[00:02] poseidon: https://help.ubuntu.com/12.04/serverguide/ [00:14] JonEdney, I alway do aptitude update then aptitude safe-upgrade [00:15] cc77, whats the difference if you dont mind me asking? [00:17] update lists the newest packages. And mind asking, found you something btw: https://help.ubuntu.com/community/AptitudeSurvivalGuide [00:19] JonEdney, basically safe-upgrade just updates all packages. And full-upgrade will also latest ubuntu version. [00:19] so safe-upgrade is what you need. [00:19] Okay, I get that. Thanks for the link also. [00:23] I assigned a static ip to my server but dns is not working. opening resolv.conf in nano the comment shows I should edit that file. But if I do anyway it works. What file should I edit for dns then? [00:24] oops same mistake as before editL [00:24] correction: the comment shows that I SHOULD NOT EDIT that file [00:43] Question - would someone please explain/help me with .deb packaging. I'm using deb helper and i've gone through the PackagingGuide.pdf -- and i still cannot understand how you move files around. For instance, if i want the .deb file when installed to have files in /etc/ and then /home/user/.filename .. where/how do i do that ? would this be the dh_installdeb helper script ? === Lcawte is now known as Lcawte|Away [00:49] pdtpatrick: you should ask that kind of question in #ubuntu-motu [00:49] Good point. THanks [00:50] you're welcome [01:50] I don't suppose there would be a gitorious package [01:50] I've been having the hardest time installing it === YankDown1nder is now known as YankDownUnder === oCean_ is now known as oCean === Myrtti_ is now known as Myrtti [08:46] i need to totally purge and remove all config files for apache2, why doesn't apt-get remove apache2 remove the config files... why doesn't apt-get purge remove the config files. why doesn't apt-get install apache2 --reinstall rewrite the config files? why doesn't dpkg-reconfigure apache2 rewrite the config files... none of the above rewrites the config files even if /etc/apache2 doesn't exist [08:46] unhappy with apt-get [08:46] none of the above rewrites or creates fresh config files if /etc/apache2 does or does not exist. [09:46] Hi all, I'm looking for help with an SSH problem, anyone here? [09:52] Hello [09:53] i need run sample bash script when postfix server send message [09:53] How can do this ? [09:56] You may well be far better asking in a postfix channel [09:56] I'm sure that will be simple === Lcawte|Away is now known as Lcawte [12:16] mysqld won't start [12:16] dumped an 11.10 reloaded on 12.04 [12:17] Plugin 'FEDERATED' is disabled. [12:42] mysqld now started but drupal cannot connect === fenris is now known as Guest83592 [14:33] Does Ubuntu Precise support TRIM on LVM? [14:35] hmm [14:36] yes [14:36] has since 10.10 it looks like [14:37] oh opps, 11.04 :) [14:38] looks like it's automatically/always enabled, for all dm devices, except for crypt, that you have to manually enable it for [14:39] https://www.redhat.com/archives/dm-devel/2011-January/msg00035.html [14:44] hrmf - just started installing precise server from usb stick, and it gives an error "unable to mount cdrom" [14:44] any idea what that might be? [14:44] nope :( wonder if it has to do with that new combined iso/usb image thing [14:45] I haven't done a usb install since 10.04 came out [14:45] o/ [15:23] I did a USB install of 12.04 last week [15:23] on a Mac, no less. Worked fine. [15:23] anybody running mysql 5.0.96 on ubuntu 8.0.4? [15:24] sorry mysql 5.0.96 on ubuntu 12.04 [15:40] Ermm.. [15:41] Why would you be running THAT old of MySQL on Ubuntu 12.04 so current? [15:41] MySQL 5.0 is NOT recommended for use. [17:41] hello, I'm having some trouble with importing templates in Maas. When I issue the command "maas-import-isos" I get "mount: Could not find any loop device." and then "failed to import precise-x86_64 profile precise-x86_64 not found, inheritance not possible". I'm running the Node on which Maas is installed in an LXC container. [17:43] carroarmato0: that's because mounting .iso files isn't allowed by default in the container [17:43] carroarmato0: easiest way to allow it is (sorry it's going to be a bit long): [17:43] 1) sudo cp /etc/apparmor.d/lxc/lxc-default /etc/apparmor.d/lxc/lxc-default-with-iso [17:43] 2) edit /etc/apparmor.d/lxc/lxc-default-with-iso, rename the profile from lxc-container-default to