| erichammond | Patrickdk: Yep, the standard EC2 AMIs are built for pv. EC2 also has an hvm instance type which requires a special hvm kernel, but I'm wondering if a pv kernel could theoretically run there. | 00:25 |
|---|---|---|
| === dendro-afk is now known as dendrobates | ||
| === sixstringsg|away is now known as sixstringsg | ||
| EvilResistance | if i want a certain program to run on boot, but it doesnt have an init.d file, and needs to be run before a user logs in, where should that program's command go? | 02:11 |
| Patrickdk | there is no guarentee it will run before a user logs in | 02:11 |
| Patrickdk | but normall rc.local | 02:11 |
| EvilResistance | Patrickdk: location of that is...? | 02:12 |
| EvilResistance | ahhh there it iis | 02:12 |
| shirt | why is dbus-daemon installed and running by default? is this a critical service? how do i know if i need it? | 02:23 |
| twb | Unfortunately it is :-/// | 02:37 |
| twb | These new-fangled fancy-pants init systems think that a running system dbus is a prerequisite to being able to do things like, oh, mount the bloody filesystem | 02:38 |
| === dendrobates is now known as dendro-afk | ||
| === dendro-afk is now known as dendrobates | ||
| twb | You can easily tell why (as far as apt is concerned) by running "aptitude why dbus" or whatever the package name is | 02:39 |
| === guampa is now known as Guest59369 | ||
| twb | Hmm, it looks like I'm lying, and that's a systemdism absent from upstart. | 02:41 |
| twb | aptitude why dbus on my systems generally indicate it's pulled in by bluez (bluetooth support) or Emacs; and the session dbus (dbus-x11) is pulled in by the GUI web browser. | 02:42 |
| === zz_FridgeBoX is now known as FridgeBoX | ||
| === guampa|2 is now known as guampa | ||
| shirt | twb: so should i remove dbus? | 03:06 |
| twb | shirt: I'm not stopping you | 03:06 |
| === yaboo007 is now known as yaboo | ||
| === FridgeBoX is now known as zz_FridgeBoX | ||
| pmatulis | shirt: is anything wrong with your system? | 03:18 |
| shirt | pmatulis: no, this is a fresh install | 03:20 |
| pmatulis | shirt: then do not remove stuff willy-nilly. learn in-depth before even considering it | 03:21 |
| shirt | only package that needs dbus is accountsservice, and i don't think i need that | 03:25 |
| shirt | hm... that's ultimately needed by ubuntu-standard. i guess i should leave ubuntu-standard installed :\ | 03:35 |
| === koolhead17|zzZZ is now known as koolhead17 | ||
| am_r | any idea how to force unmount a locked up unionfs mount point? | 04:40 |
| am_r | i keep getting "rmdir: failed to remove `dir': Device or resource busy" | 04:40 |
| EvilResistance | am_r: did you try umount -f ? | 04:41 |
| twb | Oh yeah I saw that accountsservice crap on oneiric, I went "WTF is this doing in -standard" | 04:44 |
| twb | It's as bad as plymouth being a hard Depends of init | 04:44 |
| twb | am_r: unionfs and aufs are pretty buggy, I regularly wedge them in such a way as they can't be unmounted short of rebooting | 04:45 |
| twb | am_r: if you're happy to break the system further you can umount -l which will basically pretend to umount it | 04:45 |
| twb | Obviously try umount -f first as EvilResistance suggests, also check lsof | 04:46 |
| am_r | aha that fixed it :) | 04:50 |
| twb | FSVO fix | 04:53 |
| shirt | twb: it actually goes, "ubuntu-standard" -> "language-selector-common" -> "accountsservice / dbus" | 05:01 |
| twb | Yeah I was grumpy when I ran into that one | 05:01 |
| shirt | twb: "accountsservice" also happens to depend on dbus. weird that aptitude why only mentions it | 05:01 |
| shirt | twb: so now all ubuntu servers have to run this useless dbus daemon :| | 05:02 |
| shirt | if people need and it use then they should by all means install it :) but it's there by default now | 05:03 |
| twb | Well it also gets pulled in by stuff like dnsmasq for silly reasons | 05:08 |
| robbyt | Hi All, I'm having some problems with apparmor and mysql on an EC2 box- I updated the apparmor profile for mysql because I moved the datadirectory, I even rebooted, but apparmor still blocks access to the new data directory. I'm running 10.04 | 06:54 |
| eagles0513875| | hey guys any koan and cobbler users here | 07:00 |
| smoser | adam_g, saw this again http://paste.ubuntu.com/800273/ | 07:01 |
| adam_g | smoser: jamespage saved the day on that one | 07:01 |
| adam_g | smoser: not sure i remember what it was, he would. | 07:01 |
| adam_g | smoser: im downstairs btw | 07:02 |
| adam_g | server room | 07:02 |
| smoser | it just "went away" | 07:02 |
| smoser | you retried | 07:02 |
| smoser | it ends up as a result of an apt-get update failure | 07:02 |
| smoser | i think its related to our mirrors issue that we're seeing | 07:04 |
| adam_g | smoser: hash mismatch? | 07:04 |
| smoser | i'm going to lunch then will boether you more. | 07:04 |
| smoser | well, yeah in one way or another. | 07:04 |
| smoser | s/lunch/breakfast/ | 07:04 |
| chrislabeard | I'm trying to run bind on my server, I have my domain and sub domains pointing at the public ip but it seems like the virtual servers are working if I change all the records for the domains to the local IP seems like it works locally but not publicly. | 07:31 |
| chrislabeard | aren't* | 07:32 |
| eagles0513875| | O_O | 07:32 |
| chrislabeard | oh yeah eagle s | 07:32 |
| chrislabeard | wow I can't type | 07:32 |
| eagles0513875| | chrislabeard: why don't you use your domain registrars dns | 07:32 |
| eagles0513875| | i never setup or ran bind | 07:32 |
| chrislabeard | eagles0513875|: I'm going to be creating a ton of sub domains and don't want to login everytime and create a record | 07:32 |
| eagles0513875| | chrislabeard: you will still have to login to the server | 07:33 |
| chrislabeard | thought this would be the easier/smarter way to do this | 07:33 |
| eagles0513875| | not really | 07:33 |
| eagles0513875| | more work imho | 07:33 |
| eagles0513875| | understanding how bind works how to set it up | 07:33 |
| chrislabeard | well I'm using virtualmin that creates the records for me | 07:33 |
| chrislabeard | when I create a new domain | 07:33 |
| eagles0513875| | the domain itself should have an entry @ and then the ip of the server | 07:33 |
| eagles0513875| | at least that is how mine is | 07:34 |
| chrislabeard | eagles0513875|: i have them pointing at my name server | 07:34 |
| eagles0513875| | thats wrong | 07:34 |
| chrislabeard | wait hold on | 07:34 |
| chrislabeard | I mean I have my domain pointing to my name server | 07:34 |
| eagles0513875| | your name server needs to know the ip's of the servers where the sites are on or what ever | 07:34 |
| eagles0513875| | lol try again chrislabeard | 07:34 |
| eagles0513875| | :p | 07:34 |
| eagles0513875| | you have your name servers pointing your domain to that 99 ip | 07:35 |
| chrislabeard | eagles0513875|: yeah sorry its been a long night can't type very well. My name server is pointing to my server's public ip | 07:35 |
| chrislabeard | eagles0513875|: correct | 07:35 |
| eagles0513875| | ok | 07:35 |
| eagles0513875| | you have vps servers on the same machine or different | 07:35 |
| chrislabeard | same machine | 07:36 |
| eagles0513875| | then all subdomains should point to that 99 ip address | 07:36 |
| eagles0513875| | then | 07:36 |
| eagles0513875| | wait | 07:36 |
| eagles0513875| | what are those hosting | 07:36 |
| chrislabeard | k thats how I have it set up | 07:36 |
| eagles0513875| | what is on these machines if you don't mind me asking | 07:36 |
| chrislabeard | dev environments | 07:36 |
| chrislabeard | well they will be | 07:36 |
| chrislabeard | once I get it working | 07:37 |
| eagles0513875| | you running web servers etc | 07:37 |
| chrislabeard | eagles0513875|: right | 07:37 |
| eagles0513875| | the way i have my subdomains setup then isn't a dns issue | 07:37 |
