I see [10:48] So I'll try this solution magic raphink ! [10:48] :D [10:48] let me know [10:48] pam_succeed is an existing module raphink ? [10:49] yes [10:49] !! [10:49] see ls /lib/security [10:49] pam_succeed_if.so sorry [10:49] e.g. [10:49] auth required pam_succeed_if.so uid > 60000 [10:49] ok np [10:49] * raphink thinks pam modules need a good documentation in debian/ubuntu [10:50] raphink: do you think it's possible to do a think like pam_succeed_if.so file is-in directory/ ? [10:50] How do I get support for SftpUmask in sshd_config? [10:54] SlimG: install a version supporting this directive, e. g. an openssh server patched with http://sftpfilecontrol.sourceforge.net/ [10:56] joschi: that patch is only for 5.4p1, I have 5.5p1 [11:10] hi guys, in ubuntu-server iptables, is it possible to add the source as a hostname instead of an IP? [11:12] twister004, I do not know the answer to that but I can guess. Firewall Builder uses script to first lookup hostnames that it converts to IP numbers and put it in iptables. Also, iptables/kernel does not really use hostname to make rules. So no, I do not think its possible [11:13] it works just fine [11:13] but it won't give you want you want :) [11:13] dns changes will never take effect [11:25] i just installed ubuntu server 10.04 with raid1 (software) and get error ( target filesystem doesn't have sbin/init ) after first boot, somebody can help me out? [11:34] patdk-lap.. so, if my hostname association changes, it wont work? [11:42] <_ruben> twister004: not without reloading the ruleset after the dns change [12:20] I've posted a bug and required steps taken to fix the default cobbler package on Natty Alpha 3 - https://bugs.launchpad.net/ubuntu/+source/cobbler/+bug/741661 [12:20] Launchpad bug 741661 in cobbler "Web UI does not work from default install (2.1.0~bzr-2009-0ubuntu1)l" [Undecided,New] [12:20] Do I need to post this to the mailing list? never done this before... [12:26] New bug: #741661 in cobbler (universe) "Web UI does not work from default install (2.1.0~bzr-2009-0ubuntu1)l" [Undecided,New] https://launchpad.net/bugs/741661 [12:39] zul: ⤴ [12:40] thank ill take care of it [12:58] hello, dual starup my ubuntu server hagup, any idea how to fix it or i should reinstall, because i don't get any error [13:00] alami: new install? [13:03] pmatulis: yes i mean new install, but the problem is: can i copy iptables configuration? [13:04] alami: you think the hangup is caused by iptables? [13:06] no no the hangup is caused by a update two week ago, i have install some but i forget it, after apt-get update and upgrade, the server stay on startup after bios [13:09] morning all [13:09] hello andreserl [13:10] howdy raphink [13:10] fine thanks [13:10] you? [13:11] pretty good [13:12] alami: iptables rules are created by issuing the iptables command (for each rule). best is to gain access to the list of commands and then apply them to your server === Jonny51 is now known as Jonny5 === Jonny51 is now known as Jonny5 [13:17] alami: but maybe i misunderstand what you mean by 'copy iptables configuration' [13:19] pmatulis, i don't know iptables good, and i have see if there is some to copy this rule, or something like that, but i can list this rule and copy them and do it another time after the new installation+ [13:21] alami: alright. you should also look into ufw [13:21] !info ufw [13:21] ufw (source: ufw): program for managing a Netfilter firewall. In component main, is standard. Version 0.30.0-1ubuntu2 (maverick), package size 140 kB, installed size 804 kB [13:21] ufw == uncomplicated firewall [13:21] i don't understand that ufw but i will google it [13:22] https://help.ubuntu.com/community/UFW [13:23] why i need that? i have right now i webmin acces to the server, i can excute command from there i will only excute iptables --list and copy them to a file.txt [13:25] pmatulis do you understand me [13:26] alami: i think so. but i would stay away from webmin. it can mess things up on ubuntu [13:26] !webmin [13:26] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. [13:26] webmin is evil [13:27] i thing that why my system is hanging up [13:28] hggdh: around? [13:30] just the other day webmin messed up a bind/dns setup [13:30] (as i was made aware) [13:31] webmin doesnt seem reliable [13:32] lol [13:32] Daviey, available? [13:33] hggdh, o/ [13:33] Daviey, we still have the euca bug to work out [13:34] hggdh, yeah... i am tied for the next 2.5 hours at least. :( [13:34] hggdh, did you complete that reinstall last night? [13:34] i don't found also any iptable rule or config files [13:35] Daviey, (a) I can wait; (b) yes and additionally: sometimes -- at least twice, so far -- we *did* get IP addresses [13:42] hi...I can't edit a file, alldough I have permissions to do it. The group has write access and I'm part of the group. the file system is ext4 mounted with 'default' and the files are not immutable [13:47] m_tadeu: look at the permissions of all parent directories. it can also be an apparmor issue [13:51] pmatulis: it's inside /var/www, and I'm part of www-data...the group has write permissions inside that dir and sub dirs...care to explain that apparmor issue? [13:54] m_tadeu: apparmor imposes further