[00:28] <armaan> how can i kill nohup jobs
[00:47] <patdk-lap> same way you kill any other job
[00:51] <armaan> patdk-lap: is there any way to do it automatically. i need to do it for testing purpose
[00:52] <patdk-lap> there is nothing different about nohup, and anything else
[00:52] <patdk-lap> so whatever you *normally* use, will work
[03:02] <osmosis> how do I get more info on a package from apt-cache before I install it?
[03:17] <lifeless> statik: o/
[03:36] <JonEdney> Hey, any way to remove Java from 12.04 server? "Virtual packages like 'sun-java6-jre' can't be removed"
[03:37] <qman__> dpkg -l | grep jre
[03:37] <qman__> dpkg -l | grep java
[03:37] <qman__> then apt-get remove those
[03:38] <JonEdney> Ah, wasn't aware that's how it worked, interesting.
[07:12] <thisismyname> hi guys, i got the following problem: my postgres database server is only using one cpu, even with multiple connections
[07:14] <thisismyname> using ubuntu 12.04 vanilla kernel
[07:33] <overrider> How can i go about monitoring about 50 ppp interfaces created by pptp users that dial in; id like to perform some traffic accounting. Is there any sensible way to combine those interfaces into 1 single interface for me to do my bandwith / netflow monitoring on?
[07:48] <greppy> overrider: MRTG can do aggregate data
[07:48] <greppy> you can tell it to graph the data from all the interfaces.
[07:51] <_ruben> overrider: could do with a netfilter based accounting setup hooked into the filter/FORWARD chain for instance, it could make use of ppp+ as in/out interface. etc
[09:07] <overrider> ill look at netfilter and mrtg - trouble with mrtg is that the interfaces come and go as they please, based on user login or logout, not sure how that gets handled. Thanks though
[10:12] <BuenGenio> hello
[10:12] <BuenGenio> how do I get openvpn client to connect without asking for username & password?
[11:26] <RoyK> BuenGenio: isn't that a bit like installing a new and fancy lock on your house and then glueing the key into the lock?
[11:27] <greppy> RoyK: depends on your setup.
[11:28] <RoyK> AFAGCT, there are separate builds for that http://www.acevpn.com/2009/12/07/openvpn-build-with-save-password-enabled/
[12:05] <zul> good morning
[12:42] <lynxman> zul: morning!
[12:42] <zul> lynxman: slacker
[12:53] <lynxman> zul: :D
[13:58] <roaksoax> lamont: howdy! I was wondering if bind9 supports conf.d by default, or if it would be a nice thing to have it do
[14:01] <roaksoax> lamont: the issue is basically this: 1. MAAS installs and creates dir /etc/bind/maas. 2. Adds an include line to 'named.conf.local'.
[14:02] <roaksoax> lamont: so concern1: then if I remove MAAS (not purge), files will be left in /etc/bind/maas. Then if I purge bind9, it wont remove /etc/bind because  /etc/bind/maas exists
[14:03] <roaksoax> lamont: concern2: If I upgrade bind9, I'm presumming dpkg will detect that named.conf.local has differed, and will offer to overwrite it. If that happen then MAAS config will not be accessible
[14:41] <ironm> hello. in case anyone (! erkules)  is interested: http://rsync.it-infrastrukturen.org/public-mariadb/ubuntu/.tmp/README.mysqld_multi_install_concepts.txt / http://rsync.it-infrastrukturen.org/public-mariadb/ubuntu/mysqld_multi_install_db_voip_files.tgz
[14:41] <ironm> thank you in advance for *any" feedback
[15:05] <lamont> roaksoax: 1) dunno.  2) I'm very reluctant to deviate from upsteam wrt configuration file directivces.
[15:07] <roaksoax> lamont: i see, do you have any recommendation on how to handle the case above though?
[15:09] <lamont> bind has a history of not actually purging itself all the way, largely because of the package name change between bind and bind9
[15:10] <lamont> which I expect to recreate for bind10 :(
[15:10] <lamont> what is maas doing that it wants to add files to the bind config (as opposed to replacing them wholesale)?