| eagles0513875| | it is an apache issue | 07:37 |
| eagles0513875| | i have a single server hosting about 7 sites | 07:37 |
| eagles0513875| | thats where apache Vhosts come into play | 07:37 |
| chrislabeard | eagles0513875|: Yeah, I'm using apache vhosts | 07:37 |
| eagles0513875| | ok | 07:38 |
| eagles0513875| | so whats the issue | 07:38 |
| chrislabeard | those virtual servers are pointing at the public ip | 07:38 |
| chrislabeard | however when I go to http://wwmcd.org | 07:38 |
| chrislabeard | its not actually showing the right virtual domain it should be saying "testification" | 07:38 |
| eagles0513875| | ok | 07:38 |
| eagles0513875| | easy fix | 07:38 |
| chrislabeard | and http://test.wwmcd.org should be showing "OMG" | 07:38 |
| eagles0513875| | then the vhosts aren't right | 07:38 |
| eagles0513875| | how did you setup the vhosts | 07:39 |
| chrislabeard | virtualmin did, but I might need to customize the templates for them | 07:39 |
| eagles0513875| | the way i do it is setup in sites-available the vhosts then use the ubuntu a2ensite to enable the ghosts | 07:39 |
| eagles0513875| | chrislabeard: can you paste bin me one of the vhosts | 07:40 |
| chrislabeard | yeah just a second | 07:40 |
| eagles0513875| | thanks | 07:40 |
| * eagles0513875| slides chrislabeard a coffee :) | 07:41 | |
| chrislabeard | eagles0513875|: primary domain -- http://pastebin.com/G46gKyUa | 07:41 |
| eagles0513875| | chrislabeard: i need the subdomain if you don't mind | 07:41 |
| eagles0513875| | the primary domain works or not chrislabeard | 07:41 |
| * chrislabeard drinks coffee intently | 07:41 | |
| chrislabeard | eagles0513875|: it does not work | 07:41 |
| eagles0513875| | ok let me look | 07:42 |
| chrislabeard | eagles0513875|: subdomain - http://pastebin.com/Rzfh0f5g | 07:42 |
| eagles0513875| | if it decides to load the pastebin | 07:43 |
| chrislabeard | lol | 07:43 |
| eagles0513875| | found the problem | 07:43 |
| eagles0513875| | for each alias | 07:43 |
| eagles0513875| | you need its own vhost | 07:43 |
| eagles0513875| | for each subdomain you need a vhost | 07:43 |
| eagles0513875| | let me copy you the one of my main business site :) | 07:44 |
| chrislabeard | it creates its own | 07:44 |
| chrislabeard | k | 07:44 |
| eagles0513875| | well its making a big mess of things | 07:44 |
| eagles0513875| | virtualmin = FAILMIN | 07:44 |
| eagles0513875| | how versed are you with commandline | 07:44 |
| Deathvalley122 | cobbler ftw | 07:44 |
| eagles0513875| | Deathvalley122: not what I'm helping him with | 07:44 |
| eagles0513875| | chrislabeard: how versed are you with command line | 07:45 |
| chrislabeard | eagles0513875|: decent | 07:45 |
| eagles0513875| | working via command line is a million times better then web min | 07:45 |
| Deathvalley122 | tbh | 07:45 |
| Deathvalley122 | webmin is falling apart | 07:45 |
| Deathvalley122 | with there development | 07:45 |
| eagles0513875| | Deathvalley122: its a big security hole in ones system | 07:46 |
| chrislabeard | yeah I just don't have the knowledge to right a script to do all this | 07:46 |
| eagles0513875| | chrislabeard: its not all that hard | 07:46 |
| eagles0513875| | once you get one the same concept for all | 07:46 |
| eagles0513875| | just have to adapt it for each directory | 07:46 |
| eagles0513875| | look at mine http://pastebin.com/LPkgiPwi | 07:46 |
| eagles0513875| | very easy very simple | 07:46 |
| eagles0513875| | i have a directory eagleeyet in /var/www | 07:46 |
| eagles0513875| | ignore the stuff with # chrislabeard that stuff is commented out as it doesn't seem to work but the ghosts themselves works fine | 07:47 |
| chrislabeard | the only thing that is confusing me is that I have another fake domain setup that is resolving fine and its IP is set to the server's local IP whereas the other domains are set to public and are not working | 07:47 |
| eagles0513875| | chrislabeard: i created a file in /etc/apache2/sites-available with the vhost information then used a2ensite to create the necessary symlinks etc | 07:47 |
| eagles0513875| | O_o | 07:48 |
| eagles0513875| | chrislabeard: with public ips i have a feeling that has something to do with the vhosts | 07:48 |
| twb | eagles0513875|: congratulations on evading my /ignore for nearly a whole hour. | 07:48 |
| eagles0513875| | :( | 07:49 |
| eagles0513875| | what i do wrong twb :( | 07:49 |
| chrislabeard | eagles0513875|: K i'm gonna give this a few more mins and see if I can find the wrong, if not I'll research scripts for doing this | 07:49 |
| eagles0513875| | chrislabeard: you have an example vhost which i pasted for ya | 07:49 |
| eagles0513875| | of mine | 07:49 |
| chrislabeard | eagles0513875|: yeah that one is really neat and tidy | 07:49 |
| * eagles0513875| is puzzled as to what i did to twb to upset him | 07:49 | |
| ts2 | chrislabeard: did you restart/reload apacahe after adding the virtual hosts? | 07:51 |
| chrislabeard | ts2: I believe so | 07:51 |
| eagles0513875| | ts2: look at his pastes | 07:51 |
| eagles0513875| | virtualmin butchers the vhosts | 07:51 |
| eagles0513875| | with lots of junk | 07:51 |
| eagles0513875| | mine are simple and to the point | 07:51 |
| ts2 | those *admin packages are usually more trouble than they are worth | 07:51 |
| eagles0513875| | agreed | 07:51 |
| eagles0513875| | i had installed webmin on my home server for testing it fudged up my entire installation | 07:51 |
| ts2 | chrislabeard: and they are linked in /etc/apache2/sites-enabled? (just checking some basics) | 07:52 |
| chrislabeard | ts2: not really sure how to check that one | 07:53 |
| ts2 | chrislabeard: look at the output of "ls /etc/apache2/sites-enabled" | 07:53 |
| eagles0513875| | chrislabeard: ls -la in that directory will do it | 07:53 |
| ts2 | you should have a link to the real config file from /etc/apache2/sites-available in there | 07:53 |
| chrislabeard | it shows me 000-default jenna.dev.conf test.wwmcd.org.conf wwmcd.org.conf | 07:54 |
| ts2 | you probably don't want to 000-default one, that's the one that ships with apache, and will be the default when no vhost config exists for a domain | 07:55 |
| ts2 | if you just use http://<ip-adderss>/ for example | 07:55 |
| eagles0513875| | i removed that 000-default and created a new one 000-eagleeyet | 07:55 |
| eagles0513875| | on mine | 07:55 |
| chrislabeard | so i can just delete that one | 07:55 |
| eagles0513875| | chrislabeard: the ghosts are there | 07:55 |
| twb | If you disable 000-default then some other entry will be used as the default when no vhost exists | 07:55 |
| twb | I advise you to keep it and populate /var/www/ with a single index.html file that says something like "sorry, did you want: ..." | 07:56 |
| ts2 | the default is the first (alphabetically) file/link found in /etc/apache2/sites-enabled | 07:56 |
| twb | e.g. http://epoxy.cyber.com.au/ | 07:56 |
| twb | ts2: not quite -- it uses the first one found in the httpd.conf file, which happens to #include all those .conf files in lexicographic order. | 07:57 |
| onre | the thing with web-based admin tools is that usually they sort of "lock you in", as in, you shouldn't touch relevant configurations by hand if you use them | 07:57 |
| twb | ts2: if you were silly and wrote one straight into e.g. /etc/apache/conf.d/security.conf it would be the default | 07:57 |
| ts2 | twb: I'm assuming nothing insane has been done to the default config | 07:57 |
| chrislabeard | I haven't touched the default config other than just install the modules and virtualmin | 07:57 |
| * eagles0513875| waves to twb | 07:58 | |
| chrislabeard | I just don't get why the virtual hosts work fine when I change them to the local ip | 07:58 |
| twb | chrislabeard: if you installed virtualmin you have failed, reinstall ubuntu and start again | 07:58 |
