[00:24] <AndyGraybeal> i've read that ubuntu doesn't support backing up ACL's, is this true for Ubuntu 10.04 ?
[00:27] <AndyGraybeal> nevermind, i think this is the answer: http://ubuntuforums.org/showthread.php?t=1451667
[01:03] <clusty> hey
[01:03] <clusty> how can i install an rpm in ubuntu ?
[01:05] <pmatulis> clusty: alien i think
[01:05] <pmatulis> !info alien
[01:05] <Pici> !alien
[01:05] <clusty> Pici, how dangerous ?
[01:06] <clusty> damn intel does not provide a tool to flash the raid controller for ubuntu
[01:14] <mase_wk> clusty: which raid controller is it ?
[01:15] <mase_wk> is it 'real' raid or fake raid
[01:15] <clusty> mase_wk, real raid
[01:16] <mase_wk> *most* real RAID cards have full applications available for linux usually from the controller vendors site
[01:16] <clusty> SRCSAS18E
[01:16] <clusty> this has for suse and rh
[01:16] <mase_wk> is that the kernel module ?
[01:16] <clusty> and alien chokes
[01:16] <clusty> mase_wk, no
[01:16] <clusty> there is open source mod
[01:16] <clusty> i wanted to flash to new FW
[01:17] <mase_wk> ah so you have a utility just not specific to ubuntu ?
[01:17] <clusty> cause the on the card is prolly 3 years old
[01:17] <clusty> yes
[01:17] <mase_wk> where does alien choke ?
[01:17] <clusty> and i cannot alien cause it complains about 32/64 bit issue
[01:17] <mase_wk> is it a 32b binary ?
[01:17] <mase_wk> and you have a 64bit os ?
[01:18] <clusty> dpkg-gencontrol: error: current host architecture 'amd64' does not appear in package's architecture list (i386)
[01:18] <clusty> yes
[01:18] <mase_wk> ah ok.
[01:18] <clusty> seems is 32 bit
[01:18] <clusty> the name of it does not say shit :D
[01:18] <clusty> i will try with DOS
[01:18] <clusty> never thought i woul;d fokin ever boot dos
[01:18] <mase_wk> if your having to reboot anyway you could just boot off a livecd
[01:18] <mase_wk> and install it into that environment
[01:19] <clusty> i got a DOS bootable disk already
[01:19] <mase_wk> although that's still annoying as you don't have the utilities when the server is up and running
[01:19] <clusty> cause i want to fix the TLER issue
[01:19] <mase_wk> i am surprised they don't make a 64bit version
[01:19] <clusty> mase_wk, got any experience with cheapo disks and raids ?
[01:20] <mase_wk> sure, i got a bunch of them at work :)
[01:20] <clusty> any TLER issues ?
[01:20] <mase_wk> not that i've encountered but i have different RAID cards to you
[01:20] <clusty> time limited error recovery as of
[01:21] <clusty> well this card is also oldish. 3 years old
[01:21] <clusty> but FW update is from 2010
[01:21] <mase_wk> is it on a particular port ?
[01:22] <clusty> ??
[01:22] <clusty> PCI-E ?
[01:22] <mase_wk> no i mean , raid port
[01:22] <clusty> nope
[01:22] <mase_wk> so whats the actual error message you get ?
[01:22] <clusty> alien ?
[01:22] <mase_wk> no the TLER issue
[01:23] <clusty> i did not. just got the card like 1h ago
[01:23] <clusty> just finished installing it
[01:23] <mase_wk> i'm confused
[01:23] <clusty> i got no TLER problems
[01:23] <clusty> but i read about them and WD hdds
[01:23] <mase_wk> ah ok
[01:23] <mase_wk> now i understand
[01:23] <clusty> i was asking you how common are they
[01:23] <clusty> :D
[01:24] <mase_wk> erm well we have cheap disks but i don't believe that we have any WD specifically in those boxes
[01:24] <clusty> i find it hard to believe the hdd with chew for 7 sec trying to recover an error such that the controller would drop the disk
[01:24] <mase_wk> it's not that hard to believe.
[01:24] <clusty> WD charges 100bux extra for disks with TLER turned on
[01:25] <clusty> but they also give you the util to turn it on yourself
[01:25] <mase_wk> some raid cards are fussy and disks are generically crap
[01:25] <mase_wk> and getting worse
[01:25] <mase_wk> i've gone through so many hard disks in different boxes this year
[01:26] <mase_wk> it's almost worth buying ssd's so i don't have to visit the datacenter every other month
[01:26] <clusty> ony disk i had choking was a 2.5" one in an external box
[01:26] <clusty> which i kinda abused
[01:27] <clusty> mase_wk, what sort of work do you do?
[01:28] <mase_wk> when i'm not replacing disks i'm probably what is best described as a devop
[01:28] <mase_wk> ie everything :)
[01:28] <mase_wk> i work for a small company
[01:28] <clusty> so i presume DB heavy apps ?
[01:29] <mase_wk> nah not so much
[01:29] <clusty> i also did that for 1,h years
[01:29] <clusty> 1.5
[01:29] <clusty> maintaining a small internal server
[01:29] <clusty> and 2 DB servers
[01:30] <mase_wk> i work for a solutions provider which includes a small boutique ISP
[01:30] <mase_wk> so basically it's writing applications / support tools around that
[01:31] <mase_wk> right now i'm applying language translations to a clients website
[01:31] <mase_wk> yay \0/
[01:31]  * mase_wk sighs
[01:31] <clusty> the nastiest thing i did was to configure the a router gizmo and LDAP users
[01:31] <clusty> damn that was a pain in the ass
[01:31] <ball> Why doesn't http://www.ubuntu.com/getubuntu/downloadmirrors show 10.4 ?
[01:31] <mase_wk> i like the idea of ldap, i just hate LDAP it's self
[01:31] <ball> (isn't 10.4 the most recent release of Ubuntu?
[01:31] <ball> )
[01:32] <clusty> well the LDAP lingo is damn weird
[01:32] <clusty> and all tutorials are either aimed at academics coming up with query tools or experts
[01:32] <clusty> anyways back in 5min
[01:32] <mase_wk> ball: yeh 10.04 is the lates ...who knows why that page lists that
[01:33] <ball> Does anyone happen to have a .torrent file for ubuntu-10.4-server-i386 then?
[01:34] <mase_wk> http://www.ubuntu.com/desktop/get-ubuntu/alternative-download#bt
[01:34] <mase_wk> that lists ubuntu-10.04-alternate-i386.iso.torrent
[01:35] <ball> Thanks mase_wk
[01:36] <mase_wk> np
[01:40] <clusty> mase_wk, can i just dd an image intended for a CD onto a disk?
[01:40] <clusty> and still hope to have it bootable?
[01:42] <ball> clusty: Doesn't sound likely.
[01:46] <mase_wk> clusty: no
[01:47] <mase_wk> you need a partition table and a bootloader
[01:47] <ball> This is going to take a wee while to download
[01:47] <ball> ...but it'll be worth the wait, I'm sure.
[01:48] <mase_wk> ball: does it need to be a torrent ?
[01:48] <mase_wk> it's probably faster from a local mirror
[01:48] <ball> mase_wk: Might as well be consistent.  That's how I download (and share) all my .iso images.
[01:48] <mase_wk> each to their own
[01:49] <mase_wk> most people have 10.04 already so a torrent is not going to be so speedy
[01:49] <mase_wk> but if your prepared to wait i guess it doesn't matter
[01:50] <ball> mase_wk: That's okay.  This is just for a fun project, so I have no deadline.
[03:16] <Patrickdk> ok, I'm having issues with pacemaker/heartbeat
[03:17] <Patrickdk> I am using IPv6addr to failover an ipv6 address
[03:17] <Patrickdk> but I can't figure out how to tell pacemaker to bring that ip up as deprecated, so it isn't used by default, messing ip things
[04:27] <AndyGraybeal> can someone explain the default:mask with ACLs?
[05:34] <chrislabeard> Hey guys is there anyway to improve the speed of webdav?
[05:36] <AndyGraybeal> i'm in the 'tech' group, the 'tech' group has group write access on the Tech folder.  it appears that I can't write to that folder unless the 'tech' group is default; is this normal?   if the 'tech' group is a normal group, i can't write to this folder.  it's very strange.  i need some guidance.
[05:39] <AndyGraybeal> nevermind, i think it is another issue.
[06:12] <mase_wk> chrislabeard:webdav just uses http so it should be pretty fast
[06:12] <chrislabeard> mase_wk: alright
[06:13] <mase_wk> at least as fast as http allows
[06:13] <chrislabeard> mase_wk: 700MB in about an hour
[06:13] <mase_wk> is this a local connection ?
[06:13] <chrislabeard> no
[06:13] <mase_wk> how fast is it via sftp ?
[06:13] <chrislabeard> mase_wk: not sure I'm using a mac with connect to server option
[06:14] <mase_wk> ok what is your basis for comparison then ? ie how do you know it's webdav that is slow and not just your uplink?
[06:14] <chrislabeard> Well I wasn't sure if that was normal speed or not
[06:14] <chrislabeard> I was just seeing if it was possible to bump the speed
[06:16] <mase_wk> without knowing what your uplink speed is it's very difficult to tell if your hitting a performance problem with webdav or a problem with your connection
[06:16] <mase_wk> you need some sort of reference
[06:17] <mase_wk> so you know what your theoretical maximum is and how that relates
[06:17] <chrislabeard> that would be my max upload speed right
[06:17] <mase_wk> yes
[06:18] <chrislabeard> its about 1.8-2Mb
[06:18] <mase_wk> what kind of connection are you on ? DSL / Cable ?
[06:18] <chrislabeard> DSL
[06:18] <mase_wk> annex-m ?