lxc-container-default-with-iso [17:44] 3) still in /etc/apparmor.d/lxc/lxc-default-with-iso, add a "mount fstype=iso9660," to the profile (or whatever fstype you want) [17:44] 4) sudo /etc/init.d/apparmor reload [17:44] 5) edit your container configuration (/var/lib/lxc//config) and uncomment the lxc.aa_profile line and set it to lxc-container-default-with-iso [17:45] 6) restart your container [17:45] I tested that solution for someone else with the same need at the Ubuntu Developer Summit and it worked as expected [17:47] stgraber: thank you very much, I was also suspecting it had to do with apparmour, will give it a try! [17:48] carroarmato0: you can usually confirm it by looking at the bottom of the "dmesg" output, apparmor blocking specific calls are listed in there [18:03] stgraber: hmm tried what you said but I'm not noticing anything from apparmour, but it still fails on the mounting part [18:05] carroarmato0: oh, right, I forgot one step, sorry. You also need to allow lxc to access /dev/loop* [18:06] carroarmato0: grabbing the exact line now [18:06] stgraber: thx :) [18:06] #loop [18:06] lxc.cgroup.devices.allow = b 7:* rwm [18:06] add that to your /var/lib/lxc//config [18:08] stgraber: jup, magic seems to be happening [18:08] hehe :) [18:08] stgraber: aan't maas isn't complaining about missing templates :) thank you very much [18:08] *and [18:09] trying to run a poor man's cloud on my laptop to play around with maas, juju and charms [18:10] aspiring to become a witch doctor casting voodoo on servers [18:14] cool, good to hear it all works fine now [18:26] Anyone know of a package for installing gitorious, or a complete walkthrough for installing it on ubuntu 12.04. The ones I followed with 11.04 seem to throw errors [18:37] As anyone had any problems using ssh to ubuntu servers? [18:38] DigiDaz: nope [18:39] I've been having a lot of problems and I've seen a few recent posts popping up whilst googling for an answer but if there was anything happening on a large scale I would expect you would have seen activity here [18:40] I'm beginning to think its some sort of hack or malware [18:40] DigiDaz: what kind of problem are you experiencing? [18:41] Basically, SSH not connecting, its timing out, this has just started happening in the last couple of weeks but to a number of servers [18:42] DigiDaz: Do you have a firewall running? [18:42] If the timeout in sshd config file is upped to about 10 minutes it eventually gets there, keys work fine, someone elsewhere did say they solved it but didn't specify how, they said it was something to do with pam_unix [18:43] I may well have IP tables running on them [18:43] ssh is working fine for me [18:43] I have the default config, though I only let whitelisted ip addresses connect [18:45] This is wierd, the first alert that I am getting is when servers start failing nightly backup using ssh, thing is these boxes have been running a long time, sometimes never even get logged into for weeks, perhaps months at a time [18:45] Its not like anything at all has changed [18:47] Its also happened on a couple of Debian squeeze boxes too, again this has been the last couple of weeks on boxes that have been running for months [18:48] DigiDaz: have you checkt the logs on those boxes? People who have recently logged in, the commands they issues? [18:48] *checked [18:48] No, one logs in, except me, I have checked logs, done searches for modded files etc [18:50] I have one now I can't get into, I'm just gonna reboot it into safe mode in a minute and then change the ssh timeout, I know I will get back on but its wwhats causing this thats beginning to bother me [18:50] is it just ssh which takes a lot of time? As in: does the network feel slower? [18:50] It will probably take about best part of 10 minutes to log in [18:51] How long does it take you to ping the host [18:51] If I then change it to using keys to log in it back to fast as ever [18:51] ping times are fine, everything is fine, its just password based login thats turned mega slow [18:51] ah, probably isn't network connection then [18:52] I have password login disabled [18:52] No, its also multiple servers, multiple data centers etc [18:52] I'm thinking of dong that myself, I know I have been lazy security wise [18:53] heh [18:53] Moreover, even ssh@localhost fails to so its definitely not network related [18:54] I'm lazy and use keys and ssh-agent