| chrislabeard | lol I know I know its terrible | 07:58 |
| chrislabeard | I just really like GUI | 07:58 |
| chrislabeard | at least web based | 07:59 |
| ts2 | chrislabeard: hmm, you need a "NameVirtualHost" somewhere, and you probably want to replace all the "_default_:80" with just "*" | 07:59 |
| ts2 | unless you want to have https:// versions with different config | 08:00 |
| chrislabeard | ts2: for each virtual host does it need have the wildcard address? or are they supposed to be something else? | 08:00 |
| ts2 | I simply have <VirtualHost *> for *all* virtual hosts | 08:01 |
| chrislabeard | k | 08:01 |
| twb | You cannot meaningfully do a vhost dispatch on 443 unless you have a cert that's valid for all the vhost names (or all your end users ignore cert validation). | 08:01 |
| ts2 | and I have "NameVirtualHost *" in the first, I can't remember why exactly, except that I needed to for some reason | 08:01 |
| twb | ts2: that tells it to do vhost dispatching | 08:01 |
| twb | This discussion probably should move to #httpd | 08:02 |
| chrislabeard | well, thanks for all yalls help. I gotta get some sleeeeep. | 08:03 |
| hallyn | kirkland: just wondering, does mutt not handle gpg on gmail for you? | 08:14 |
| eagles0513875| | hey guys is there an oneric ppa with koan in it cuz the standard 11.10 version is bugged | 08:19 |
| Deathvalley122 | !koan | 08:21 |
| Deathvalley122 | lol | 08:21 |
| ts2 | eagles0513875|: https://launchpad.net/ubuntu/+ppas find out | 08:22 |
| eagles0513875| | thanks ts2 :) | 08:23 |
| eagles0513875| | hey fwereade | 08:26 |
| eagles0513875| | hey guys how can i disable selinux on ubuntu server | 08:54 |
| eagles0513875| | ignore last question | 08:57 |
| === koolhead11 is now known as koolhead17 | ||
| === ejat- is now known as ejat | ||
| eagles0513875| | any selinux experts in here | 09:33 |
| hallyn | yo | 09:33 |
| eagles0513875| | hallyn: you know much about selinux | 09:34 |
| hallyn | zul: failing 0.9.9, do you fancy merging 0.9.8-2? :) I hadn't realized i hadn't done that yet... | 09:46 |
| hallyn | eagles0513875: it's been awhile | 09:46 |
| hallyn | the theory, yes. practice on debian/ubuntu, not that much | 09:47 |
| zul | hallyn: sure i can probably start it this afternoon | 09:47 |
| hallyn | zul: awesome. let me push one more 0.9.7 version right now | 09:47 |
| zul | hallyn: okie dokie | 09:47 |
| hallyn | zul: libvirt_0.9.7-2ubuntu12_source.changes is on its way. thanks. | 09:47 |
| hallyn | i'm going to spend a bit of time trying to straighten out the udd tree! | 09:48 |
| eagles0513875| | well hallyn I'm trying to use koan to provision a guest using koan on 11.10 but have encountered a bug which i have fixed and documented detailed steps to fix it | 09:49 |
| eagles0513875| | but now I'm having a problem with lvm and selinux returning SELinux security context setting to LVM partition failed | 09:49 |
| eagles0513875| | is that something i setup incorrectly in the system in cobbler? | 09:50 |
| eagles0513875| | or an issue with selinux | 09:50 |
| hallyn | more details - what exactly fails. userspace is getting eperm? the *wrong* context is assinged to /dev/dm-X or /dev/mapper/Y? | 09:50 |
| eagles0513875| | basically hallyn thats the issue i am having | 09:51 |
| hallyn | which | 09:51 |
| hallyn | what is the context on /dev/dm-X and /dev/mapper/Y? | 09:51 |
| eagles0513875| | would you like me to paste bin the entire output i get and what I'm passing to koan | 09:51 |
| hallyn | unlabeled_t? | 09:51 |
| hallyn | sure - to pastebin, not here :) | 09:51 |
| hallyn | btw, you might get the best info out of #ubuntu-hardened | 09:52 |
| eagles0513875| | i know :) | 09:52 |
| eagles0513875| | what exactly is that channel | 09:53 |
| eagles0513875| | hallyn: http://pastebin.com/zuKLck1j | 09:55 |
| hallyn | SpamapS: . | 09:56 |
| eagles0513875| | ? | 09:56 |
| eagles0513875| | hallyn: ubuntu-hardened is rather dead | 09:57 |
| hallyn | so is zeus an ubuntu box, running selinux? I know nothing about koan. It sounds like it's selinux policy is out of sync with the host's. | 10:00 |
| hallyn | sorry, not much help | 10:00 |
| eagles0513875| | ubuntu box | 10:00 |
| eagles0513875| | how can i get them in sync | 10:01 |
| eagles0513875| | morning corm | 10:03 |
| eagles0513875| | Cromulent: | 10:03 |
| Cromulent | hi | 10:04 |
| uvirtbot | New bug: #914676 in lxc (main) "Support clean container reboot vs shutdown disambiguation" [Undecided,New] https://launchpad.net/bugs/914676 | 10:26 |
| === alaing is now known as funkymonk | ||
| hallyn | rbasak: how is the lxc testing going? | 10:32 |
| rbasak | hallyn: in progress. looks good so far | 10:33 |
| hallyn | fwiw I tested (on amd64) with lp:~serge-hallyn/+junk/lxc-test | 10:33 |
| eagles0513875| | wish i could say i was having better luck hallyn :( | 10:34 |
| hallyn | rbasak: if you have tests i could add to that script lemme know | 10:34 |
| hallyn | eagles0513875|: does "touch ab; chcon -t virt_iamge_t ab" work? | 10:35 |
| ikonia | eagles0513875|: how are you running SELinux on ubuntu ? | 10:35 |
| uvirtbot | New bug: #914679 in lxc (main) "/run is not setup as tmpfs" [Undecided,New] https://launchpad.net/bugs/914679 | 10:35 |
| rbasak | hallyn: your test script looks good, I'll try that on arm | 10:35 |
| hallyn | the last I knew, we didn't even have a ssh policy working in the standard selinux setup... | 10:36 |
| hallyn | (i'd meant to add it, but then lost my setup) | 10:36 |
| ikonia | that's why I'm curious to how eagles0513875| is running selinux on ubuntu | 10:36 |
| hallyn | yeah | 10:36 |
| hallyn | what does 'id -Z' show? | 10:37 |
| eagles0513875| | its fixed some issues with koan i was experiencing | 10:37 |
| eagles0513875| | unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c255 | 10:37 |
| hallyn | is anything confined? (ps -Z) | 10:38 |
| ikonia | what did ? | 10:38 |
| eagles0513875| | its set to permissive hallyn | 10:38 |
| hallyn | ok: so does my chcon command work for you? | 10:38 |
| hallyn | if not, then my recommendation woudl be to insert a policy module defining that type. | 10:39 |
| eagles0513875| | it returned nothing | 10:40 |
| hallyn | ls -lZ ab shows what then? | 10:45 |
| alex88 | hello guys, i have ubuntu natty server, and i need to upgrade libvips-tools as the version that's in oneiric..how can i do that? it would be nice to use the oniric repo only for that instead of downloading and installing debs | 10:46 |
| koolhead17 | Daviey: hello there | 10:47 |
| SpamapS | alex88: install ubuntu-dev-tools and use 'backportpackage' | 10:47 |
| eagles0513875| | ls: cannot access ab: No such file or directory hallyn | 10:48 |
| === bicranial_ is now known as bicranial | ||
| alex88 | smoser: seems what i need, will it also be updated with apt-get in case it's updated in oneiric repo? | 10:49 |
| alex88 | ups.. it was for SpamapS | 10:51 |
| SpamapS | alex88: no | 10:51 |
| eagles0513875| | wait a min | 10:52 |
| alex88 | SpamapS: says unable to find package libvips-tools in release oneiric.. but it's there in packages.ubuntu.com | 10:53 |
| eagles051387| | i hope that koan is getting upgraded version wise in 12.04 | 11:21 |
| alex88 | i've installed and configure smbd on ubuntu natty server, but "start smbd" says Unknown job: smbd, /etc/init.d/smbd start says the same | 11:24 |
| ikonia | it's just smb isn't it | 11:31 |
| ikonia | ls -la /etc/init.d/smbd will show no such file | 11:31 |
| eagles051387| | ikonia: I'm still at a loss as to why I'm still having issues with koan | 11:33 |
| ikonia | I've not been following your issue | 11:33 |
| ikonia | what is the issue ? (summary) | 11:33 |
| eagles051387| | nut shell is started out as an issue in koan where it wanted a xen kernel which is pre version 3 | 11:33 |
| eagles051387| | which i was helped and got a solution | 11:34 |