[06:18] <chrislabeard> Its uverse
[06:19] <chrislabeard> I'm not sure if its annex-m
[06:19] <mase_wk> just wondering how your getting 1.8 up on a dsl connection is all
[06:20] <mase_wk> iirc the max upload is about <1Mb without annex-m
[06:20] <chrislabeard> We have the 18MB/D and 2MB/Up
[06:20] <mase_wk> k.
[06:21] <mase_wk> ok does your ISP give you any hosting space ?
[06:21] <chrislabeard> not that I know
[06:21] <chrislabeard> of
[06:21] <mase_wk> do you have somewhere local you can use as a speed test ?
[06:21] <mase_wk> so you can work out roughly how much you can actually upload in a sustained fashion
[06:22] <mase_wk> my guess is that the webdav server is not your problem
[06:22] <chrislabeard> Yeah I'm on the same network as my server
[06:22] <chrislabeard> If I download locally it get about 2.xMB a second
[06:22] <mase_wk> forget download
[06:22] <mase_wk> upload is different.
[06:23] <chrislabeard> So upload a file locally
[06:23] <chrislabeard> I uploaded a 700MB file it took about 2 Mins but I couldn't see the transfer rate
[06:23] <mase_wk> what do you mean by your on the same network, i thought you said this was a remote upload ?
[06:24] <chrislabeard> I am on the same network however my friends when they access it
[06:24] <mase_wk> ah ok so you can upload a file in 2 mins but it takes your friends ( who are remote ) a long time ?
[06:24] <chrislabeard> Right
[06:25] <mase_wk> not really sure how to respond to that.
[06:25] <chrislabeard> ... I know its going to take them a lot longer than me since they are remote
[06:26] <mase_wk> why would you think that is a webdav issue though ?
[06:26] <chrislabeard> I just wanted to know if its possible to bump the speed but from what you are telling me. Its not possible
[06:26] <mase_wk> well they could compress the content but you can't upload faster than your physical limit
[06:27] <mase_wk> regardless of protocol
[06:27] <chrislabeard> K well thats good to know
[06:27] <mase_wk> well...isn't that obvious ?
[06:27] <mase_wk> you can only jam so much down a pipe before you can't fit any more
[06:28] <chrislabeard> Right, Well it didn't seem like they were getting max download
[06:28] <mase_wk> well if its from your server and your server is attached to your DSL
[06:29] <mase_wk> then they can only download at the rate your server can upload at
[06:29] <mase_wk> which is limited by your DSL connection
[06:29] <chrislabeard> right
[06:29] <chrislabeard> So they should get something like a meg a second
[06:29] <chrislabeard> rougly
[06:29] <chrislabeard> roughly
[06:31] <chrislabeard> They are getting like 60-70KB/s
[06:31] <chrislabeard> That just didn't sound right to me
[06:32] <mase_wk> i can pretty much assure you it's not related to your webdav server
[06:32] <mase_wk> unless you have some sort of firewall active doing rate limiting
[06:32] <chrislabeard> K so pretty much wysiwyg
[06:33] <mase_wk> but  i assume you would know that as you would have to set it up
[06:33] <mase_wk> if anything it would be a network issue.
[06:33] <mase_wk> you need to diagnose that if you think there is an issue
[06:33] <mase_wk> you have already proved your server is functioning correctly
[06:33] <chrislabeard> right
[06:43] <robertpayne> anyone here use nginx frontend + apache2 backend ( dynamic content )
[06:46] <SpamapS> robertpayne: I have in the past...
[06:46] <SpamapS> robertpayne: do you have a specific question about it?
[06:46] <robertpayne> Did you use https with it?
[06:47] <robertpayne> Just wondering if I need to run both apache and nginx under the same ssl certificate for the proxy_pass.. or just nginx? apache2 is only running on 127.0.0.1:8080
[06:48] <SpamapS> no
[06:48] <SpamapS> You should not need to use the same cert
[06:48] <SpamapS> the SSL<->user conn will be handled entirely by nginx's ssl
[06:49] <SpamapS> In fact you won't even need SSL for the local apache2
[06:49] <robertpayne> thats what I thought
[06:49] <robertpayne> only nginx needs to be running the SSL
[06:50] <robertpayne> since apache isn't exposed it wouldn't be a security thing
[06:50] <robertpayne> I just wanted to make sure :)
[07:42] <ball> What would I type at the command line to see if I had any supported temperature sensors?
[07:49] <ball> Goodnight everyone
[07:49] <xampart> morning
[07:50] <WalterN> very early moring
[07:50] <WalterN> morning
[07:57] <binBASH> moin
[08:07] <xampart> does "mdadm --manage /dev/md0 --add /dev/sdb2" add sdb2 as a spare if i already have 2 active synced devices in my raid1?
[08:13] <xampart> nevermind
[08:22] <Ongavezyr> Hosanna bruderz
[08:51] <WalterN> got my webserver up now, woo
[08:52] <kaushal> hi
[08:52] <kaushal> Can Ubuntu Hardy Heron be hardened in case of vulnerabilities
[08:52] <kaushal> ?
[08:53] <kaushal> I mean Ubuntu 8.04 Server
[08:53] <joschi> kaushal: ubuntu 8.04 can be hardened like any other linux distribution...
[08:54] <kaushal> joschi: are there Howtos to do it ?
[08:59] <joschi> kaushal: you can mostly follow http://www.debian.org/doc/manuals/securing-debian-howto/
[09:03] <kaushal> joschi: but it is debian specific
[09:03] <kaushal> I know Ubuntu is debian sid based
[09:03] <kaushal> is there ubuntu specific ?
[09:04] <joschi> kaushal: nothing really useful I'm aware of (links appreciated ;) )
[09:21] <Ongavezyr> Viva Mr.Yasser Arafat the Great Leader !
[09:26] <Jeeves_> Isn't he dead?
[09:27] <ikonia> Jeeves_: please dont feed the troll, freenode are aware
[09:27] <Jeeves_> ikonia: I like feeding the trolls :)
[09:27] <ikonia> Jeeves_: don't do it
[09:27] <Jeeves_> I like Shrek too! ;)
[09:34] <dominicdinada> quick question guys
[09:35] <huats> morning
[09:35] <xampart> g'day
[09:36] <ikonia> go for it
[09:36] <Ongavezyr> ikonia bruder:S
[09:36] <Ongavezyr> hello Matthew brotha
[09:36] <ikonia> !ops
[09:37] <xampart> is it possible to have onboard video + seperate video card used at the same time with kubuntu?
[09:37] <Ongavezyr> Hush now baby!
[09:37] <ikonia> xampart: that depends on your hardware, try asking in #kubuntu as this is for ubuntu server
[09:37] <xampart> of course.
[09:38] <Ongavezyr> ikonia brotha you are a stooge?
[09:38] <Ongavezyr> that is dirty work!
[09:39] <Ongavezyr> so... my ssh server why do not working correctly?
[09:41] <joschi> Ongavezyr: you mind giving *any* information?
[09:43] <ikonia> please don't feed the troll
[09:44] <Ongavezyr> joschi: SSH AND OPENBSD_SSH SERVER INSTALLED
[09:44] <Ongavezyr> in the localhost the services working correctly
[09:45] <Ongavezyr> when I try connect from another pc in local area network : Connections refused
[09:45] <Ongavezyr> I was check the sshd config already
[09:45] <Ongavezyr> The keys regenerated
[09:45] <Ongavezyr> What is the problem? ( 10.04 )
[09:46] <ikonia> so I suggest you take it elsewhere
[09:46] <ikonia> there is no problem, as you'e made it up as you're here to troll, nothing more, the same way you where doing in ##slackware
[09:47] <Ongavezyr> ikonia bro: hush now please
[09:47] <Ongavezyr> please let's play with your legos
[09:47] <Ongavezyr> it is a real problem and I need to some help
[09:48] <robertpayne> Is there any way to automate the input required by tasksel install lamp-server?
[09:48] <ikonia> then you shouldn't troll channels
[09:48] <ikonia> robertpayne: never tried that to the be honest, you can do it within the package with answer files
[09:49] <robertpayne> ikonia: ok that should work.. doesn't really matter how just weird bug when running it in a shell script ( other tasks after it ) it doesn't install the root user was thinking that may fix it
[09:50] <Ongavezyr> When I try to connect with telnet from another pc, the connections refused too
[09:51] <WalterN> ok, I just took a picture of my server box
[09:51] <WalterN> time to put picture on my server and show all
[09:52] <WalterN> :D
[09:53] <jpds> Ongavezyr: Then it's probably a firewall in the way.
[09:53] <Ongavezyr> the firewall ( ufw ) uninstalled already
[09:56] <dominicdinada>  if someone has installed gnome and xorg on a server is it safe to remove the interfaces ?
[09:57] <robertpayne> ufw annoys me.. seems it's worse than just creating the iptables.rules file
[09:59] <Ongavezyr> I will delete the iptables?
[10:01] <dominicdinada> WalterN: What are your specs?
[10:02] <Ongavezyr> ikonia bruder?
[10:02] <WalterN> dominicdinada: intel atom based
[10:02] <WalterN> 6 drives attached
[10:03] <dominicdinada> how much space?
[10:03] <WalterN> 3x 2tb RAID5 and RAID1 boot
[10:04] <WalterN> 5th drive is just something I stuck in there, might attach it to RAID 5 sometime
[10:04] <WalterN> erm, 6th drive, rather
[10:04] <dominicdinada> ahhh
[10:04] <WalterN> just a sec
[10:06] <dominicdinada> so you have 3 2 tb arrays ?
[10:08] <WalterN> 4, including the one that is not currently doing anything useful :P
[10:08] <WalterN> and 5 including the one thats in this computer
[10:08] <WalterN> XD
[10:11] <WalterN> http://jewelcreekkennels.com/img_0063.jpg
[10:11] <WalterN> http://jewelcreekkennels.com/img_0068.jpg
[10:11] <ikonia> WalterN: please stop
[10:11] <ikonia> WalterN: this is a support channel
[10:14] <Ongavezyr> ikonia do not hysterical, please
[10:16] <ximal> Q : Is there a way to run programs such as deluge thru the cli or as a process that way on my server I can just set it up for remote web viewing etc ?