[15:11] <roaksoax> lamont: yes, so basically create /etc/bind/maas and place everything under it
[15:12] <roaksoax> lamont: and add an include on /etc/bind/named.conf.local
[15:12] <roaksoax> that includes above directory
[15:12] <lamont> yes, but WHAT is it doing that it needs to add files?
[15:13] <lamont> gratuitously adding a subdirectory for something that you're configuring completely strikes me as just silly
[15:13] <roaksoax> lamont: it will be used as the DNS server for the MAAS client machines
[15:13] <roaksoax> rvba: ^^
[15:14] <roaksoax> lamont: so it is the MAAS DNS server instead of having an external DNS
[15:14] <lamont> so is this the maas server install, and wanting to interoperate with the admin's previously-configured bind9 install?
[15:14] <lamont> s/interoperate/not trash/
[15:14] <roaksoax> lamont: yes
[15:14] <lamont> so s/instead of/in addition to/
[15:15] <roaksoax> lamont: yes, so in cases where in the local network there's no DNS (which is needed by juju), then MAAS would be the DNS
[15:15] <roaksoax> lamont: so it is just part of the replacement of cobbler handling DNS/DHCP
[15:17] <roaksoax> lamont: so, that being said, do you think it would be better approach to simply replace named.conf instead of includding config files?
[15:25] <lamont> I'm going to guess that your admin will want to be involved in a number of decisions about what address blocks and what bogus-TLD/real-suffix is being used for maas, and even after ripping maas out of the server... how much automation are you expecting to do?
[15:28] <roaksoax> lamont: well obviously ripping it out would mean leaving this usable
[15:28] <roaksoax> rvba: ^^
[15:28] <rvba> otp
[15:37] <lamont> roaksoax: replacing the files in /etc/bind with the maas stuff may be the simplest.  There are policy issues with package B modifying package A's conffiles directly, which you're stuck facing no matter how you go about it. (the policy compliant way to solve it is to provide me some code for the bind9 package to add/remove zones from the config, which maas-server would then call to make the mods
[15:40] <roaksoax> lamont: i see. Thanks for the input. I'll run this with the MAAS team.
[15:42] <lamont> roaksoax: there is some stuff inthe postfix stack that does just that, fwiw
[15:42] <lamont> (modifying postfix config, totally separate from anything to do with bind9)
[15:43] <roaksoax> lamont: ok cool, I'll look into that. Thanks again :)
[16:52] <Daviey> Anyone have the exact size of a default ubuntu server 12.04 to hand?
[16:59] <patdk-wk> Daviey, I have 963megs
[17:03] <Daviey> patdk-wk: vanilla cd install?
[17:09] <patdk-wk> daviey, yep, 64bit, using last alpha
[17:10] <Almindor> hgello
[17:10] <Daviey> patdk-wk: thanks
[17:10] <Almindor> is there a known bug with IO operations/copying on 10.04 LTS? (probably kernel bug with scheduler/IO)
[17:11] <Almindor> I was copying ~300gb (normal cp command) and then I tried to install one small python module from apt-get and it stalled on "unpacking" until the CP was done
[17:11] <Almindor> node that the copy was disk to different disk and the python module installs to a 3rd disk
[17:11] <patdk-wk> your using lvm/dmraid/...?
[17:12] <Almindor> I'm not using raid on any of the 2 disks being copied from/to
[17:13] <patdk-wk> are any of the *devices* under /dev/mapper?
[17:13] <Almindor> patdk-wk: the main disk (the one python module was installing to) is
[17:14] <patdk-wk> not sure then
[17:14] <patdk-wk> I know when writing to one, it will completely hog the system, haven't tried it on newer kernels
[17:15] <Almindor> hmm
[17:15] <Almindor> so there's a mapper bug in this kernel?
[17:15] <patdk-wk> no one said it was a bug
[17:16] <Almindor> well the "main" drive is actually two 500gb ones mapped (from what I can say, I didn't set this up)
[17:16] <Almindor> but the secondary drives are all direct mounts via fstab
[17:16] <patdk-wk> works, and works optimally, and works without affecting others, so many different levels, that don't even account for bugs
[17:16] <Almindor> sorry but a cp command blocking other io operations is a major kernel bug
[17:17] <patdk-wk> if that is the case
[17:17] <Almindor> it's like back to DOS :)
[17:17] <patdk-wk> atlesat in my cases, it didn't block, but it did bring it to a crawl
[17:18] <Almindor> well I was installing python-argparse (few kb size) and it stopped on unpacking until the whole 300gb copy was done (about an hour IIRC)
[17:18] <Almindor> so I think that's less than a crawl
[17:18] <patdk-wk> an hour for 300gigs?