| ikonia | ok - so you're using ubuntu what version ? | 11:34 |
| eagles051387| | the next issue once that was fixed was a python one which relates to selinux | 11:34 |
| ikonia | lets get the basic info | 11:34 |
| eagles051387| | 11.10 | 11:34 |
| ikonia | eagles051387|: ok - 11.10 - with what kernel | 11:34 |
| eagles051387| | 3.0 kernel | 11:34 |
| ikonia | the ubuntu stock kernel or a different one | 11:34 |
| eagles051387| | Linux zeus 3.0.0-14-generic #23-Ubuntu SMP Mon Nov 21 20:28:43 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux | 11:34 |
| ikonia | the ubuntu stock kernel or a different one | 11:35 |
| ikonia | just asking to be %100 clear | 11:35 |
| eagles051387| | once i installed selinux it stopped complaining of python bugs in the code now its complaining about lvm | 11:35 |
| ikonia | !info linux-image | 11:35 |
| ubottu | linux-image (source: linux-meta): Generic Linux kernel image.. In component main, is optional. Version 3.0.0.14.16 (oneiric), package size 1 kB, installed size 32 kB (Only available for i386 amd64 all armel powerpc) | 11:35 |
| ikonia | ok - so your kernel is that one ^ ? | 11:35 |
| eagles051387| | correct | 11:35 |
| ikonia | great, so all stock, no changes, | 11:36 |
| ikonia | what's the issue with lvm | 11:36 |
| eagles051387| | when i run koan its giving me an error with selinux and lvm | 11:36 |
| ikonia | what error ? | 11:36 |
| eagles051387| | hold on | 11:36 |
| eagles051387| | now ikonia it seems the error has changed | 11:39 |
| eagles051387| | its a libvirt error now | 11:39 |
| rbasak | hallyn: my lxc test succeeds on your md5 437beee14c4f48006e0d5dd096aa6bcf lxc-rally-p.debdiff, running your tests now | 11:39 |
| hallyn | rbasak: lxc-rally-p6.debdiff is the one i was planning on pushing. (which fixes lxc-ps and lxc-ls, which will probably fail for you with running my tests, but let's see) | 11:50 |
| SpamapS | hallyn: any luck on your libvirt branch? | 11:51 |
| hallyn | rbasak: only lxc-ls is different | 11:52 |
| hallyn | so no need to rerun all tests | 11:52 |
| hallyn | SpamapS: it's going, slowly | 11:52 |
| rbasak | ok thanks | 11:52 |
| eagles051387| | hallyn: you in charge of libvirt python lib | 11:53 |
| eagles0513875 | ikonia: the error im getting is SELinux security context setting to LVM partition failed | 12:30 |
| __godhelpme | hello | 12:38 |
| __godhelpme | i could use some help installing the server edition with raid from usb | 12:41 |
| pmatulis | __godhelpme: what do you mean by 'with raid from usb' | 12:45 |
| __godhelpme | trying to install onto a raid system but usb is my source not cd | 12:46 |
| __godhelpme | here's my problem it asks me if i want to load the raid drivers i say yes but when i go to partition it, it wont see the raid or the drives at all | 12:50 |
| __godhelpme | pmatulis, any ideas | 12:54 |
| kerframil | __godhelpme: why not open a terminal and check the status of the device and whether the appropriate driver has initialised. | 12:57 |
| __godhelpme | kerframil, and how do i do that from the install screen | 12:58 |
| kerframil | __godhelpme: Alt+F2 should work, iirc | 12:58 |
| kerframil | __godhelpme: also, Alt+F4 should show a log (including messages emitted by the components of the installer) | 12:58 |
| __godhelpme | device-mapper target "mirror" is not in the kernal | 13:01 |
| kerframil | that's not relevant | 13:01 |
| kerframil | can't remember is lspci is available from the shell here, but if it is - try running "lspci -k" | 13:01 |
| kerframil | look for the RAID controller there. if it is currently supported/active, it will also show the kernel module supporting it. | 13:02 |
| __godhelpme | what did you want to know from lspci | 13:03 |
| kerframil | which controller you're trying to support would be a good start. maybe you should start with just "lspci" - no parameters. | 13:04 |
| kerframil | look for SCSI controllers | 13:04 |
| __godhelpme | hudson amd raid controler | 13:04 |
| kerframil | hmm, I don't know much about this controller. can you run "lspci -nn | grep SCSI" and paste the line here if possible? | 13:05 |
| __godhelpme | no scsi | 13:06 |
| __godhelpme | only raid | 13:06 |
| kerframil | ok, whatever. the line correlating with your controller - with -nn | 13:07 |
| kerframil | the identifies in square brackets at the end of the line are of interest | 13:08 |
| kerframil | indentifiers* | 13:08 |
| __godhelpme | RAID bus controller [0104]: Amd Hudson sata controller [Raid mode][1022:7802 (rev 40)] | 13:08 |
| __godhelpme | RAID bus controller [0104]: Amd Hudson sata controller [Raid mode][1022:7802] (rev 40) | 13:09 |
| ikonia | eagles0513875: what are you actually doing to cause this error | 13:11 |
| __godhelpme | installing | 13:11 |
| __godhelpme | sorry | 13:11 |
| kerframil | __godhelpme: hmm. my guess is that it's a 'fakeraid' controller. if not in RAID mode, it should be supported by the AHCI driver. if that assumption is correct, it probably needs dmraid. I don't believe in dmraid and would be inclined to use the RAID implementation in Linux instead. | 13:12 |
| __godhelpme | so change it from raid to ahci the setup the raid via dmraid right? | 13:13 |
| kerframil | __godhelpme: almost. if I were in your position and wanted to use RAID, I'd toggle it to AHCI in BIOS but use md RAID (that's the native implementation in the kernel which is controller agnostic). | 13:13 |
| kerframil | __godhelpme: if it were a 'proper' hardware controller, then I would say otherwise. but it doesn't look like it. | 13:14 |
| __godhelpme | md raid? | 13:14 |
| kerframil | yes | 13:14 |
| __godhelpme | explain please | 13:15 |
| kerframil | briefly, md means multiple device. that's the native (software) implementation in Linux. you can assemble an array from component devices (/dev/sda, /dev/sdb ...) into a device which follows a naming scheme like /dev/md0 ... | 13:16 |
| kerframil | dmraid is something entirely different and is used to support half-baked 'fakeraid' controllers in their native RAID mode (which I personally don't recommend) | 13:16 |
| __godhelpme | could you point me to a guide for md raid? | 13:17 |
| kerframil | best off asking someone else. I refuse to use it. | 13:18 |
| __godhelpme | md raid | 13:18 |
| kerframil | a search for ubuntu mdraid will probably turn up some answers | 13:18 |
| kerframil | oh, sorry | 13:18 |
| kerframil | https://help.ubuntu.com/community/Installation/SoftwareRAID | 13:18 |
| kerframil | if you switch to AHCI, the individual drives should show up | 13:18 |
| kerframil | then it should be fairly straightforward | 13:18 |
| __godhelpme | ok sounds good thanks ya it detects the drives thanks for your help | 13:20 |
| kerframil | sure. good luck with your install. | 13:20 |
| eagles0513875 | ikonia: nothing i had to make an addition to a single line in the koan app.py and then all im trying to do is run koan to provision the vps | 13:20 |
| ikonia | ok - so show me the line you've added | 13:22 |
| a_ok | I have set umask in both /etc/login.defs and /etc/profile but I still get the default umask | 13:29 |
| === bladernr_afk is now known as bladernr_ | ||
| === sixstringsg is now known as sixstringsg|away | ||
| gary_poster | hallyn, sent you the details, fwiw. Thanks. | 14:50 |
| hallyn | gary_poster: in email, or bug? | 14:51 |
| hallyn | i see it, thanks | 14:55 |
| gary_poster | cool | 15:05 |
| hallyn | gary_poster: jinkeys. | 15:05 |
| gary_poster | hallyn, heh, what? | 15:06 |
| hallyn | on mine it just completely, cleanly draws the console over X, rather than the ugly effects you were getting. | 15:06 |
| gary_poster | wow | 15:06 |
| gary_poster | still not ideal, sounds like :-) | 15:07 |
| eagles0513875 | ikonia: the and not is what i added but this was giving me the issue about needed a xen enabled kernel which is pre 3.0 kernel version | 15:09 |