[10:17] <ximal> Q : Is there a way to run programs such as deluge thru the cli or as a process that way on my server I can just set it up for remote web viewing etc ?
[10:33] <Ongavezyr> Where is my good friend, ikonia?
[10:33] <WalterN> chewing me out? (heh)
[10:38] <alexharrington> hey. I've installed 10.04 server 64 bit and am now installing OpenLDAP. I've followed the official docs to get things going but now I want to change the cn=config root password to something else but i can't figure out how to do it?
[10:50] <Ongavezyr> ikonia brotha , the perpetual light shine upon you
[10:51] <Ongavezyr> Allah is accompanied by on your way badjew brotha
[10:57] <alexharrington> nm. I got an answer over on OpenLDAP. Thanks
[11:09] <ximal> Q : Is there a way to run programs such as deluge thru the cli or as a process that way on my server I can just set it up for remote web viewing etc ?
[11:10] <ximal> kinda like running it as a service etc ... where it's running but no gui verification .. only cli ?
[11:13] <topyli> ximal: transmission maybe?
[11:14] <topyli> it has web and terminal interfaces and can be daemonized
[11:14] <ximal> yeah
[11:14] <ximal> but i could just install flux if i want web functionality
[11:15] <topyli> if you don't want the web inteface, just don't use it. use the terminal one
[11:15] <ximal> just curious
[11:16] <ximal> upon adding stuff to transmission can you set it to ask you where to put each file ?
[11:22] <kaushal> hi
[11:22] <kaushal> when i do apache2ctl -M i see mpm_worker_module (static)
[11:23] <kaushal> is there a way to set it to mpm_prefork_module (static)
[11:23] <kaushal> ?
[11:24] <RoyK> kaushal: apt-get install apache2-mpm-prefork
[11:24] <kaushal> do i need to uninstall apache2-mpm-worker ?
[11:24] <RoyK> I don't think you can have both installed, no
[11:24] <RoyK> it's somehow one out of two (or three?)
[11:25] <kaushal> oh ok
[11:25] <RoyK> either prefork, or event or worker
[11:25] <kaushal> so if i do apt-get install apache2-mpm-prefork, it will uninstall apache2-mpm-worker ?
[11:25] <RoyK> iirc php only works with prefork, at least that's default
[11:25] <kaushal> automatically
[11:25] <joschi> kaushal: MPMs can be dynamically be changed in apache httpd 2.4
[11:25] <RoyK> kaushal: try it
[11:25] <joschi> kaushal: with apache httpd 2.2 you can only have one
[11:26] <RoyK> joschi: Lucid only has 2.2
[11:26] <joschi> RoyK: yes. apache httpd 2.4 is not released yet ;)
[11:26] <RoyK> ah :)
[11:29] <robertpayne> what's new in 2.4? :P
[11:30] <robertpayne> reading docs myself haha stupid question
[11:50] <robertpayne> Can you completely wildcard a ssl certificate? It's for access to dev utilities like phpmyadmin postfixmyadmin etc.. so the browser warnings wont really be a bother
[12:04] <Iceman_B> hi
[12:05] <Iceman_B> how much is a server installation dependant on the hardware it was installed with? can i just pick up my hdd, put it in a new system and expect it to run fine?
[12:05] <jpds> Yes.
[12:05] <jpds> You'll have to 'rm -vf /etc/udev/rules.d/70-persistent-net.rules' so that eth0 comes back.
[12:06] <Iceman_B> what does that file do?
[12:06] <jpds> Locks down network interface names to MAC addresses.
[12:07] <Iceman_B> ah, right
[12:07] <Iceman_B> any additional steps after removing that file?
[12:07] <Iceman_B> reboot? restart network? do a little dance?
[12:07] <jpds> Shutdown the machine and move the drive across?
[12:07] <jpds> Dance is optional.
[12:08] <Iceman_B> aw :( not mandatory?
[12:08] <Iceman_B> o you meant remove the file BEFORE I move the drive
[12:08] <Iceman_B> they thats clever
[12:08] <Iceman_B> *hey
[12:09] <Callum__> jpds: does Ubuntu regenerate 70-persistent-net.rules?
[12:10] <jpds> Yes.
[12:10] <Callum__> Ah
[12:10] <patdk-lap> any idea how to make pacemaker/heartbeat bring ip and ipv6 ip as non-primary?
[12:11] <patdk-lap> bring up :)
[12:11] <patdk-lap> just woke up like 3min ago, guess I'm not thinking yet
[12:12] <Callum__> I should go to bed soon...
[12:13] <binBASH> timezones timezones........
[12:13]  * Iceman_B dcc's coffee.jpg to patdk-lap
[12:13] <Callum__> Aug 10 11:13pm...
[12:13] <patdk-lap> na, shower always wakes me
[12:13] <Iceman_B> oh
[12:13] <Iceman_B> well there goes a perfectly good cup of coffee then :(
[12:13] <Callum__> shower didn't help me this morning =P still felt like crap
[12:13] <binBASH> Aug 10 13:13:38 CEST
[12:13] <patdk-lap> but pacemaker is annoying me :)
[12:14] <Callum__> UTC+12 here
[12:14] <patdk-lap> using it with IPv6addr
[12:14] <Callum__> sorry, can't help you patdk-lap =P
[12:14] <Callum__> don't know anything about pacemaker
[12:14] <Iceman_B> UTC+2(1+dst) here
[12:14] <patdk-lap> but can't figure out how to bring the ip up with preferred_lft/depress option
[12:14] <patdk-lap> so linux doesn't use that new ip as the default
[12:15] <Callum__> its really cold here :<
[12:15] <binBASH> patdk-lap: They have chan at #linux-cluster ?
[12:15] <patdk-lap> dunno
[12:16] <binBASH> at least according to their website ;)
[12:18] <Iceman_B> did the shutdown command change in 10.04?
[12:18] <Iceman_B> "shutdown now" doesnt seen to power down the system completely
[12:18] <Iceman_B> it goes back to some menu
[12:19] <patdk-lap> heh, I always just type halt
[12:20] <binBASH> poweroff for me :D
[12:25] <jussi> Iceman_B: "shutdown -h now"  iirc
[12:25] <jussi> and -r for a restart.
[12:32] <Iceman_B> thanks
[12:37] <misha> can I make dns server with one ip? and can you give me a guide for making dns server? thanks:)
[12:39] <misha> anyone here?P:
[12:40] <Pici> misha: I'd start with the server guide: https://help.ubuntu.com/10.04/serverguide/C/dns.html
[12:41] <misha> I have 9.04 the guide will work for me?
[12:42] <Pici> misha: https://help.ubuntu.com/9.04/serverguide/C/dns.html
[12:42] <misha> thanks
[12:43] <misha> also maybe you know, I had postfix but it didn't send all the mail and sometimes falls. Do you have any other good mail server? thanks again
[12:44] <patdk-lap> heh, I highly doubt that is a postfix issue
[12:44] <patdk-lap> the two big issues would be, didn't configure postfix correctly
[12:45] <alex88> hi, i'm tring to use aria2c to download files...but after give command it remains on "[#1 SIZE:0B/0B CN:1 SPD:0Bs](4more...) [TOTAL SPD:0Bs]"
[12:45] <patdk-lap> or you attempted to run it without a smarthost/relay, and have no idea of proper edicate on the internet when talking to other email servers, so they banned you
[12:45] <alex88> after some time it says timeout, but with wget it works
[12:45] <misha> I installed with the ubuntu guide and with ehcp and both of them the postfix did a problems
[12:46] <misha> (ehcp = control panel) but I deleted it
[12:46] <patdk-lap> did a problem?
[12:46] <misha> no I just didn't like the panel
[12:47] <misha> I now don't using any panel its better for me
[12:48] <patdk-lap> did you setup forward and reverse dns for your email server ip? did you setup spf? did you setup your helo name to match your ip?
[12:49] <patdk-lap> heh, helo name match your dns forward name :)
[12:49] <misha> no, no and about the last question i'm not sure
[12:49] <patdk-lap> well, then 90% of the world won't accept email from you
[12:49] <patdk-lap> so postfix won't be able to send it
[12:49] <alex88> someone know another download accellerator?
[12:49] <patdk-lap> alex88, I always perfered to use wget
[12:50] <misha> gmail accepted the messages excellent and the messages was in the main folder not in spam
[12:50] <alex88> patdk-lap: but from server i can download at higher rates..i'm on a vps..and aria not start, wget works...
[12:51] <alex88> is there any way to debug?
[12:52] <patdk-lap> higher rates? I can download using wget at full gigabit speeds easily
[12:52] <patdk-lap> I haven't tried a computer with 2gb or higher connection yet though
[12:52] <alex88> for me aria get 2-3 mbyte/s higher...maybe uploading server limits per connection speed
[12:53] <alex88> maybe it tries for some kind of proxy..but i can't found config files..
[12:54] <misha> hi I didn't understand what to do on caching nameserver in https://help.ubuntu.com/9.04/serverguide/C/dns-configuration.html
[12:55] <misha> nevermind just a question can I do it with one ip?
[12:56] <alex88> misha: sure..you mean to set one ip as forwarder?
[12:58] <misha> yes
[12:58] <patdk-lap> caching dns server is what looks up dns names for your stuff
[12:58] <patdk-lap> it looks up dns names that you *don't* serve
[12:58] <xampart> why just not use 8.8.8.8
[12:58] <alex88> google sucks..use opendns..
[12:58] <patdk-lap> 8.8.8.8 is slow for me, and it's negative cache is horrible
[12:59] <patdk-lap> opendns is slow, it's negative cache is good though
[12:59] <patdk-lap> my server does a few thousand per second, really helps to have a few local :)
[12:59] <patdk-lap> to cache all the repeats
[13:00] <misha> maybe some can help me with the Apache cause on #httpd no one answer me
[13:01] <misha>  I have a problem is someone enters https://mydomain.com:80 its write: ssl_error_rx_record_too_long and I want it will be redirect to https://mydomain.com/ how can I do that?