[17:18] <Almindor> it was started in the middle
[17:18] <patdk-wk> still, horrible slow performance there
[17:19] <patdk-wk> ah, from single disks
[17:19] <patdk-wk> I'm too used to using 20+ disk arrays
[17:20] <Almindor> :)
[17:20] <Almindor> this one only has 4 :D
[17:20] <MoleMan> where is the system bash config stored?
[17:21] <patdk-wk> bash has a config?
[17:21] <patdk-wk> you know the system doesn't *normally* use bash right?
[17:21]  * MoleMan starts thinking
[17:22] <patdk-wk> it uses dash, unless you explicitally tell it to use bash
[17:23] <patdk-wk> if there was a bash config, it would be the bash startup file, /etc/bashrc
[17:23] <MoleMan> all the users are set to use /bin/bash    in /etc/passwd and I've not set that... unless I'm getting confused somewhere?
[17:23] <patdk-wk> oh, /etc/bash.bashrc
[17:24] <patdk-wk> (was actually on a redhat system, one of the 3 I have to deal with)
[17:24] <MoleMan> thanks :)
[17:26] <LoT> SpamapS: question for you, if you're alive.
[17:26] <MoleMan> patdk-wk: do you know where the user specific config is as well pleasE?
[17:27] <patdk-wk> in the users home folder
[17:27] <SpamapS> LoT: whats up?
[17:28] <LoT> just a small question... not sponsorship related but directly related to this bug: https://bugs.launchpad.net/ubuntu/precise/+source/php5/+bug/1014044
[17:28] <LoT> shouldnt that be a crit one?  i was reading through the bugs triage guide, it said regressions should be crit
[17:28] <LoT> unless i'm misreading it :/
[17:29] <MoleMan> patdk-wk: strange, I can't find it but am sure I have used it before :/
[17:29] <patdk-wk> it should make it in your folder
[17:29] <LoT> wait, might've had an older version up :P
[17:30] <patdk-wk> .bashrc
[17:30] <MoleMan> only thing there is .bash_history :/
[17:31] <MoleMan> uuunless... maybe it was my netbook I've changed it on before not my server :/
[17:31] <patdk-wk> I haven't changed it, it was *installed* there by default for me
[17:31] <patdk-wk> maybe when yo uadded users you didn't tell it to use the template
[17:32] <LoT> SpamapS: nevermind, my browser had an (old) triage guide cached
[17:32] <LoT> so apparently the thing was showing me ancient data
[17:32]  * LoT cleared the cache and saw that there was no such requirement for regressions
[17:32] <LoT> although this one probably needs regression tagging
[17:33] <SpamapS> LoT: huh?
[17:33] <LoT> SpamapS: nevermind.
[17:34] <LoT> SpamapS: although i will poke you about the sponsoring of that debdiff, there's 6 groups using webhosting and the error logs arent useful because of php not reporing.
[17:34]  * LoT has his own motivations for poking the debdiff on LP Bug 1014044 into precise
[17:35] <SpamapS> LoT: patience. There are many sponsors, and only one of me. ;)
[17:35] <LoT> :P
[17:36] <LoT> patience is hard when you've got ISOs for each group contacting you daily saying "What's up with it not reporting"
[17:36] <LoT> they keep thinking its hacks, i say "I don't know, there's a PHP regression"
[17:37] <SpamapS> LoT: indeed, the fix will get out to them soon. :)
[17:37] <LoT> that's what I said
[17:37] <LoT> guess what, they dont listen :/
[17:37]  * LoT is not pleased
[17:37] <LoT> perhaps i should email the sponsors mailing list, poking them saying "This needs processing"?
[17:37] <LoT> (probably not, but...)