| eagles0513875 | ikonia: # correct kernel on dom0? | 15:09 |
| eagles0513875 | if uname_str.find("xen") == -1 and not os.path.exists("/proc/xen/capabilities"): | 15:09 |
| eagles0513875 | raise InfoException("kernel-xen needs to be in use") | 15:09 |
| ikonia | eagles0513875: sorry say that again | 15:14 |
| ikonia | eagles0513875: what are you talking about now ? | 15:15 |
| ikonia | eagles0513875: the last thing you said was you where having a problem with selinux and lvm and I asked for the line you changed, now you are showing me something that checks for a xen kernel ??? | 15:15 |
| uvirtbot | New bug: #914306 in openssh (main) "sorry, the program"parted_server"closed unexpectedly" [Undecided,Invalid] https://launchpad.net/bugs/914306 | 15:16 |
| ikonia | anychance you could backup and explain a little or have I missed a post ? | 15:16 |
| eagles0513875 | you wanted to know what i changed initially | 15:18 |
| eagles0513875 | and the lines above were the issue i was having with it wanting a xen kernel which is pre 3.0 | 15:19 |
| ikonia | ok, so how does this releate to selinux ? | 15:19 |
| ikonia | or your problem | 15:19 |
| eagles0513875 | ahh now all that does is bring on the next problem of it complaining about the python selinux code | 15:20 |
| eagles0513875 | which is fixed by installing selinux | 15:20 |
| eagles0513875 | now when i run this command to have koan provision the guest i get this output | 15:20 |
| eagles0513875 | ikonia: this is the error http://pastebin.com/xyvhE9kQ | 15:21 |
| hallyn | stgraber: ^ control groups appear to be not working for containers, at least for chardevs | 15:22 |
| ikonia | why are you using a xen initrd ? | 15:22 |
| ikonia | eagles0513875: ^ | 15:22 |
| eagles0513875 | O_o | 15:22 |
| ikonia | why are you using a xen model when you are not using xen ? | 15:22 |
| ikonia | or are you using xen ? | 15:22 |
| eagles0513875 | we are | 15:22 |
| ikonia | ok - so you are using xen, great. | 15:22 |
| eagles0513875 | xen is part of 3.0 kernel hence going back to the initial code paste | 15:22 |
| eagles0513875 | where it wanted a pre 2.6 kernel but thats not necessary with 3.0 and later | 15:23 |
| ikonia | ok, I'm starting to follow | 15:23 |
| eagles0513875 | now i have no idea why selinux is having issues with lvm when its set to permissive mode | 15:23 |
| ikonia | eagles0513875: ok - hang on | 15:24 |
| ikonia | eagles0513875: so 1.) why are you doing vgs -o vg_name ? | 15:24 |
| ikonia | do you have a volume group called "vg_name" ? | 15:24 |
| eagles0513875 | no | 15:24 |
| eagles0513875 | and where do i specify that | 15:24 |
| ikonia | so why are you doing that | 15:24 |
| ikonia | look at line 10 | 15:24 |
| eagles0513875 | i specified the volume group name in cobbler | 15:24 |
| eagles0513875 | let me check something in cobbler | 15:24 |
| eagles0513875 | fixed the volume group issue | 15:25 |
| eagles0513875 | now the issue has moved onto lib virt | 15:26 |
| ikonia | eagles0513875: do you have a volume group called vg_free ? | 15:26 |
| hallyn | feh. i think i see. | 15:26 |
| ikonia | ahh I see what you are doing | 15:26 |
| eagles0513875 | im using koan to provision a system i created with cobbler | 15:27 |
| ikonia | vg_free is the option | 15:27 |
| ikonia | why are you using vgs ?? | 15:27 |
| ikonia | yes, I understand that, but you're actual process doesn't seem to tie togther | 15:27 |
| eagles0513875 | for the xen guests | 15:27 |
| ikonia | together | 15:27 |
| eagles0513875 | to be able to increase their size | 15:27 |
| eagles0513875 | would you like me to pastebin the libvirt issue | 15:28 |
| ikonia | so you have a volume group called xenguestlvm ? | 15:28 |
| ikonia | no | 15:28 |
| eagles0513875 | correct | 15:28 |
| ikonia | do you also have an logical volume call xenguestlvm ? | 15:28 |
| ikonia | as well as a volume group | 15:29 |
| eagles0513875 | correct | 15:29 |
| eagles0513875 | this is the partition /dev/sda5 85940224 1953523711 933791744 8e Linux LVM | 15:30 |
| eagles0513875 | VG Name xenguestlvm | 15:30 |
| ikonia | ok - so you have a volumegroup called xenguestlvm, and a logical volume called xenguestlvm | 15:30 |
| eagles0513875 | how do you name the logical volume? | 15:31 |
| eagles0513875 | logical partition | 15:31 |
| ikonia | when you create it | 15:31 |
| eagles0513875 | ahh ok | 15:31 |
| ikonia | so looking at that error, you are specifying a security policy called virt_image_t | 15:32 |
| eagles0513875 | when you give the partition a name you need to give it a location correct? | 15:32 |
| ikonia | however it also looks like you have no se policy called virt_image_t | 15:32 |
| eagles0513875 | ok | 15:32 |
| eagles0513875 | how do i go about creating one | 15:32 |
| ikonia | ??? why are you using it if you don't know how to use it | 15:32 |
| ikonia | don't bother | 15:32 |
| ikonia | just don't use selinux if you don't need it or know how to use it | 15:33 |
| ikonia | eagles0513875: where did you get the info that you should use -t virt_image_t ? | 15:33 |
| eagles0513875 | ikonia: if i dont use it koan throws python errors | 15:33 |
| eagles0513875 | ikonia: it seems to be hardcoded into koan | 15:33 |
| ikonia | are you following a guide ? | 15:34 |
| ikonia | (for the setup of this stuff) | 15:34 |
| eagles0513875 | i had someone helping me in cobbler with the initial problem yesterday | 15:35 |
| eagles0513875 | im just wondering if its worth all this hassel file a bug with what i have fixed so far and push for a newer version in 12.04 | 15:35 |
| ikonia | why file a bug ? | 15:35 |
| eagles0513875 | this was a trunk pull for 11.10 pulled back 2nd june | 15:35 |
| ikonia | who says there is a bug ? | 15:36 |
| eagles0513875 | ikonia: this is a clean install on the server | 15:36 |
| ikonia | ???? so ? | 15:36 |
| eagles0513875 | only ubuntu orchestra server and koan | 15:36 |
| eagles0513875 | btw here is the virt_image_t hard coding | 15:36 |
| eagles0513875 | http://pastebin.com/20khbEqx | 15:37 |
| ikonia | I don't see a bug though | 15:37 |
| eagles0513875 | ok | 15:37 |
| ikonia | ok - so there is a check to see if selinux is enabled, that suggests you can use it without selinux, as if there is a check that means there must be two options, one for se enabled and one for without | 15:38 |
| ikonia | also 0 suggests that selinux is not enabled | 15:38 |
| eagles0513875 | well i disabled se linux and it still didnt work for me | 15:38 |
| eagles0513875 | let me try and disable it | 15:38 |
| ikonia | I may have missed some of your back story, so apologies if I'm asking you to repeat stuff you've said before | 15:40 |
| eagles0513875 | no worries i just really want to get this resolved | 15:40 |
| eagles0513875 | i just disabled it now it seems to be spitting out libvirt errors again | 15:40 |
| eagles0513875 | should i try uninstall selinux all together and put back apparmor | 15:40 |
| ikonia | no, what is it complaining about | 15:42 |
| ikonia | lets look if it's a genuine error, or a miss-configuration | 15:42 |
| roaksoax | smoser: i retried the odev without success | 15:42 |
| smoser | roaksoax, what failed? | 15:42 |
| smoser | i walked through it myself on a fresh install | 15:43 |
| smoser | :-( | 15:43 |
| roaksoax | smoser: nothing seems to fail but rather, I just can't log into the VM | 15:43 |
| lynxman | roaksoax: hey | 15:43 |
| roaksoax | smoser: http://paste.ubuntu.com/800673/ | 15:43 |
| roaksoax | lynxman: howdy man | 15:43 |
| eagles0513875 | ikonia: http://pastebin.com/7JdaBBrf | 15:44 |
| ikonia | eagles0513875: what should be listening on port 8000 ? | 15:44 |
| ikonia | that's the libvirt daemon isn't it ? | 15:44 |
| ikonia | yes it is | 15:45 |
| ikonia | you've not got libvirtd running | 15:45 |
| eagles0513875 | just started it and tried running koan again and its not working | 15:47 |