[13:01] <misha> also .hta doesn't work
[13:02] <alex88> well, loading domain.com:80 not works? because every redirect is served after the https connection is established..if i remember right..
[13:03] <misha> no you didn't understand
[13:03] <patdk-lap> yes he does :)
[13:03] <misha> i use only https
[13:03] <Pici> Why are you specifying :80 for https connections?
[13:03] <misha> no hh
[13:03] <misha> you didn't understand
[13:03] <patdk-lap> misha, your test failed :)
[13:04] <misha> again i just want to make redirector if someone enters on https://mydomain.com:80
[13:04] <patdk-lap> https://mydomain.com:80 is invalid (unless you really really are attempting to break things on purpose)
[13:04] <misha> its possible or not?
[13:04] <patdk-lap> yes, it's possible, but you don't want to
[13:04] <Pici> misha: No sane browser will attempt to connect to https on that port.
[13:04] <patdk-lap> http://mydomain.com:80
[13:05] <misha> anyway ok
[13:05] <Pici> http is port 80, https is 443.
[13:05] <misha> i know
[13:05] <patdk-lap> why do you want to do ssl on port 80?
[13:05] <patdk-lap> and if you do, why bother redirecting?
[13:05] <patdk-lap> that error is telling you, you are not running ssl on port 80
[13:05] <patdk-lap> and you shouldn't be :)
[13:06] <misha> nevermind
[13:06] <misha> also if someone enter http://domain.com:443
[13:06] <Pici> If you're trying to redirect http/80 users to https/443 thats a different story.
[13:06] <patdk-lap> you want to redirect, http://mydomain.com to https://mydomain.com
[13:06] <patdk-lap> so your test is invalid
[13:06] <Pici> misha: Again, that shouldn't work and no one should expect it to.
[13:07] <misha> ok
[13:07] <misha> also id you know how to make redirector if someone enters http://domain.com:443
[13:07] <misha> cause its enter document 400
[13:07] <Pici> Why are people trying to connect to those ports?
[13:08] <misha> there are people that tries
[13:08] <misha> I don't know
[13:08] <misha> I just want to make redirector
[13:08] <patdk-lap> you can't :)
[13:08] <patdk-lap> you can't talk none-ssl to an ssl port
[13:08] <patdk-lap> and you can't talk ssl to a non-ssl port
[13:09] <patdk-lap> that is why they invented tls
[13:09] <patdk-lap> but no http client/server I know of uses tls
[13:09] <alex88> you cant from https://site:80 to :443, just to http too https...
[13:09] <misha> i guess you right
[13:10] <misha> anyway heres what it show:
[13:10] <misha> http://mshell.info:443
[13:10] <Pici> Try connecting to any popular site like that... like https://google.com:80/ , it won't work and it shouldn't.
[13:10] <patdk-lap> hell, why stop there
[13:10] <patdk-lap> http://gmail.com:25
[13:11] <misha> hh thanks anyway:)
[13:11] <misha> i did a redirector from http to https
[13:11] <misha> and from www to non www
[13:12] <Pici> Those two are perfectly valid.
[13:12] <alex88> if someone enter www.site where he should go?
[13:12] <misha> see your self
[13:12] <misha> http://www.mshell.info
[13:12] <patdk-lap> alex88, home :)
[13:13] <alex88> patdk-lap: ROFL...btw...certificate invalid for www..
[13:13] <misha> i know
[13:13] <Pici> www.mydomain.com is an additional A record for mydomain.com
[13:13] <misha> cause I forgot to backup the key for www. certification
[13:14] <Pici> er, for MY domain.
[13:14] <alex88> :) that's the problem...redirects must be done after https..
[13:14] <misha> and now I can't use the crt for www
[13:14] <alex88> lol
[13:14] <misha> so thats why I make redirector from www to non wwww
[13:15] <Patrickdk> alex88, you mean redirect before :)
[13:15] <Patrickdk> in this case
[13:15] <alex88> i think that if someone accept certificate there is no reason to redirect after..
[13:15] <Patrickdk> paypal had the same issue forever, it annoyed me, i only ever type in paypal.com
[13:16] <Patrickdk> and get invalid cert
[13:16] <alex88> Patrickdk: no, the problem is that redirects stands after ssl..
[13:16] <alex88> Patrickdk: how much time ago? sure you were not in a MITM? :)
[13:17] <Patrickdk> alex, well ever since I started using paypal (pre-2000), till about 2years ago
[13:17] <Patrickdk> and that was from several different ip's
[13:17] <alex88> oh...strange..i thank they have money to buy 2 certs :)
[13:17] <Patrickdk> and the cert was always the same one, for paypal.com and www.paypal.com, but only a cn=www.paypal.com
[13:18] <alex88> oh kk..
[13:19] <Patrickdk> people just don't get how certs work, dunno why
[13:19] <Patrickdk> well, I do know why, no one ever bothers to explain it to people :)
[13:20] <alex88> maybe they're afraid some ppl break ssl :)
[13:21] <alex88> that would be a bad thing :)
[13:23] <misha> on the dns configuration in ubuntu I see there "ns      IN      A       192.168.1.10" and I don't have it in my file, add it?
[13:24] <alex88> misha: you just need a dns server to set for your home network?
[13:25] <misha> to set my domain on my server
[13:25] <misha> I hate to use free dns service
[13:25] <alex88> oh...don't think so
[13:26] <misha> i'm using now http://freedns.afraid.org/
[13:31] <misha> Replace 1.168.192 with the first three octets of whatever network you are using.
[13:31] <misha> octets?
[13:34] <misha> where the support gone?P:
[13:34] <Pici> An octet is each part of the ip address 192, 168, 0, 1 are all octets.
[13:35] <misha> lets say my ip is 32.21.634.6
[13:36] <misha> I need to write there 32 21 and 634?
[13:36] <Pici> Well that would be weird because you can't have an octet higher than 255, but yes, those would be the first three.
[13:36] <misha> thanks:P
[13:42] <alex88> if i want to do a bash script to process all *.001 files in a dir?
[13:43] <alex88> i don't remember how, because i have to run lxsplit -j to each file ending with .001
[13:43] <maswan> Is it possible to change the default kvm network, or am I better off adding another rfc1918 network if I don't want nat/forwarding?
[13:47] <alex88> ok got it
[13:49] <misha> cya:P
[13:49] <alex88> BYE :)
[14:01] <raubvogel> Dumb question: if www-data's homedir is /var/www, why it does not own/have rights in that directory
[14:06] <joschi> raubvogel: because www-data could change the permissions of /var/www otherwise
[14:06] <joschi> raubvogel: and since your web server runs in the user context of www-data and might allow script execution within that context, that's a bad idea ;)
[14:06] <patdk-wk> unless your using webdav
[14:08] <raubvogel> joschi: So, if I want to run stuff such as reprepro, I should run that as another user and then make the directory being offered by the web server owned by this user (read-only by www-data) ?
[14:09] <joschi> raubvogel: yes. just chgrp the directory to www-data group and make it g+rx
[14:10] <joschi> raubvogel: as an alternative: make it world readable, but that's sort of last resort ;)
[14:30] <alexharrington_> i've got an OpenLDAP directory running on 10.04 and am using libnss-ldap to connect to it. When I do "getent passwd" I only see a few hundred of the thousands of entries that should be returned
[14:31] <alexharrington_> in the past I had to add a line to the slapd.conf to return more results on searches - but we're now using the recommended cn=config scheme and I can't figure out how to apply the setting
[14:31] <alexharrington_> i know the setting is now called olcSizeLimit and I want it set to 5000, but no matter where I apply it it doesn't seem to do anything
[14:32] <alexharrington_> equally limiting it to 10 results doesn't seem to cut the number returned either so perhaps there's another limit somewhere?
[14:33] <alexharrington_> i've tried asking on #openldap but the silence in there is deafening :D
[14:38] <smoser> Daviey, ping
[14:38] <Daviey> smoser: Y'all
[14:39] <smoser> 2 items
[14:39] <smoser> bug 615529 has a branch attached that i would like you to take
[14:39] <smoser> tiny code change for the loader stuff to work around a kvm bug.
[14:40] <smoser> second question, is how would you like me to send you patches for such things
[14:40] <hggdh> Daviey, ttx: please have a looka t bug 615646
[14:41] <Daviey> smoser: branching lp:~ubuntu-core-dev/eucalyptus/devel and proposing merge to that makes it better IMO... or for a trivial change you could just bounce a patch over :)
[14:41] <ttx> hggdh: looking
[14:41] <Daviey> smoser: using ~ubuntu-branches is better for SRU's in this instance.. i feel
[14:42] <smoser> Daviey, i looked there, that branch is old
[14:42] <smoser> at least i thought
[14:43] <smoser> ah. i must have looked at some other one.
[14:43] <smoser> are you core dev, Daviey ?
[14:43] <Daviey> smoser: no
[14:43] <smoser> it seems like thats a bad location then :)
[14:43] <smoser> so you have to ask someone to pull from yours ?
[14:43] <Daviey> smoser: Exactly...
[14:44] <Daviey> smoser: Hopefully use ~ubuntu-server-dev soon
[14:44] <Daviey> but there isn't anywhere else i can think of that is collaborative
[14:45] <smoser> Daviey, virt group i think most are members of.
[14:45] <smoser> or very easily could be
[14:46] <Daviey> yeah
[14:47] <Daviey> smoser: that makes good sense..
[14:47] <smoser> well, only due to lack of -server-dev
[14:47] <smoser> yours is the real solution.
[14:47] <Daviey> smoser: We have it now... just nobody has applied :)
[14:47] <smoser> anyway, that patch is there, please review and pull.