for everything [18:54] I don't even know what my root password is [18:54] lol [18:54] I have it written down somewhere [18:54] Some long ass thing [18:55] There are a couple of posts flying around with exactly the same problem and they are recent too [18:56] http://askubuntu.com/questions/134298/how-do-i-set-up-gitorious-server-on-ubuntu [18:56] gahh, hope this one works [18:56] I want to try and reinstall pam, see if that fixes it but I don't know how to do it properly, apt get remove it gives some badass warning that you are just about to trash your system === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [19:17] stgraber: have you encountered the problem where nodes would stay in Commissioning permanently? === dendro-afk is now known as dendrobates === dendrobates is now known as dendro-afk [20:20] Anyone here using the dovecot-postfix/mail-delivery-stack packaged setup? I'm having a complication with it not being able to deliver mail to subfolders in the mail using username+extension@mydomain [20:20] It gives me errors trying saying the directory doesn't exist, yet, it does. [21:02] Psi-Jack, never used the combined package, I use them both though [21:02] Hmmm. [21:03] why not use sieve? [21:03] Patrickdk: It's supposed to be, actually. heh [21:03] hmm? what lda are you using? [21:03] dovecot-lda currently. [21:03] what line for it in your master.cf? [21:04] It's not, it's setup as a mailbox_command, by that mail-stack [21:04] mailbox_command = /usr/lib/dovecot/deliver -c /etc/dovecot/conf.d/01-mail-stack-delivery.conf -m "${EXTENSION}" [21:04] oh, your using only local users [21:04] Correct. [21:05] seems too simple [21:05] I'm /also/ trying to get Dovecot to actually use Kerberos binded LDAP connections for userdb+passdb [21:05] but, I'm failing that too, it seems. heh [21:07] odd [21:07] that usecase isn't in the dovecot config for local users [21:07] and I can't believe that is an accident [21:07] heh [21:08] http://wiki.dovecot.org/LDA/Postfix [21:08] are you actually using deliever? or dovecot-lda? [21:08] cause they are largely different [21:09] Yeah.. That particular mailbox_command was setup by the mail-delivery-stack package. [21:09] Ahh, deliver [21:09] like I said, I configure dovecot and postfix, myself [21:09] you don't have a dovecot-lda program at all? [21:09] I have it, yes. [21:09] then your using dovecot-lda, not deliever [21:09] using dovecot 2.x [21:10] http://wiki2.dovecot.org/LDA/Postfix [21:10] best use those examples [21:10] Err, wait. [21:10] dovecot-lda is a 2.x thing, it replced deliever, from 1.x [21:10] I just have dovecot-lmtp [21:10] d [21:10] lmtp is a 2.x thing :) [21:11] This is all Ubuntu 12.04, so it's all very current stuff. [21:11] 12.04 is a dovecot 2.x :) [21:12] heh [21:12] So, no dovecot-lda, just deliver and optionally lmtpd, cause I was considering, once I get this all working, having front-ended smtpds to deliver mail by lmtpd [21:13] heh? [21:13] are you sure? [21:13] the file /usr/lib/dovecot/dovecot-lda doesn't exist? [21:13] dovecot --version [21:14] it should be like 2.0.13 I think [21:15] Hmmm [21:15] Yeah, it does. [21:15] Heh [21:15] deliever is only a symlink anymore, as it doesn't exist in 2.x [21:15] And deliver is a symlink yeah [21:16] I would change that config to the 2.x format, instead of the old 1.x format [21:16] atleast [21:17] mailbox_command = /usr/lib/dovecot/dovecot-lda -c /etc/dovecot/conf.d/01-mail-stack-delivery.conf -f "$SENDER" -a "$RECIPIENT" [21:17] but I still dunno if that -m ${extension} would work or not [21:17] most likely not cause of permissions [21:18] Hmmm, or maybe it needs to be INBOX/${EXTENSION} [21:18] as local users normally go into /var/mail/, where as the extention thing, would require putting it into the user home folder [21:18] No. [21:18] that depends on your dovecot config and folder layout [21:18] All my mail is Maildir in their homes. [21:18] I don't put it into INBOX/* atleast [21:20] Hmmm [21:21] I think I found my problem with LDAP/GSSAPI ;) [21:21] Suwheeeeet! [21:22] So far, it works, but I wonder if it'll properly renew it's Kerberos ticket now. [21:23] May 12 17:23:09 testmail01 dovecot: lda(psi-jack): msgid=<4FAED4BF.5070606@hostdruids.com>: save failed to open mailbox testpool: Mailbox doesn't exist: testpool [21:23] That's the error I get though with delivery. [21:23] May 12 17:23:09 testmail01 dovecot: lda(psi-jack): msgid=<4FAED4BF.5070606@hostdruids.com>: saved mail to INBOX [21:23] That's the end result. [21:24] and what does your maildir folder look like? [21:25] It's got .INBOX.Testpool, and .Testpool directories in it. [21:25] with uppercase? [21:25] Correct [21:27] Oh sheash.. [21:27] May 12 17:27:07 testmail01 dovecot: lda(psi-jack): msgid=<4FAED5AC.7030902@hostdruids.com>: saved mail to testpool [21:27] Why the heck is it case sensative!? [21:27] cause your filesystem is :) [21:28] or just make it sane, and tell dovecot to lowercase everything :) [21:28] it should be smarter than that!! [21:28] use zfs? [21:28] Hmmm [21:28] Well, I /do/ intend to hook this up to my NAS storage home directory. [21:29] But.. That also shares out my /home/& through autofs [21:30] Bah.. That's just silly though.. [21:31] I mean, the idea of using +extensions is for like, signing up to mailing lists with that and not having to bother with actual filtering. [21:33] heh, I don't use it for that [21:33] I use sieve for filtering [21:33] I defently don't want my +stuff autofiltered [21:33] the idea of it though, is so I can easily filter it later if I wanted to though [21:34] atleast the way I use it [21:34] * Psi-Jack nods. [21:34] I think you can make a simple sieve command to autofilter it like that [21:35] not sure, but I think my brother uses it that way === almaisan-away is now known as al-maisan [21:37] Heh [21:37] Well, I just installed a sieve extension to Thunderbird. [21:37] Course, I have NO idea how to do manual sieve. ;0 [21:40] heh, it's hard to get the handle on all the options [21:40] but not hard after that [21:40] Well, I'll be learning it. I'm trying to replace my Zimbra mail server solution with my own more managable, upgradable solution. [21:41] And a /lot/ less resources required for it. heh [21:41] dunno [21:42] my mail server requires 10gigs of ram these days :( [21:42] Patrickdk: 10G ? [21:42] how many users ? [21:42] normally, about 800 imap sessions active [21:42] about 60 active pop [21:42] Patrickdk: heh, mine would never need that much. ;) [21:42] I never looked to see how many per day or whatever [21:43] But, so far, finally have my Kerberos+OpenLDAP solution, with autofs on most of my servers now, even my desktop has Kerberos auth and autofs mounting. :D [21:45] What anyone would recommend for usage with VDS for Ubuntu Server ? I mean some kind of web interface or what ? [21:46] glosoli: and VDS stands for ?¡ [21:47] Virtual Dedicated Server [21:47] Yay! [21:47] great, thanks [21:47] A new bling acronym. [21:48] what exactly is a virtual dedicated server? [21:48] just by it's name, it can't happen :) [21:48] another version Of CLoud Server, virtual private server or whatever, some kind of market trick [21:48] anyway, what web interface anyone of you recommend ? [21:49] recommend not using a web interface [21:50] qman__: hmm just curious why [21:50] ? [21:50] because they're almost universally broken and insecure [21:50] hmm sounds fair, thanks for advice then [21:51] qman__: do you use virtualenvs for server or something like that ? [21:51] I am thinking to deploy OpenERP, not sure if VirtualENV would be smart [21:51] or not [21:51] sounds like buzzword of the month to me [21:51] what are you really trying to accomplish? [21:53] qman__: well, I need to deploy OpenERP in server, with basic modules, and then develop some in some kind of period, the whole thing will be used for production :) [22:01] glosoli, still doesn't mean much to me, I'm not a software developer and ERP is just another buzzword to me, but I see no reason why virtualenv would give you trouble, it's a dev tool [22:01] qman__: well I never did something with clouds servers, I see the dir tree, there is a home dir, and I am not aware if it would be good to create something like /home/workspace [22:01] :? [22:01] but web interfaces for general server management like webmin and friends or website managers like plesk are generally crap, see also !webmin [22:02] qman__: well I trust you, won't even check it :P [23:03] anyone can help me with locales problem "http://dpaste.org/bcrwO/" ? [23:03] http://dpaste.org/bcrwO/ [23:04] I can't seem to find in which file they are located to fix the errors === al-maisan is now known as almaisan-away [23:55] glosoli, iirc, you need language-pack-en and any other language packs you wish to have installed