| smoser | roaksoax, thats output of the build-zimmer stuff ? | 15:47 |
| smoser | the console log? | 15:47 |
| eagles0513875 | ikonia: im wondering if its worth it keeping at this issue when there is a much newer version then what is in the 11.10 repos | 15:48 |
| smoser | can you run that comand i gave yesterday to extract the ovf-env.xml file from the .iso file you're handing it please ? | 15:48 |
| ikonia | eagles0513875: stop saying "not working" and give details] | 15:48 |
| eagles0513875 | ikonia: same issue as what i pasted above | 15:48 |
| ikonia | so still can't connect to port 8000 ? | 15:48 |
| eagles0513875 | correct | 15:48 |
| ikonia | ok, so telnet localhost 8000 what error do you get | 15:49 |
| ikonia | ooh | 15:49 |
| roaksoax | smoser: yes | 15:49 |
| roaksoax | smoser: I added ssh-import-id for my lp id to the cloud-init config, but now It doesn't even seems to be installing openssh-server | 15:50 |
| smoser | roaksoax, ^ can you get me the ovf-env.xml file | 15:54 |
| roaksoax | smoser: hold on, can't pastebinit | 15:57 |
| === andreas__ is now known as ahasenack | ||
| hallyn | aHA - gary_poster: it's overlayfs itself. It doesn't do the LSM checks. I can put myself into a locked down device cgroup, overlay-mount /dev to /mnt, and then I can write to /mnt/tty7 but not /dev/tty7 | 16:01 |
| hallyn | apw is going to look into why | 16:01 |
| eagles0513875 | ikonia: any ideas | 16:02 |
| gary_poster | hallyn, ah-ha! great, thanks. Right, that's what William had guessed. Thanks! | 16:02 |
| roaksoax | smoser: http://people.canonical.com/~andreserl/ovf-env.xml | 16:03 |
| smoser | roaksoax, your user-data is "hi world" | 16:04 |
| smoser | you didn't copy and paste | 16:04 |
| smoser | :) | 16:04 |
| smoser | ../../../build-info.txt | 16:05 |
| smoser | err... | 16:05 |
| smoser | ../../ud-build.txt | 16:05 |
| ikonia | eagles0513875: sorry, one moment please. | 16:07 |
| ikonia | eagles0513875: can you please telnet localhost 8000, what do you get | 16:07 |
| eagles0513875 | connection refused | 16:07 |
| eagles0513875 | crapola | 16:07 |
| eagles0513875 | wait | 16:07 |
| eagles0513875 | the DC has a firewall setup to provide anti Ddos protection and all that would the port need to be opened on that | 16:08 |
| eagles0513875 | ignore that i just realized i used the local host machine that is having this issue | 16:08 |
| roaksoax | smoser: heh... ok will re do it lol | 16:08 |
| eagles0513875 | ikonia: i dont have any firewall setup on the local host no ufw or iptables | 16:09 |
| ikonia | "localhost" | 16:09 |
| ikonia | there should be no firewall | 16:09 |
| ikonia | eagles0513875: ps -ef | grep virtd | 16:10 |
| eagles0513875 | root 1150 1 0 02:16 ? 00:00:00 /usr/sbin/libvirtd -d | 16:10 |
| ikonia | ok, so the issue is why can't your koan setup invoke the connection to libvirtd | 16:11 |
| ikonia | re-looking at your log, the xen daemon wants to talk to libvirtd, what on your setup could stop that ? | 16:12 |
| eagles0513875 | nothing | 16:12 |
| ikonia | does your machine run a gui ? | 16:12 |
| eagles0513875 | only thing i installed was the ubuntu-orchestra-server | 16:12 |
| ikonia | as in a desktop of some sort ? | 16:12 |
| eagles0513875 | no it does not | 16:12 |
| ikonia | that's annoying | 16:12 |
| ikonia | ahhh, hang on | 16:13 |
| ikonia | can you pastebin the output of "netstat -a | grep LIST | grep 8000" please | 16:14 |
| eagles0513875 | ikonia: no output is returned | 16:15 |
| ikonia | eagles0513875: grep 8000 /etc/services | 16:16 |
| eagles0513875 | nothing returned there either | 16:17 |
| ikonia | that's fine, that's not a worry, ok, so libvirt isn't listening, there is the problem | 16:18 |
| ikonia | so that is the root cause to this problem at this moment, libvirtd is not listening | 16:18 |
| eagles0513875 | ikonia: what do i need to do to fix the problem | 16:19 |
| === dendrobates is now known as dendro-afk | ||
| smoser | adam_g, lp:~smoser/+junk/check-archive is the check-archive script | 16:22 |
| ikonia | eagles0513875: depends on the reason | 16:25 |
| eagles0513875 | what can we do to determine it | 16:25 |
| ikonia | come on - you should know basic debugging | 16:26 |
| ikonia | you're selling your services as a professional Linux admin for paying customers | 16:26 |
| ikonia | what have you checked so far ? | 16:26 |
| === dendro-afk is now known as dendrobates | ||
| ikonia | eagles0513875: are you there ? | 16:32 |
| eagles0513875 | ya i am helping dad on skype really fast | 16:35 |
| ikonia | eagles0513875: it's worth checking what the listen address is for libvirtd - it may be set to all interfaces or a speific one, so if you try to connect to it on localhost, but it's saying listening on 2.2.2.2 instead | 16:35 |
| eagles0513875 | i hear ya | 16:37 |
| ikonia | just reading the docs now and it looks like libvirt listening on a socket | 16:38 |
| ikonia | so netstat -a won't show it on port 8000 it will show it as a socket file | 16:38 |
| ikonia | so netstat -a | grep libvirt should show the socket file | 16:38 |
| eagles0513875 | its listening on 8154 and 8156 | 16:40 |
| eagles0513875 | ikonia: http://pastebin.com/JarXUYCF | 16:41 |
| ikonia | ok, thats good, so it is listening | 16:41 |
| eagles0513875 | is it listening on the right socket though? | 16:43 |
| eagles0513875 | what is the 8154 and 8156 | 16:43 |
| ikonia | just reference numbers, think of it as process id's | 16:45 |
| ikonia | so the question is why can't xen connect to it | 16:45 |
| ikonia | (also check in your libvirt config file that libvirtd is not bound to 1 ip and localhost is valid) | 16:45 |
| eagles0513875 | ikonia: from the looks of it its not bound to anything specific | 16:51 |
| eagles0513875 | http://paste.ubuntu.com/800749/ | 16:51 |
| ikonia | eagles0513875: great, it's on all interfaces, so localhost should be fine | 16:51 |
| ikonia | the question is why can xen not connect then. | 16:51 |
| eagles0513875 | could it be the way i have bridging setup | 16:52 |
| eagles0513875 | i was told in xen not to use xend scripts as those tend to be buggy but setup the bridge in the network interfaces file | 16:53 |
| smoser | roaksoax, did you try again ? | 16:54 |
| ikonia | I guess a question that I can't see the answer to, is how is xen trying to connect ? | 16:54 |
| eagles0513875 | gd question | 16:54 |
| ikonia | I don't use xen so don't know how it normally connects | 16:55 |
| roaksoax | smoser: on it | 16:56 |
| eagles0513875 | ikonia: im asking in xen for some help | 16:57 |
| eagles0513875 | appreciate the help though ikonia :) | 16:57 |
| * eagles0513875 slides ikonia a coffee | 16:57 | |
| ikonia | eagles0513875: out of interest, check the permissions on the socket files | 16:57 |
| eagles0513875 | ikonia: srwxrwx---. 1 root libvirtd 0 2012-01-11 02:16 /var/run/libvirt/libvirt-sock and srwxrwxrwx. 1 root libvirtd 0 2012-01-11 02:16 /var/run/libvirt/libvirt-sock-ro | 17:00 |
| Pici | 60 | 17:00 |
| eagles0513875 | Pici: ? | 17:00 |
| ikonia | eagles0513875: try virsh -c qemu:///system list | 17:03 |
| ikonia | eagles0513875: it's not xen, but it will test your connection to libvirtd | 17:03 |
| eagles0513875 | error: unable to connect to '/var/run/libvirt/libvirt-sock', libvirtd may need to be started: Permission denied error: failed to connect to the hypervisor | 17:03 |
| roaksoax | smoser: this still fails: ( cd cloud-init/doc/ovf/ && ./make-iso ovf-env.xml.tmpl ../../../ud-build.txt -o - ) > build.iso | 17:03 |
| roaksoax | wrote iso to - | 17:03 |
| === sixstringsg|away is now known as sixstringsg | ||
| === sixstringsg is now known as sixstringsg|away | ||
| smoser | roaksoax, how does it fail? | 17:05 |
| eagles0513875 | ikonia: actually needs sudo | 17:05 |
| smoser | paste that console log and that ovf-env.xml | 17:05 |
| smoser | please | 17:06 |