[14:48] <Daviey> smoser: Okay.. probably doing an upload later today.. I'll land it in my branch then get ttx to pull and push it (if that is OK ttx)
[14:51] <ttx> Daviey: I should be able to do that.
[14:51] <Daviey> super!
[14:59] <hallyn> kirkland: fyi, still building (will take awhile here), but I'm testing https://code.launchpad.net/~serge-hallyn/ubuntu/maverick/qemu-kvm/merge-0.12.5
[15:00] <kirkland> hallyn: excellent -- push a PPA package, and i'll test here too
[15:00] <sergevn> aargh, darwin calendarserver is broken for a year now :(
[15:01] <hallyn> kirkland: alas, one thing i failed to do was bring my gpg keys :)  I'll generate new ones and get a ppa in the next few hours
[15:02] <kirkland> hallyn: doh!
[15:02] <hallyn> yup
[15:02] <kirkland> hallyn: 1024?  just ask kees to crack it for you :-)
[15:02] <kirkland> :-P
[15:02] <hallyn> kees: ^  that should take you waht, 2 mins?
[15:03] <smoser> on your transmeta laptop it will take longer than that
[15:03] <smoser> slightly
[15:03] <kirkland> haha
[15:03] <hallyn> smoser: joker.  mind you, THAT one has my gpg keys :)
[15:04] <patdk-wk> does openssl support larger keys correctly now?
[15:04] <patdk-wk> I know the last time I played iwth keys >8k it has issues, it would make them, but I couldn't use them
[15:07] <hallyn> NOt enough entropy.  Need to pl;ay some klondike to get more.
[15:08] <patdk-wk> run vmware, it seems to generate all kinds of entropy on my system
[15:08] <patdk-wk> atleast that is the only difference between this computer and the one at home, and the home one never has any entropy
[15:10] <hallyn> hm, kvm alas doesn't seem to do the same
[15:25] <kpettit> what's a good jabber server that's good for scripting?  I plan on using python to do a bunch of scripting with a jabber server, but there are alot of them out there
[15:36] <sergevn> has anyone calendarserver working on lucid?
[15:45] <ttx> hggdh: re: bug 615646 -- that used to work, right
[15:49] <SKuhaneck> what is the best way to notify the package manager about software installed from source?
[15:49] <jpds> There isn't a way?
[15:50] <lau> what mean the first digit in ubuntu dpkg version ? 4:2.11.3-1ubuntu1.3
[15:50] <jpds> lau: An epoch.
[15:51] <Pici> SKuhaneck: checkinstall?
[15:51] <Pici> Not exactly what you're asking though..
[15:51] <lau> jpds: a sort of date or timestamp ?
[15:52] <lau> I think I understand the 2.11.3-1 but the 4: Oo ?
[15:52] <jpds> lau: http://www.debian.org/doc/debian-policy/ch-controlfields.html
[15:52] <SKuhaneck> Pici: the main thing I am looking for is if I install a webserver from scratch and I install a package that requires a webserver, how do I stop it from trying to install apache as a dependency?
[15:52] <jpds> lau: Look at 5.6.12 Version.
[15:53] <lau> ok thx
[15:54] <lau> so now if I want to upgrade to a newer version what are my ways to do ?
[15:54] <lau> I want ot keep distribution version but use for this specific package a n+1 distribution version
[15:54] <lau> ie. keep hardy but use lucid version for this specific package
[15:55] <jpds> Append ~hardy1#
[15:55] <jpds> lau: Append something like ~hardy1 to the version string?
[15:56] <bogeyd6> SKuhaneck, just download the package from packages.ubuntu.com
[15:57] <Pici> SKuhaneck: I don't recall if checkinstall allows you set a 'provides' field.  I'm sure its frowned upon, but I suppose you could create a dummy package that just provides an httpd server.
[15:57] <hggdh> ttx: yes indeed. I did not test on 2.0 until now, but on 1.6.2 it worked
[15:57] <hggdh> ttx: and... this sounds like a blocker
[15:57] <ttx> noted
[15:58] <SKuhaneck> bogeyd6: the package I am compiling does not exist
[15:59] <soren> lool: I see you've touched user-mode-linux semi-recently.
[15:59] <soren> lool: Does it actually work for you?
[16:00] <soren> lool: I'm getting stuff like this: [   82.905031] linux[2845] general protection ip:7f80d64b962c sp:6282fcb0 error:0 in libc-2.12.so[7f80d6482000+17a000]
[16:01] <impi> hello, I am following this howto: http://cumu.li/2008/5/13/recompiling-php5-with-bundled-support-for-gd-on-ubuntu I have only one question. how can i install php 5.2.X instead of 5.3 ? i want to run magento and this software doesnt work on php 5.3
[16:05] <SKuhaneck> Pici: from the reading I just did, it looks like checkinstall will register it with the package manager and should do what I want, thanks
[16:05] <bogeyd6> SKuhaneck, you told pici you were trying to install a package that had a dependency and you didnt want the dependency
[16:05] <Pici> bogeyd6: I understood what he meant.
[16:05] <lool> soren: I used it many years ago in my previous company because it seemed more secure than the solutions requiring root, but the performances were abysmal; I remember fixing it to build in Ubuntu, but that's about the involvement I had with the package
[16:06] <impi> anyone that can help me with my php issue?
[16:07] <soren> lool: All of what you just said is true for me, too, just offset a couple of years :)
[16:07] <soren> lool: But ok, no worries.
[16:07] <lau> jpds: I was thinking about /etc/apt/preferences but have no clue how to implement-it
[16:07] <Pici> SKuhaneck: I just checked the manpage, checkinstall does have a --provides argument, providing 'httpd' should prevent your other packages from pulling in apache2.
[16:08] <lool> soren: 2/3 years ago, I had noticed that it wasn't maintained anymore; it still gets updated upstream to build, but I wonder whether anybody cares for it; is it relevant for you?
[16:08] <lool> soren: BTW met with a couple of coworkers of yours Saturday evening in NY (Debconf); I've asked one to wave hello  :-)
[16:10] <soren> lool: It might be relevant. I'm not willing to invest a whole lot of time in it, but it may be really convenient for testing Openstack stuff.
[16:10] <papertigers> lool: debconf was in NY, i wish i would of went
[16:10] <soren> lool: Noone has waved yet :)
[16:13] <impi> how can i get apt-get source php5 but get php 5.2.X instead of 5.3 ????????
[16:16] <tomsdale> impi http://ubuntuforums.org/showthread.php?t=1447401     couple of posts down with the Pin-Priority
[16:17] <impi> tomsdale, thank you, let me read this quick.
[16:18] <Daviey> impi: Offically, you need to use hardy, jaunty or karmic :)
[16:18] <impi> DavidLevin, officially, i should slap my client and his dev ;)
[16:18] <impi> but yea, thats about as official as i can get
[16:20] <tomsdale> had the same issue installing vtiger, requires 5.2 for some reasons.
[16:22] <Daviey> hggdh: Are you ready to test a new upstream checkout, hopefully fixing registration issues?
[16:24] <hggdh> Daviey: most certainly :-)
[16:24] <Daviey> hggdh: Rockin', building in the PPA atm..
[16:24] <Daviey> (queued)
[16:24] <hggdh> Daviey: k. Meanwhile, I will take 20 min
[16:24] <Daviey> hggdh: That is a fantastic idea :)
[16:32] <impi> tomsdale, it seems you have saved my 'ss
[16:32] <impi> i have been sitting here for like nearly 6 hours now, thanks for that tip dude
[16:34] <tomsdale> :-) np - actually someone else just saved my a lot of time on howtoforge - guess karma does exist.
[16:51] <impi> mmm anyone that know how i can get around Patch suhosin.patch does not remove cleanly (refresh it or enforce with -f)
[16:52] <ttx> mathiaz: please see https://bugs.launchpad.net/ubuntu/+source/openldap/+bug/571057/comments/27 with high priority, so that it doesn't block 10.04.1 release
[16:52] <mathiaz> ttx: working on it
[16:52] <ttx> mathiaz: great  :)
[16:56] <Kaelten> Anyone able to give any recomendation between apt-mirror, apt-proxy, or apt-cacher
[16:58] <patdk-wk> ya, I didn't like any of them
[17:01] <qman__> yeah, after a lot of frustration with those I opted for a squid caching proxy
[17:01] <Kaelten> patdk-wk: lol, well I have to use something to create a local repository
[17:02]  * patdk-wk is using squid
[17:02] <qman__> zero maintenance that way
[17:02] <patdk-wk> before that, I was just doing a mirror
[17:02] <qman__> just crank up the cache size and point your apt clients at it, or better yet, transparently redirect with your firewall
[17:03] <Kaelten> I'd be more likely to abuse varnish for this purpose than squid I suppose.
[17:04] <patdk-wk> why?
[17:04] <patdk-wk> varnish wasn't made for this, it's a reverse proxy, not a normal proxy
[17:05] <patdk-wk> "You can, but you probably don't want to. Doing it requires significant amounts of DNS magic and a huge Varnish VCL file. "
[17:06] <qman__> with squid, all you have to change are the size settings so it'll cache the bigger debs
[17:06] <qman__> and then mark the packages.gz files as never cache
[17:07] <Kaelten> well I'll already have varnish running and up so adding a few lines to the vcl won't hurt
[17:09] <Kaelten> and honestly, and this is of course just my opinion, I find squid to be antiquated and cumbersome, I haven't used it in several years, so my opinions may be stale.
[17:09] <qman__> well, the config file is enormous
[17:09] <qman__> but it's set up correctly out of the box
[17:10] <qman__> would be nice if they split it up
[17:11] <patdk-wk> heh, I just switched to squid3, I like it
[17:12] <patdk-wk> I haven't used squid for years like 10years, till recently
[17:12] <qman__> $ wc -l /etc/squid/squid.conf
[17:12] <qman__> 4433 /etc/squid/squid.conf
[17:12] <Kaelten> ugh 4.4k lines?