[17:38]  * LoT does have horrible impatience when being poked via email daily :/
[17:41] <LoT> actually, wait, i have a PPA i can use for this until it shows up in -updates xD
[18:13] <zul> adam_g: so we need to transition glance-client to python-glanceclient in the debian packaging fyi
[18:13] <adam_g> zul: yes
[18:13] <adam_g> zul: the thing is, python-glanceclient completely broke the user interface from the original glance client
[18:13] <zul> yeah
[18:14] <adam_g> so doing so is going to break anyones scripts/configuration managemenet/charms/etc
[18:14] <adam_g> i have a half-done patch i want to get upstream that provides a compatability layer
[18:14] <zul> cool beans
[18:15] <zul> glanceclient is in main by the way so i just did the recommndes: from glance-client to python-glanceclient, when you compat layer gets in we will do a dummy transitional package in python-glanceclient
[18:16] <adam_g> zul: what about dropping glance-client entirely?
[18:16] <adam_g> zul: oh
[18:16] <adam_g> i think the original client will disappear from the glance source at some point in the future
[18:16] <zul> adam_g: glance-client will be dropped in gaglance
[18:16] <zul> adam_g: right
[18:17] <adam_g> zul: where are you committing this stufF? i have a bug number that can be added
[18:18] <zul> adam_g: openstack-ubuntu-testing branches
[18:18] <adam_g> zul: lemme know when you've pushed it there
[18:18] <zul> adam_g: done
[18:28] <MoleMan> I'm probably being stupid here, but I'm trying to create a symlink and it won't work, the command I'm using is 'ln -s Dropbox/schematics MC-Chocobo/plugins/WorldEdit/schematics'
[18:30] <MoleMan> well it appears to be trying to make it but not doing what I expect, if I try to list the contents of the symlink it says that it is not a directory, how do I create a symlink to a folder?@
[18:46] <MoleMan> what do I need to do to get a symlink to a folder to work properly? it is currently refusing to acknowlege it is a folder and list from it...  the command I'm using is 'ln -s Dropbox/schematics/ MC-Chocobo/plugins/WorldEdit/schematics'
[18:46] <MoleMan> oops
[19:12] <r3dLunchb0x> anyone successfully setup up email alerts with nagios on ubuntu server 11.04? I have an issue where the email gets sent BUT it is using the wrong host name. I have checked all config files for nagios that even mention the host name and still it sends as wrong hostname.
[19:19] <SpamapS> Hm, Ubuntu cloud mirror in us-west-1 going sllllooww..
[19:19] <SpamapS> 87% [28 Packages bzip2 0 B] [34 Translation-en 372 kB/3341 kB 11%]                                                   3823 B/s 12min 56s
[19:20] <Daviey> SpamapS: nothin' we can do about that.. i think you want to speak to IS :)
[19:20] <SpamapS> isn't it S3?
[19:27] <Daviey> SpamapS: haven't we handed it over?
[19:28] <Daviey> arosales ^ ?
[19:28] <SpamapS> Daviey: yes, I'm just surprised S3 would go so slowly
[19:28] <SpamapS> of course, now its going afaster
[19:28] <SpamapS> 16% [65 wget 0 B/277 kB 0%]                                                                                          2352 kB/s 1min 16s
[19:34] <smoser> SpamapS,
[19:34] <smoser> http://pad.lv/966577
[19:35] <smoser> that is milestoned for 12.04.1 (by you on july 12). is that going to get there by next week?
[19:35] <SpamapS> smoser: "maybe" :-P
[19:36] <smoser> ok. then i leave that to you, and will forward on a very strong maybe
[19:52] <smoser> stgraber, ping
[19:52] <arosales> Daviey: sorry, afk for some lunch. But, yes as SpamapS said, IS is now handling the S3 mirrors.
[19:54] <SpamapS> thanks guys, I was just making a comment
[19:54] <SpamapS> not a support request :)
[19:54] <smoser> why is on 850443 http://status.qa.ubuntu.com/reports/kernel-bugs/reports/rls-p-tracking-bugs.html . is it solely the rls-p-tracking? is it appropriate for me to just take that off ?
[19:54]  * ScottK demands support for SpamapS .
[19:55] <bladernr_> hey, just out of curiosity, on a default Ubuntu Server install, is there a dbus method for getting info on network devices ( like driver and such)?  We can do it when NetworkManager exists, but that's not installed on server.