| eagles0513875 | but when i run koan command i am using sudo which shouldnt give me this issue or so i would think it shouldnt | 17:06 |
| ikonia | eagles0513875: yes, sorry, I'm not using ubuntu, I forgot sudo | 17:06 |
| ikonia | eagles0513875: why are you now referncing koan ? and why would sudo remove the problems with libvirtd ? | 17:07 |
| roaksoax | smoser: the ovf-env-xml.tmpl is the same as I linked a bit ago | 17:07 |
| roaksoax | there's no console log | 17:07 |
| smoser | i'm confused. | 17:07 |
| roaksoax | smoser: http://pastebin.ubuntu.com/800784/ | 17:08 |
| smoser | ok. | 17:08 |
| smoser | so that looks fine. | 17:08 |
| smoser | and when you start kvm with tha t? | 17:08 |
| Deathvalley122 | it could be | 17:08 |
| eagles0513875 | ikonia: ignore my last comment regarding sudo | 17:08 |
| smoser | you get no console output? | 17:08 |
| Deathvalley122 | that is needs to be chmod | 17:08 |
| Deathvalley122 | it** | 17:08 |
| eagles0513875 | Deathvalley122: shhh im working on getting this issue resolved :p | 17:09 |
| Deathvalley122 | -is | 17:09 |
| Deathvalley122 | j/s | 17:09 |
| roaksoax | smoser: if I do this, it works: http://pastebin.ubuntu.com/800785/ | 17:09 |
| smoser | roaksoax, i dont understand how those 2 things are different | 17:10 |
| roaksoax | smoser: never mind :) | 17:11 |
| smoser | can you get the ovf-env.xml out of both ? | 17:11 |
| roaksoax | smoser: http://paste.ubuntu.com/800788/ | 17:11 |
| roaksoax | build-console | 17:11 |
| roaksoax | build-console.log | 17:11 |
| roaksoax | smoser: ok I think I found the problem | 17:13 |
| roaksoax | smoser: for some reason it didn't finish building the image | 17:13 |
| roaksoax | not it seems to resume normally | 17:13 |
| smoser | roaksoax, can you get the ovf-env.xml file for me to see? | 17:14 |
| roaksoax | smoser: there's not ovf-env.xml but there's a ovf-env.xml.tmpl | 17:14 |
| smoser | no. there is ovf-env.xml inside the iso | 17:15 |
| smoser | roaksoax, isoinfo -i build.iso -x "/OVF_ENV.XML;1" | pastebinit | 17:16 |
| roaksoax | smoser: http://people.canonical.com/~andreserl/ovf-env.xml | 17:16 |
| smoser | that says "hello world" | 17:17 |
| smoser | roaksoax, the command you're running inside cloud-init takes the user-data in 'ud-build.txt' and shoves it inside of an iso it creates | 17:17 |
| smoser | the one you've got there has user-data that says 'hello world' | 17:18 |
| roaksoax | smoser: right, but that file is not being used | 17:18 |
| roaksoax | err where is it being used? | 17:18 |
| smoser | then you're doing something wrong | 17:18 |
| smoser | :) | 17:18 |
| smoser | you create that iso file | 17:18 |
| smoser | and then you run kvm..... -cdrom that.iso | 17:19 |
| roaksoax | smoser: hold on as I said, for some reason kvm could finish the process, but now it seems it has | 17:19 |
| smoser | oh | 17:19 |
| smoser | i know whats wrong | 17:19 |
| smoser | you need to re-create your image file | 17:19 |
| smoser | qemu-img create -b ... | 17:19 |
| smoser | do that again | 17:19 |
| smoser | get a fresh image to kvm | 17:19 |
| smoser | because the one you messed up thinks it has alrady run the first-boot stuff for the instance-id that is inside the ovf file | 17:20 |
| smoser | so cloud-init is not doing anything | 17:20 |
| smoser | that make sense? | 17:20 |
| roaksoax | smoser: ok found the problem. As i presumed the kvm command failed for some reason before it finished doing its cloud init stuff | 17:25 |
| smoser | roaksoax, so theres probably a silent grub 5 second wait before it laods the kernel | 17:37 |
| smoser | i just realized that | 17:37 |
| roaksoax | smoser: seems like it | 17:47 |
| === dendrobates is now known as dendro-afk | ||
| uvirtbot | New bug: #914920 in ipmitool (universe) "Please merge ipmitool 1.8.11-5 (universe) from Debian testing " [Undecided,Confirmed] https://launchpad.net/bugs/914920 | 18:26 |
| === dendro-afk is now known as dendrobates | ||
| iToast | Im upgrading to ubuntu server 11.10. | 18:41 |
| pmatulis | iToast: good stuff | 18:41 |
| iToast | Should I hold out for 12.0 | 18:43 |
| ikonia | it's up to you | 18:44 |
| ikonia | it's 4 months away | 18:44 |
| sanilm | Hello | 18:49 |
| sanilm | I have PHP5 installed on my Ubuntu Server, and when I try to send an request using cURL to Google Maps Geocode API, the page timeouts | 18:50 |
| sanilm | I suspect that Ubuntu blocks the PHP5 requests | 18:50 |
| sanilm | How can I check / change that? | 18:50 |
| ikonia | why do you think ubuntu blocks php5 ? | 18:51 |
| ikonia | php5 is just a language, what ports is it communicating on | 18:51 |
| ikonia | thats the key if you think it's being blocked | 18:52 |
| sanilm | I suppose 18, because I use apache | 18:52 |
| sanilm | It's an ordinary LAMP server | 18:52 |
| mgw | what is the url for the geocode api? | 18:52 |
| ikonia | 18 ? | 18:53 |
| mgw | can you ping it? | 18:53 |
| ikonia | port 18 ? | 18:53 |
| ikonia | I don't think it will be communicating on port 18 | 18:53 |
| mgw | I think he meant 80 — no? | 18:53 |
| sanilm | Yes 80, sorry | 18:53 |
| mgw | http that is | 18:53 |
| sanilm | Did not notice that I made a mistake | 18:53 |
| ikonia | sanilm: ok, so port 80 is open by default | 18:53 |
| ikonia | ubuntu won't be blocking that | 18:53 |
| sanilm | Well, when I use cURL to send a request, I get a false boolean... I don't know what else could be the reason why it does not work | 18:54 |
| sanilm | If I open that URL from Chrome, it perfectly works | 18:54 |
| sanilm | Does firewall have to do something with that? | 18:56 |
| sanilm | I just remembered what else could be the reason | 19:02 |
| sanilm | I chagned my network settings so that ubuntu uses the static ip | 19:02 |
| sanilm | But when I try to update it or download any package, it won't work until I change that back to dhcp | 19:02 |
| sanilm | Could that be the reason? | 19:02 |
| === david is now known as Guest50124 | ||
| Guest50124 | I want to call screen with a command and I want to command to immediately detach | 19:07 |
| mgw | sanilm: yes, if your network is failing to route or resolve dns, you won't be able to connect to a host | 19:07 |
| Guest50124 | something like 'screen ping google.ca &' | 19:07 |
| mgw | that's why i asked if you could ping the google url | 19:08 |
| sanilm | I figured out the problem | 19:10 |
| uvirtbot | New bug: #914954 in mysql-dfsg-5.1 (main) "package mysql-server-5.1 5.1.41-3ubuntu12.10 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/914954 | 19:11 |
| pdtpatrick_ | Question -- I'm trying to update the motd on my server. /etc/motd.d has the files, i've change of the files in this directory, i understand pam.d/login has option to display the motd upon successful login, however, it does not display the new change | 19:26 |
| pdtpatrick_ | there''s also no update-motd binary to restart the service (if it is) | 19:26 |
| kirkland | pdtpatrick: it's /etc/update-motd.d/ | 19:26 |
| pdtpatrick_ | right i've edited a file in there, how do i make it appear on the next login? | 19:27 |
| pdtpatrick_ | kirkland: btw -- good work on the new byobu - quite nice | 19:27 |
| kirkland | pdtpatrick: thanks | 19:27 |
| kirkland | pdtpatrick: it should automatically do so | 19:27 |
| kirkland | pdtpatrick: if you file a bug and assign it to me, i'll get a /usr/sbin/update-motd shell script that manually updates | 19:28 |
| kirkland | pdtpatrick: in the mean time, i can give you a one-liner to do so, if you like ... | 19:28 |
| pdtpatrick_ | sure | 19:29 |
| pdtpatrick_ | interesting .. creating /etc/motd.tail | 19:31 |
| pdtpatrick_ | will auto append to the file based on the 99-footer shell script | 19:31 |
| uvirtbot | New bug: #914971 in samba (main) "package samba 2:3.5.8~dfsg-1ubuntu2.3 failed to install/upgrade: sub-processo script post-installation instalado retornou estado de saída de erro 1" [Undecided,New] https://launchpad.net/bugs/914971 | 19:36 |