[17:12] <Kaelten> that's pretty hefty
[17:12] <patdk-wk> ya, 90% of it is comments
[17:12] <qman__> yeah
[17:13] <qman__> and setting it up, you'll only change maybe 10 of them
[17:13] <qman__> grep -v ^\# /etc/squid/squid.conf | wc -l
[17:13] <qman__> 335
[17:14] <patdk-wk> squid2.conf 402 lines
[17:14] <patdk-wk> squid3.conf 374 lines
[17:14] <patdk-wk> atleast in my config
[17:14] <Kaelten> wc -l default.vcl
[17:14] <Kaelten> 226 default.vcl
[17:14]  * patdk-wk smacks kaelten, that is a totally different kind of proxy :)
[17:15] <patdk-wk> you can compare apache proxy to squid
[17:15] <Kaelten> even that's significantly less config
[17:15] <patdk-wk> apache?
[17:16] <patdk-wk> it has tons of modules it has to load, each 2 config lines :)
[17:16] <Kaelten> hrm, now that I think about it probably not, but it feels it because they chop it up
[17:16] <sponzor> hi. how to make logs for every ip that is trying to connect to my server... and what service is he trying to connect etc...?
[17:16] <hggdh> mathiaz: mind if I set bug 615646 as critical? This has to be resolved before FF
[17:16] <patdk-wk> sponzor, ulog :)
[17:16] <jpds> sponzor: iptables?
[17:17] <qman__> grep -R -v ^\# /etc/apache2 | wc -l
[17:17] <qman__> 952
[17:17] <qman__> and that's not even a proxy, just a basic apache setup
[17:17] <Kaelten> I'm moving away from apache as well though,  in our new deployments we're going almost exclusively nginx
[17:17] <sponzor> hmm never tryed with iptables.. so they can do logs?
[17:18] <patdk-wk> sponzor, yes, alittle limited, unless you use ulog instead
[17:18] <hggdh> Daviey: should we just wait (scheduled to start building in 1 hour), or should we try a gentle poke?
[17:18] <sponzor> ok. i will geave it a try :P. tnx
[17:19] <Kaelten> grep -R -v ^\# /etc/nginx/ | wc -l
[17:19] <Kaelten> 540
[17:19] <Kaelten> for a proxy setup
[17:19] <qman__> I use apache because I know it well, but I openly admit it's quite heavy
[17:19] <patdk-wk> heh, my lighttpd proxy is like 50lines
[17:19] <Kaelten> patdk-wk: well I was including all of the files inside nginx
[17:19] <qman__> that apache server is running a single site for vnstat php frontend
[17:19] <Kaelten> the proxy config itself is only 25 lines
[17:20] <patdk-wk> Kaelten, dunno, my whole lighttpd.conf file, is 50lines
[17:20] <patdk-wk> and it includes no other files
[17:20] <Kaelten> lol
[17:20] <Kaelten> fair enough
[17:20]  * patdk-wk hates the debian/ubuntu version of lighttpd config file
[17:21] <Kaelten> I always had problems with lighttpd messing up with php processes
[17:21] <patdk-wk> never had that issue
[17:23] <patdk-wk> I've been running from lighttpd 1.4.6 though 1.5.x, no issues with php ever
[17:23] <patdk-wk> but many many many people join #lighttpd with php completely broken
[17:24] <jpds> qman__: So; disable some modules, and tune it? :-)
[17:25] <qman__> jpds, CPU time is cheaper ;)
[17:25] <mathiaz> hggdh: Critical - I'm not sure about it
[17:25] <mathiaz> hggdh: https://wiki.ubuntu.com/Bugs/Importance
[17:25] <mathiaz> hggdh: I don't think it's critical
[17:26] <mathiaz> hggdh: and I'm not sure how FeatureFreeze impacts it
[17:26] <FunnyLookinHat> Ok you crazy hackers - how can I diff the output of two different ls commands?  i.e. ls /a/* and ls /b/* ?
[17:28] <jpds> Well, neither nginx/lighttpd are in main so ...
[17:28] <qman__> funkyHat, ls /a/ > a; ls /b/ > b; diff a b
[17:29] <qman__> er
[17:29] <qman__> FunnyLookinHat, ^
[17:30] <qman__> there might be a way to do it without creating files, but after thinking for a moment, I decided it wasn't worth figuring out
[17:31] <Kaelten> patdk-wk: I was running a site that during high traffic php would starting erroring every time until I restarted lightty
[17:32] <Kaelten> at the time I found an old bug report describing the issue that had been around for a few years, and last I heard was still an issue
[17:32] <patdk-wk> how where you starting php?
[17:32] <Kaelten> as fcgi
[17:33] <patdk-wk> heh?
[17:33] <patdk-wk> fcgi is a protocol, not a way to start php
[17:34] <patdk-wk> did you start php yourself, spawnfcgi, or have lighttpd do it
[17:34] <FunnyLookinHat> qman__: right right - but temp files are lame...   how about diff <(ls -lh /a/*) <(ls -lh /b/*) ?  :)
[17:35] <patdk-wk> I kept having mod_php leaking memory like nuts for me
[17:35] <patdk-wk> switched the webmail servers to lighttpd, no more issues
[17:36] <qman__> FunnyLookinHat, good to know
[17:36] <smoser> kirkland, ping
[17:36] <FunnyLookinHat> Yeah found it via twitter - heh
[17:36] <patdk-wk> that line fits on twitter?
[17:36] <FunnyLookinHat> easily?
[17:36] <patdk-wk> man, I totally don't get pacemaker
[17:36] <tomsdale> for a smaller webserver - would you recommend bind or mydns.
[17:36] <patdk-wk> I just want to add an option to the ip up command
[17:37] <patdk-wk> tomsdale, neither
[17:37] <jpds> tomsdale: Surely you mean DNS server?
[17:37] <patdk-wk> why install dns on a webserver?
[17:37] <tomsdale> :-) I have to decide between one of them for my hosting server. with DNS on the server I can generate records from the same panel.
[17:38] <patdk-wk> hopefully mydns is fixed, it used hangs on my friends install, everytime mysql connection drops
[17:38] <Kaelten> patdk-wk: er ya sorry, it's been about 5 years ago I was running that site, so I'm not too sure, I think it was managing the processes though
[17:38] <jpds> tomsdale: BIND is in main and supported by Ubuntu.
[17:38] <tomsdale> I'm using ispconfig 3. patdk-wk hm. Doesn't sound too convincing.
[17:39] <patdk-wk> tomsdale, well, it's easy to fix
[17:39] <tomsdale> jpds: that's enough for me :-) thx u2
[17:39] <patdk-wk> I'm just to lazy to fix it for them
[17:40]  * patdk-wk also notes that mydns hasn't been updated on that server for atleast 7years
[17:41] <tomsdale> bind it is
[17:41] <Fidelix> Guys, what email filtering you can recommend?
[17:41] <patdk-wk> hmm, /dev/null does a great job
[17:42] <Fidelix> patdk-wk, what do u mean?
[17:42] <Fidelix> i can redirect to /dev/null, ok. But how do i check message headers to do that?
[17:42] <patdk-wk> oh, I wasn't checking
[17:42] <jpds> Fidelix: procmail?
[17:43] <patdk-wk> there are many many options, depends on what you want to write the rules in
[17:43] <patdk-wk> procmail, maildrop, sieve, ...
[17:43] <Fidelix> I want the easy one. hehehehe
[17:43] <patdk-wk> there is no easy one
[17:44] <Fidelix> OK. Then the less hard.
[17:44] <patdk-wk> none of them have anything better than manual editing of the config files
[17:45] <Fidelix> OK. I am using postfix with Maildir. Can i write rules to procmail for delivering to certain folders on the user's Maildir?
[17:45] <patdk-wk> sure
[17:45] <Fidelix> Excelent. I guess there is no gui to do that...
[17:45] <patdk-wk> not that I know of
[17:46] <patdk-wk> I personally use maildrop, and I setup the company to use seive
[17:46] <Fidelix> patdk-wk, please dont be mad with what i'll ask.
[17:46] <Fidelix> What is maildrop?
[17:46] <patdk-wk> heh
[17:46] <Fidelix> The agent that delivers the email to the users?
[17:47] <patdk-wk> yep
[17:47] <patdk-wk> instead of procmail
[17:47] <Fidelix> OK. And what is seive?
[17:47] <patdk-wk> same thing
[17:47] <Fidelix> how do i know which one i'm using?
[17:47] <Fidelix> main.cf?
[17:47] <patdk-wk> depends on your mailserver
[17:48] <patdk-wk> see what one it's setup to use
[17:48] <Fidelix> Postfix / dovecot. I'll check the config files.
[17:48] <patdk-wk> unless you isntalled one, I doubt it's using any
[17:48] <patdk-wk> well, then you have two things
[17:48] <patdk-wk> either postfix is going direct, not using anything
[17:49] <patdk-wk> or it's configured to use the dovecot deliever lda, in that case it could, or could not be using sieve, depends on dovecot config
[17:49] <Iceman_B> how much does a clean install differ from a version that's been upgraded from 8 -> 9 -> 10 ?
[17:49] <Fidelix> It seems to be using maildrop, because maildrop's line is uncommented
[17:49] <Fidelix> maildrop  unix  -       n       n       -       -       pipe
[17:49] <Fidelix>   flags=DRhu user=vmail argv=/usr/bin/maildrop -d ${recipient}
[17:49] <patdk-wk> that doesn't mean it's using it
[17:49] <patdk-wk> it just means it knows how to use it :)
[17:49] <Fidelix> Oh, true,
[17:50] <Fidelix> And do you know a way to check this?
[17:51] <patdk-wk> postconf | grep transport
[17:51] <Fidelix> default_transport = smtp
[17:52] <Fidelix> It seems to be making the deliver directly.