restrictions on r/w on some directories. check under /etc/apparmor [13:57] hggdh, Oh great.... I'd rather totally failure or total success... sometimes getting an IP addresses sounds painful to debug. :/ [13:58] Daviey, fun, is it not? Here [13:58] hggdh, It's more than fun, it's total excitement.. [13:59] Daviey here´s what I found so far: when you start fresh euca, all instances fail -- sorta: they start, but never get metadata [13:59] then, magically, the get both metadata *and* ip addresses [13:59] then, also magically, they do not anymore [14:00] during *all* this time euca's own dhcpd fails to start. Consistently. At least we do have a consistent error somewhere [14:01] so... who is providing euca with ip addresses -- and mind you, the correct ones? === Jonny51 is now known as Jonny5 [14:04] hey guys how can i vew a list of cron jobs that i have on my system? [14:04] crontab -l [14:05] and check /etc/crontab and /etc/cron.d/ [14:09] zul: howdy [14:10] kirkland: whats up? [14:10] zul: do you want to take a look at https://bugs.launchpad.net/ubuntu/+source/cobbler/+bug/741661, or would you rather i fix it? [14:10] Launchpad bug 741661 in cobbler "Web UI does not work from default install (2.1.0~bzr-2009-0ubuntu1)l" [Undecided,New] [14:10] kirkland: yeah i was going to add it to a Readme.Debian [14:10] zul: are we taking it as far as we can, programmatically already? [14:11] kirkland: yes i think so [14:11] zul: i'd think we should be able to do the a2enmod and apache2/conf.d changes, no? [14:11] kirkland: well i already talked to clint about this and he thought no [14:12] zul: hmm, i'd think we should be able to do all of the steps he details in there, except for the authentication change [14:12] * RoAkSoAx recalls having seen postinst scripts doing a2enmod [14:12] zul: and that one we could improve upon in 11.10, perhaps with a debconf question [14:12] yeah but we removed them [14:12] zul: at SpamapS request? [14:12] kirkland: doesnt matter to me really if you want go ahead [14:13] zul: yeah, i'd like to get this as absolutely far as we can; probably everything but the auth change [14:13] zul: and even that one, I might add a low priority debconf question for username and password [14:14] I can login with cobbler user/pass right now [14:14] without having to do the htdigest [14:16] kirkland: im open to anything but im on a call right now [14:16] zul: okay, no problem [14:16] zul: mind if I take a crack at it later today? [14:16] kirkland: go ahead [14:16] zul: i'll run my changes by you, if that's cool [14:17] kirkland: yep [14:19] kirkland / zul.. I think we can do it, but to do it smart is harder... :)... essentially.. a2enmod is probably not ideal as we don't remember the state it was before we installed. So if you removed cobbler, we'd leave it not as the user is expecting (if i had a module enabled for another app). [14:19] I thought we did something clever in the conf.d... did that not get committed? [14:20] no it didnt it kind of broke your apache config [14:27] New bug: #741748 in postfix (main) "package postfix 2.7.0-1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 75" [Undecided,New] https://launchpad.net/bugs/741748 === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates [14:29] I can't edit /boot/grub/grub.cfg as root...howcome? [14:31] m_tadeu: you seem to be having a lot of permissions problems. odd. but if you read that file you'll see you're not supposed to hand-edit it [14:31] is there a grub.cfg? [14:31] hwo are you trying to open it? [14:31] how [14:31] sudo -s....then vi grub.cfg [14:33] should work [14:33] what error? [14:34] vi says it's open with readonly permissions [14:35] @ALL: Anyone have TACACS+ (tac-plus) experience? I am confused with the Config File settings [14:37] Im not familure with -s - something to do with the shell? [14:40] as that directed toward me? [14:40] was* [14:40] no [14:40] k [14:40] do you think this is the correct room for TACACS+ help? [14:41] can never tell - you just need one person who knows what it is [14:42] I'm starting to think not a lot of people implement it on linux.......this has been a hard thing to research [14:43] mailing lists are often a good source [14:43] more ppl there [14:43] is there a website for a list of them? How do I use that to reach out? [14:44] all the major distros have them [14:44] try to google ubuntu mailing list [14:44] ok, I'll give that a go, thanks [14:45] I'm on 10.04 LTS, when I set a crontab for root, with it set for 12:00, with: script.sh >> /folder/log, it spawns 5 - 10 instances of the script that needs to run. I can run the command manually and it works fine without issues. Any ideas? Thank you. === dendrobates is now known as dendro-afk [14:45] dasunsrule32: Can you paste/pastebin the crontab line that you used for that job? [15:00] which mail server you guys recomend? [15:01] m_tadeu: postfix [15:02] pmatulis: thanx === dendro-afk is now known as dendrobates [15:28] Pici: [15:28] yes, one sec [15:29] Pici: http://pastebin.com/u1icgeMV [15:30] if you need to see the script too, let me know [15:31] dasunsrule32: Well, that will run every minute when the hour matches 0. This will run once at midnight: http://pastebin.com/6BybMhLw [15:33]