| pdtpatrick_ | kirkland: i figured it out .. they have /var/run/motd.new so I'm just going to symlink that to /var/run/motd and that should work. | 19:39 |
| kirkland | pdtpatrick: cool | 19:40 |
| jandro | does anyone know of a good alternative to torrentflux-b4rt with multi-user support and is web-based? cheers | 19:42 |
| === sixstringsg|away is now known as sixstringsg | ||
| === bladernr_ is now known as bladernr_afk | ||
| === sixstringsg is now known as sixstringsg|away | ||
| peterrus | hey | 20:31 |
| peterrus | my realtek RTL8111/8168B gigabit ethernet card works fine under the ubuntu server bootcd | 20:31 |
| peterrus | (during install) | 20:31 |
| peterrus | and on the ubuntu desktop live cd | 20:31 |
| peterrus | 11.10 that is | 20:32 |
| peterrus | but in my install it does not work | 20:32 |
| peterrus | it does show up in lspci of course | 20:32 |
| peterrus | but the interface is not there in ifconfig | 20:32 |
| peterrus | any ideas? | 20:32 |
| jmarsden|work | peterrus: is /etc/network/interfaces set up and doing the right things for that NIC? | 20:38 |
| peterrus | jmarsden|work: it shouldnt do anything | 20:39 |
| peterrus | except get a static ip | 20:39 |
| peterrus | but dhclient doesnt get a ip either | 20:39 |
| jmarsden|work | Whether it is static or dynamic IP depends on what you write in /etc/network/interfaces :) | 20:40 |
| peterrus | people | 21:31 |
| peterrus | my networkcard is not showing up in ifconfig | 21:31 |
| peterrus | its a | 21:31 |
| peterrus | RTL8111/8168B gigabit ethernet card | 21:31 |
| peterrus | the module is loaded | 21:31 |
| peterrus | where else could the problem lie? | 21:31 |
| peterrus | is there something I can check? | 21:31 |
| peterrus | it is working under both the ubuntu server and desktop 11.10 livecd | 21:32 |
| peterrus | but the local install does not work | 21:32 |
| peterrus | same modules on the livecd loaded | 21:33 |
| peterrus | as on the local install | 21:33 |
| peterrus | namely | 21:34 |
| peterrus | r8169 | 21:34 |
| peterrus | hm | 21:35 |
| peterrus | hmm | 21:35 |
| peterrus | this is strange | 21:35 |
| peterrus | the card shows up as eth1 | 21:35 |
| peterrus | and there is no eth0 | 21:36 |
| peterrus | how would I fix this? | 21:36 |
| JanC | did you use another NIC before? | 21:36 |
| peterrus | yes | 21:36 |
| peterrus | but it is not there currently | 21:36 |
| peterrus | and I dont plan on using it again | 21:36 |
| peterrus | so I guess eth0 is mapped somewhere | 21:36 |
| RoyK | peterrus: rm /etc/udev/rules.d/70-persistent-net.rules | 21:36 |
| RoyK | iirc | 21:36 |
| JanC | something like that indeed | 21:36 |
| RoyK | peterrus: debuntu binds ethX to a mac address | 21:37 |
| peterrus | RoyK: exactly, super! | 21:37 |
| peterrus | debuntu | 21:37 |
| peterrus | what is that? | 21:37 |
| RoyK | debian/ubuntu :P | 21:37 |
| peterrus | haha XD | 21:37 |
| JanC | udev links nicks to device names, because by default they are numbered in the order they come up otherwise | 21:37 |
| peterrus | ow | 21:37 |
| peterrus | I see | 21:38 |
| peterrus | awesome | 21:38 |
| peterrus | thanks alot! | 21:38 |
| RoyK | that file is generated on bootup if it doesn't exists | 21:38 |
| RoyK | s/s$// | 21:38 |
| peterrus | ah | 21:38 |
| peterrus | then I will just delete it | 21:38 |
| RoyK | I had to disable the generation of that once on a netboot system, but don't remember where | 21:39 |
| JanC | peterrus: alternatively, you can edit that file, I guess | 21:39 |
| peterrus | I could do that | 21:39 |
| peterrus | but its to late now XD | 21:39 |
| guntbert | peterrus: deleting it is fine | 21:39 |
| RoyK | you still need to reboot, and then removing it is easier :P | 21:39 |
| JanC | RoyK: well, in some cases it's easier to edit (e.g. if you have 2 nicks and replaced one, but the remaining one needs to keep its device name) | 21:40 |
| RoyK | true... | 21:41 |
| RoyK | but in his case, where a nic has been replaced, I'd just remove it | 21:41 |
| JanC | it would be more useful if NICs were named after the PCI slot they are in, I guess... | 21:42 |
| JanC | s/after/based on/ | 21:43 |
| RoyK | not at all | 21:43 |
| RoyK | that is, they are, from linux' point of view | 21:44 |
| RoyK | udev renames them later | 21:44 |
| RoyK | which is good indeed | 21:44 |
| RoyK | since without that, suddenly eth1 can become eth2 after inserting a new board etc | 21:44 |
| RoyK | which is usually not what you'd want | 21:45 |
| JanC | the idea is that a replaced NIC would get the same device name automatically | 21:45 |
| RoyK | that doesn't work in practice | 21:45 |
| RoyK | since the new nic may be another chipset, meaning the order may be differenc | 21:46 |
| RoyK | different | 21:46 |
| RoyK | so better tie the nic to something that identifies it, meaning its MAC address | 21:46 |
| JanC | RoyK: the "persistent rules" thing is because now they are ordered on the time their driver gets innitialized, what I mean is naming them based on the fysical slot they are in | 21:47 |
| JanC | physical | 21:47 |
| RoyK | that doesn't make sense | 21:47 |
| JanC | it would make a lot of sense on servers, I think, but maybe less so on desktops | 21:47 |
| RoyK | anyway - you can probably script that up as well | 21:48 |
| RoyK | SMOP | 21:48 |
| JanC | probably ☺ | 21:48 |
| peterrus | did you know | 21:57 |
| peterrus | I actually trashed and reinstalled my server | 21:57 |
| peterrus | which was quite painstaking | 21:57 |
| peterrus | because of this XD | 21:57 |
| peterrus | http://www.speld.nl/2012/01/06/natuurorganisaties-akkoord-met-teruggave-groningen-aan-de-natuur/ | 22:03 |
| peterrus | hm | 22:03 |
| peterrus | wrong window | 22:03 |
| === sixstringsg|away is now known as sixstringsg | ||
| uvirtbot | New bug: #915073 in open-vm-tools (multiverse) "open-vm-dkms 2011.07.19-450511-0ubuntu2: open-vm-tools kernel module failed to build (dup-of: 915068)" [Undecided,New] https://launchpad.net/bugs/915073 | 22:36 |
| === sixstringsg is now known as sixstringsg|away | ||
| skone | do most people enable the root account on server or just use sudo? On a server it seems more realistic to use root accounts | 22:55 |
| maxb | define 'enable' | 22:56 |
| skone | turn it on | 22:57 |
| xranby | skone: when the system boot the first program that start the init program are root | 22:57 |
| xranby | skone: all security are invented by avioding having all priveliges | 22:57 |
| xranby | skone: so init makes sure that the user processes gets locked down | 22:58 |
| xranby | and to answer your question | 22:58 |
| xranby | andministrators use sudo | 22:58 |
| skone | but I have ssh limited to this account and if someone get my password they have root access. If I have a seperate root account and disable ssh on the root account then they would need two passwords to get in | 22:58 |
| skone | that was my main concern | 22:58 |
| pangolin | by enabling a root account and setting a root password you give an attacker 2 possible ways in | 22:59 |
| pangolin | instead of one | 22:59 |
| skone | i guess I can see it both ways | 22:59 |
| pangolin | if you need extended elevated privileges use sudo -i but remember to "exit" when you are done. | 23:01 |
| xranby | skone: how about you add a second user with sudo access | 23:01 |
| xranby | skone: and prevent ssh longin for this second user | 23:01 |
| skone | yeah, thats what I was thinking I wold do | 23:02 |
| kaushal | Hi | 23:24 |
| kaushal | Can i seek help on rsyslog application on Ubuntu 10.04 LTS ? | 23:25 |
| uvirtbot | New bug: #915089 in tomcat6 (main) "Tomcat 6 start up fails with unknown dir /var/lib/tomcat6/work" [Undecided,New] https://launchpad.net/bugs/915089 | 23:46 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!