[17:52] <Fidelix> mailbox_transport =
[17:52] <patdk-wk> look at local_transport and virtual_transport
[17:52] <patdk-wk> and that :)
[17:52] <patdk-wk> yep, direct
[17:52] <Fidelix> local_transport = local:$myhostname
[17:52] <Fidelix> I guess thats ugly
[17:52] <Fidelix> Its not even going through dovecot, right?
[17:53] <patdk-wk> if you did, virtual-transport=maildrop, mailbox_transport=maildop, it would use maildrop instead
[17:53] <patdk-wk> nope
[17:53] <patdk-wk> for mine I have, virtual_transport = dovecot
[17:53] <patdk-wk> then dovecot uses sieve
[17:53] <Fidelix> Is sieve easier than maildrop?
[17:53] <patdk-wk> dovecot unix    -       n       n       -       -      pipe
[17:53] <patdk-wk>   flags=DRhu user=nobody:dovecot argv=/usr/libexec/dovecot/deliver -f ${sender} -d ${user}@${nexthop} -n -m ${extension}
[17:54] <Fidelix> btw, i found a GUI to procmail
[17:54] <patdk-wk> I dunno sieve
[17:54] <Fidelix> Its named Procbuilder.
[17:54] <patdk-wk> the webmail package does sieve though, so it's user friendly
[17:54] <patdk-wk> I manually write all my maildrop, have >10k lines of maildrop rules
[17:54] <Fidelix> z0mg
[17:55] <patdk-wk> my procmail script was getting way to complex, so I switched to maildrop
[17:55] <Fidelix> But ur writing maildrop regex manually?
[17:55] <patdk-wk> yep
[17:55] <Fidelix> Thats... well... something
[17:55] <Fidelix> I guess maildrop has no gui either
[17:56] <patdk-wk> dunno
[17:56] <patdk-wk> webmail package will do them all
[17:56] <patdk-wk> seieve is just easier for it to modify the scripts
[17:56] <patdk-wk> thunderbird has sieve support (still have to edit manually though)
[17:56] <Fidelix> wow. So i can create the rules on thunderbird and sieve will import 'em?
[17:57] <Fidelix> Thats wonderful.
[17:57] <patdk-wk> kindof :)
[17:57] <nxvl> kirkland: what does it means when i've a {!} on my byobu?
[17:57] <patdk-wk> you create then in thunderbird sieve plugin, and it will upload them and active them on the server
[17:57] <Fidelix> Thats really awesome.
[17:58] <patdk-wk> but seieve was made to be nicely done, just no one has bothered on the client side to make it nice yet
[17:58] <Fidelix> Should i just apt-get install sieve?
[17:58] <patdk-wk> it's not a seperate package
[17:58] <Fidelix> Its in webmail package?
[17:58] <patdk-wk> it's in dovecot
[17:58] <Fidelix> oh
[17:58] <Fidelix> ok then. Should i set mailbox_transport = dovecot then?
[17:59] <patdk-wk> if you wish
[17:59] <Fidelix> or default_transport ?
[17:59] <patdk-wk> but I would setup dovecot first
[17:59] <patdk-wk> no, I would think you want smtp to be default
[17:59] <Fidelix> What has to be done in dovecot?
[17:59] <Fidelix> ok.
[17:59] <patdk-wk> you need to configure the lda section
[17:59] <patdk-wk> setup sieve, ...
[17:59] <patdk-wk> whole sections of the wiki on dovecot's website do this step by step
[18:00] <Fidelix> ok. i'll look for it
[18:02] <Fidelix> Is this it? http://wiki2.dovecot.org/LDA
[18:03] <patdk-wk> that, then the LDA/Postfix page, and at the bottom the sieve page
[18:04] <Fidelix> OK. Thank you very much patdk-wk.
[18:04] <Fidelix> patdk-wk ++
[18:31] <smoser> kim0, so i loaded http://foss-boss.blogspot.com/2010/08/ubuntu-server-10041-virtual-release.html from an ec2 instance. in us-east-1, but i iddn't see a new dot show up on the map
[18:31] <smoser> is there a way that i can see where the server says that my ubuntu server is ?
[18:33] <smoser> oh. i see now, that there may be a dot.  from the initial view it looked like only 3 dots in US.
[18:33] <smoser> but a zoom in shows many more
[18:35] <raubvogel> Anyone using reprepro? I created a gpg key and now how do I intert it in the distribution file (per http://www.porcheron.info/setup-your-debianubuntu-repository-with-reprepro/)?
[18:49] <Fidelix> patdk-wk, are u there?
[18:49] <patdk-wk> ya
[18:51] <hggdh> Daviey: there?
[18:52] <Daviey> hggdh: as always :)
[18:54] <kim0> smoser: that's true .. you need to zoom in .. enjoy :)
[18:54] <smoser> kim0, but is there a way to know where it placed my server ?
[18:54] <kirkland> nxvl: see `man byobu` ...  apport
[18:54] <kirkland> nxvl: you have an apport crash report
[18:55] <kim0> smoser: not easily ..
[18:55] <kirkland> smoser: thanks for the errno improvements
[18:55] <kim0> smoser: I could tell by looking at the database
[18:55] <kim0> smoser: or you could by looking up your server's IP into geoIP :)
[18:55] <smoser> thats what i figured.
[18:55] <Daviey> kim0: Do you have access to the database now?
[18:55] <kirkland> smoser: one regression though....
[18:55] <kim0> smoser: maybe next time we'll make it green or something :)
[18:55] <kirkland> smoser: "errno 3"
[18:56] <kim0> Daviey: no I don't :D :D why is your position that important to you :D
[18:56] <kirkland> smoser: i had fixed it to omit matches in the description, if you search for a number explicitly
[18:56] <smoser> hm..
[18:56]  * kim0 remembers Daviey 's city name was something funny right ?
[18:56] <Daviey> kim0: I live in the backwaters of hidden country :).. would be nice to see it on the map :)
[18:57] <kim0> so next time .. people want to see themselves in a different marker .. and be able to change it
[18:57] <kim0> feel free folks to hack on mapuntu to add those :)
[18:58] <zul> so if im in a diferent city it will show up on the map but not the actual city where my server is in right?
[18:58] <smoser> kirkland, i dont see why thats matching
[18:58] <kim0> zul: that's why it's best to hit it from your servers
[18:58] <martian> I have a 9.10 server install that I connect to via samba, and every few days it starts timing out only on samba connections. I'll restart samba but it doesn't help. Only a system restart helps. Any ideas?
[18:58] <kim0> zul: http://maps.ubuntu.com/hit.html
[18:59] <kim0> zul: wget that from the box
[18:59] <zul> heh too busy right now
[18:59] <RoyK> martian: wierd - never seen that, and I've been (and am) running samba on quite a few platforms
[18:59] <RoyK> martian: check dmesg
[18:59] <RoyK> or /var/log/messages about the time of the hang
[19:00] <hggdh> Daviey: can you use your powers of persuasion to get the eucalyptus pacakges (at least the amd64) to build soon?
[19:00] <Daviey> hggdh: yeah.. i'm tugging on that favour quite alot atm :(
[19:00] <martian> RoyK: yeah, it's almost like the port is blocked. Checking those now..
[19:00] <Daviey> hggdh: but it's ridiculous wait atm
[19:00] <RoyK> martian: dmesg won't be of much use if you've already rebooted
[19:01] <hggdh> Daviey: indeed. BTW, you remember we have a meeting now, right?
[19:01] <kirkland> smoser: you're seeing it though, right?
[19:01] <martian> RoyK: I have not. The system has been online for a few weeks, and just a few minutes ago I lost the file sharing
[19:01] <RoyK> ok, pastebin dmesg
[19:01] <RoyK> if something fishy is going on in the kernel, it should be there
[19:01] <Daviey> hggdh: i'm there :)
[19:02] <hggdh> SpamapS: meeting now, sir ;-)
[19:02] <smoser> yeah.
[19:02] <martian> RoyK: neither dmesg nor messages contain 'smb' or 'samba' :(
[19:02] <RoyK> that wasn't waht I was looking for :)
[19:03] <RoyK> rather something related to memory or i/o
[19:03] <RoyK> or some crash
[19:03] <patdk-wk> anything unusual, assuming you are used to reading dmesg often :)
[19:03]  * RoyK is :)
[19:03] <patdk-wk> well, I know that :)
[19:04] <patdk-wk> hmm, I haven't had any samba issues, running samba on 8.04, 9.10 and 10.04, currently
[19:04] <patdk-wk> 10.04 isn't getting stressed, but the other two are :)
[19:04] <RoyK> martian: can you pastebin it? it won't reveal anything security-wise unless you've done something creative
[19:05] <patdk-wk> dmesg | pastebinit :)
[19:05] <smoser> kirkland, so: sed -n "s,^#define\s\+\([^[:space:]]\+\s\+${code}\s.*\),\1,p" ${headers}
[19:05] <smoser> that works.
[19:05] <martian> RoyK: (had to install pastebinit) http://pastebin.com/HqMMDP42
[19:05] <smoser> it seems that [^\s] isn't really working correctly. i replaced with [^[:space:]]
[19:06] <kirkland> smoser: ack
[19:06] <kirkland> smoser: that looks better
[19:06] <RoyK> martian: cat /proc/uptime
[19:06] <kirkland> smoser: okay, i like your changes;  it seems, though, that there's some bikeshedding about what package to put this utility in
[19:06] <smoser> yeah, which is silly.
[19:06] <smoser> i stated to respond with "not all tools there are ubuntu specific"
[19:07] <smoser> where else would it possibly go ?
[19:07] <martian> RoyK: uptime: 1112744.62 8686773.93
[19:07] <kirkland> smoser: i considered:  ubuntu-dev-tools, devscripts, coreutils, and linux-tools
[19:07] <kirkland> smoser: i have half a mind to put it in byobu, actually
[19:08] <smoser> :)
[19:08] <RoyK> martian: hrmf
[19:08] <RoyK> martian: and nothing in the samba logs either?
[19:08] <RoyK> martian: what does netstat -ln --inet say - does it still listen? or lsof -p `pidof smbd` ?