[19:55] <stgraber> smoser: pong
[19:55] <smoser> ^
[19:57] <SpamapS> bladernr_: not that I know of.
[19:57] <bladernr_> SpamapS: that's what I thought, just wanted to verify with someone... not a big deal, just an idea I had
[19:57] <bladernr_> thanks!
[20:06] <smoser> SpamapS, are you oging to do 945176 by next thursday?
[20:07] <SpamapS> smoser: that, and the other one, are on my list to try and knock out by Monday. So, maybe again. :)
[20:07] <smoser> SpamapS, deal.
[20:07] <smoser> i promise no more nagging for at least 24 hours.
[20:09]  * SpamapS tags smoser with 'geek-stop-nagging'
[20:10] <Daviey> SpamapS: arosales has us mixed up :)
[20:18] <arosales> Daviey: ah, SpamapS had the initial comment
[20:18]  * arosales should read the back scroll a little closer
[20:43] <Combatjuan> Can anyone suggest a way to determine the age of a particular network connection?  (in particular this connection goes through stunnel).  Bonus points if you can say when was the last time it was used.
[20:54] <dannf> Combatjuan: ps -eo etime $(pidof stunnel) ?
[20:55] <Combatjuan> dannf: Well, that did something... and it involved times.  I'll have to read a bit before I understand it.  Thanks.
[20:56] <dannf> 'ps -eo cmd,etime | grep stunnel' maybe
[20:57] <Combatjuan> dannf: It shows me the times, but not of what.  I guess I need another column containing the FD?
[20:58] <dannf> its the elapsed time the process has been running
[20:59] <dannf> assuming the stunnel process is running for one connection only (i don't know if multiplexing is possible - if so, this doesn't help)
[20:59] <Combatjuan> dannf: The stunnel process hold 500 connections.  Well, almost.
[21:00] <Combatjuan> I was hoping I could look at the mtimes on their file descriptors or something, or maybe use netstat to tell me age...  Presumably there is a way.
[21:00] <Combatjuan> Anyway, I'm looking for the age of the individual socket connections that stunnel is holding.  I need to cause some of them to idleout early.
[21:07] <phillw> Hi guys, re: http://ubuntuforums.org/showthread.php?p=12146694#post12146694 should I just reboot the crippled VM so as to get it resonable service for https://wiki.ubuntu.com/Thomas%20Ward ?
[21:07] <phillw> you guys know servers, this is an unhappy one
[21:08] <phillw> oh, sod it... if it is dying continously I'll reboot it.
[21:08] <phillw> znc @ trekweb is going down for a reboot.
[21:19] <phillw> the VNC server has its owner now back in charge.
[22:34] <AlphaWolf> I am running Plex Media Server, but I can't get it to read one of my folders, and I think it's a permissions issue. The information from "ls -l" is as follows: Folder it can't read "drwxrwxr-- 1 joseph plugdev 24576 Aug  2 23:26 GeneralMedia", folder it can read "drwxrwxr-x 1 joseph      99    14 Aug  2 23:29 TimeMachine". I'm guessing it's the "x" at the end (which is other execute? so anyone can execute?), but running "
[22:34] <AlphaWolf> chmod o+x GeneralMedia" does not change anything. One other note: GeneralMedia is an NTFS partition, whilst TimeMachine is a HFS+ partition.
[23:14] <Takyoji> Is it sane to upgrade from 10.04 to 12.04, rather than waiting a few days for 12.04.1?
[23:19] <smw_work> Takyoji, by "upgrade" you mean clean install?
[23:19] <Takyoji> do-release-upgrade, specifically.
[23:19] <smw_work> Takyoji, not sure I would do that on a server...
[23:19] <Takyoji> Webserver
[23:20] <smw_work> I would not trust it...
[23:20] <smw_work> but whatever. I see no difference in waiting vs not waiting
[23:21] <smw_work> 12.04.1 is just a snapshot in time
[23:21] <smw_work> Takyoji, so I do not see how waiting a few days would change anything.
[23:21] <Takyoji> it's just that it was stated in the documentation "It is generally recommended that users of Ubuntu 10.04 LTS wait until the first point release, due in July, before upgrading."
[23:28] <Takyoji> Well, with Linode, I suppose I'll start a new instance, copy things over, then swap IPs over