[19:09] <martian> RoyK: Well, I actually (perhaps a bad idea?) just wiped the samba logs before restarting the service in hopes to catch the problem in a fresh log
[19:09] <RoyK> ops
[19:09] <RoyK> move them away next time :)
[19:10] <patdk-wk> mv /var/log/samba /var/log/samba.old :)
[19:11] <martian> yeah, I kinda smacked myself for that before even coming here... frustrated thinking makes for mistakes
[19:11] <martian> Netstat: http://pastebin.com/Qxhm7Ep0
[19:11] <patdk-wk> have you adjusted any sysctl settings?
[19:12] <martian> No sir
[19:12] <RoyK> martian: ok, have you restarted samba, and it still doesn't work?
[19:12] <martian> correct
[19:12] <RoyK> I mean restarted, not reloaded...
[19:12] <martian> even stopped... waited... started
[19:13] <RoyK> stop ; ps axf|grep smb ?
[19:13] <martian> and I'm not the only client machine unable to get in, so afaik it's not a client issue
[19:13] <RoyK> martian: can you check if the process is really restarting? check its pid
[19:13] <patdk-wk> hmm, martian, how do you authenicate users?
[19:14] <hggdh> Daviey: don't know what is it you did, but amd64 just built!
[19:14] <patdk-wk> on one of my systems (the 8.04 ubuntu one), I'm using AD
[19:14] <ball> Okay, that's strange.  I just booted an Ubuntu Server box and it's giving me a graphical login screen.  Wonder if it booted from the wrong disk.
[19:14] <patdk-wk> and winbind fails assert sometimes, randomly, once a month
[19:14] <patdk-wk> if I start winbind, all works again
[19:14] <ball> Ah yes, that's it.
[19:14] <martian> oh, hello processes: http://pastebin.com/Ti99SPEr
[19:15] <martian> patdk-wk: kerberos
[19:15] <patdk-wk> so via winbind
[19:15] <patdk-wk> next time you can't login, or is that now?
[19:15] <patdk-wk> make sure winbind is running :)
[19:15] <martian> I'm sorry, let me rephrase that; the system is kerberized, but samba is not.
[19:15] <Daviey> hggdh: /me rocks :)
[19:15] <martian> I still smbpass
[19:16] <RoyK> martian: lots of those processes are old, can you kill them and try to start samba again after that?
[19:16] <patdk-wk> mine, samba uses the system users
[19:16] <patdk-wk> and the system nsswitch, uses winbind :)
[19:16] <RoyK> martian: if you remove the log files, old samba processes will still write to the erased files, not the new ones
[19:17] <martian> RoyK: Is there a simple way to kill all those processes?
[19:17] <RoyK> killall -9 smbd
[19:17] <patdk-wk> those processes die when users logout/disconnect the share
[19:18] <RoyK> patdk-wk: they won't be of much use if he's debugging them and they write to deleted files :þ
[19:18] <martian> Hmm, after stopping the service and killing the processes, they still remain
[19:18] <patdk-wk> ya, just explaining normal usage
[19:18] <RoyK> kk
[19:19] <RoyK> martian: ps 21257 ; kill -9 21257 ; sleep 1 ; ps 21257
[19:19] <RoyK> does that still show the process?
[19:19] <sponzor> who is http://pastebin.com/paLXd7Gk
[19:19] <sponzor> :1
[19:20] <martian> RoyK: yes :o
[19:20] <patdk-wk> LOCALHOST :)
[19:20] <martian> RoyK: zombies!
[19:20] <RoyK> martian: not in Z state
[19:20] <patdk-wk> sponzor, you must really like to ftp yourself a lot :)
[19:20] <sponzor> yeah but i m not using it? :/
[19:20] <RoyK> martian: try strace -fp 21257 and try to see what it's doing
[19:21] <patdk-wk> sponzor, once every 5 min? looks like cacti or something checking it
[19:21] <RoyK> it really should be killable in S state
[19:21] <martian> RoyK: nothing. Process is attached, then silence
[19:22] <tomsdale> I gave an lvm logical volume the name var-www. when I reboot it cannot mount it and throws an error. Can my naming convention with the - lead to a problem.?
[19:22] <sponzor> could be webalizer?
[19:22] <patdk-wk> doubt it
[19:22] <martian> RoyK: and I mean zombies figuratively, not literally ;)
[19:22] <RoyK> martian: that sucks
[19:22] <RoyK> martian: yeah, but zombie processes exist in unix :)
[19:22] <sponzor> becouse i dont have cacti :P
[19:23] <smoser> Daviey, your sponsor request has my change?
[19:23] <martian> yeah, I should have chosen a better word like... cranky toddler processes.. so, any hope for this?
[19:23] <RoyK> martian: mad hatters
[19:23] <martian> aye!
[19:23] <RoyK> martian: you _are_ root, right?
[19:24] <martian> RoyK: sudoing
[19:24] <RoyK> k
[19:24] <Daviey> smoser: I haven't placed it yet.. but the package hggdh is testing *doesn't* have it
[19:24] <RoyK> martian: just sudo -i
[19:24] <Daviey> smoser: If you want to push it to ~ubuntu-virt branch, that would be super.
[19:24] <Daviey> smoser: Unless you want to peer review every change?
[19:25] <smoser> what branch ?
[19:25] <smoser> lp:ubuntu/maverick/eucalyptus/2.0 ?
[19:25] <martian> RoyK: good call, but no avail. as actual root, ps 21257 ; kill -9 21257 ; sleep 1 ; ps 21257 the process still remains
[19:25] <smoser> i need no review of my changes :)
[19:25] <Daviey> smoser: Noooooo
[19:26] <Daviey> smoser: lp:~ubuntu-virt/
[19:26] <RoyK> martian: can you get anythinf from ltrace?
[19:26] <Fidelix> patdk-wk, know any guides for setting postfix with dovecot > procmail ?
[19:26] <RoyK> ltrace -p pid
[19:26] <Daviey> smoser: maintain the UNRELEASED status in debian/changelog please :)
[19:26] <smoser> Daviey, ok. you pasted me wrong.
[19:26] <patdk-wk> if you use procmail, you won't be using dovecot
[19:26] <RoyK> Fidelix: I think the admin guide has one
[19:26] <patdk-wk> so you just need a postfix + procmail one
[19:26] <Daviey> smoser: really?
[19:26] <mathiaz> SpamapS: on the topics of graphing bugs, I'm planning to load all of the ubuntu-server-bugs mailing list archive into a hadoop cluster
[19:26] <martian> RoyK: Operation not permitted
[19:27] <RoyK> Fidelix: use sieve
[19:27] <RoyK> martian: oh
[19:27] <RoyK> as root?
[19:27] <mathiaz> SpamapS: it should be interesting to gather some information/graph from there
[19:27] <martian> RoyK: yes, as root
[19:27] <RoyK> martian: ltrace -p 21257 ?
[19:27] <SpamapS> mathiaz: !! cool
[19:27] <Daviey> smoser: ah yes, seems i did.. sorry
[19:27] <Fidelix> RoyK, i was going to set Sieve, as patdk-wk suggested. But i found that i dont want to script for each rule i want to add.
[19:28] <RoyK> Fidelix: I think procmail is disregarded in ubuntu/dovecot
[19:28] <Fidelix> hmm
[19:28] <RoyK> Fidelix: and you don't need a separate script, it'll just be one large one
[19:28] <Fidelix> There are GUIs for procmail, and thats important to me.
[19:28] <Fidelix> I'm not good with regex, not at all
[19:28] <RoyK> well, dunno then
[19:29] <martian> RoyK: http://pastebin.com/5q1kECdJ baffled :x
[19:29] <RoyK> you don't need to use regex with sieve
[19:29] <SpamapS> mathiaz: take a look at Flume for loading
[19:29] <RoyK> martian: out of interest - pastebin ps axfu
[19:30]  * patdk-wk loves regex, I'm not good with it, but I can do most all I need to
[19:30] <Fidelix> Well, there is a GUI for setting procmail with Dovecot.
[19:30] <Fidelix> http://wiki2.dovecot.org/procmail?highlight=%28procmail%29
[19:30] <Fidelix> But i guess i'll go with Sieve.
[19:30] <Fidelix> I dont want to regret it later.
[19:30] <RoyK> Fidelix: you probably won't
[19:30] <Fidelix> And as you say, sieve seems more reliable.
[19:30] <RoyK> Fidelix: I think it's like 10 years since I used procmail :)
[19:30] <Fidelix> hahaha
[19:30] <martian> RoyK http://pastebin.com/kpqRTn5h Hmm, the smbd processes are me
[19:31] <RoyK> martian: what happens if you try to kill them with your own user, not root?
[19:31] <RoyK> I'm really at loss here - just guessing
[19:32] <RoyK> also, perhaps trying to stop/start winbindd might loosen up some - no idea - at all
[19:32] <martian> RoyK: no luck
[19:32] <RoyK> ok, start an strace -fp somesmbdpid
[19:32] <martian> RoyK: Yeah, I'm guessing I may want to simply restart the system and save those logs :)
[19:33] <RoyK> send a sigterm (kill xx) to that, and see if you get something
[19:33] <RoyK> if not, send a SIGSEGV (-11 iirc) and see if something happens
[19:33] <RoyK> or even SIGHUP or whatever
[19:34] <martian> RoyK: neither
[19:34] <RoyK> martian: also make sure you have a recent kernel for that distro version - apt-get update && apt-get dist-upgrade
[19:34] <Fidelix> Well, i'm not beeing happy with this Dovecot > Sieve wiki.
[19:34] <Fidelix> Anyway, i'll try harder
[19:35] <martian> RoyK: yeah, been holding off for a bit as it's hosting a new site (which went live TODAY ahh!)
[19:35] <martian> RoyK: I think I may need to hold off on any service interrupts until this evening.
[19:35] <tomsdale> I don't understand what 'nosuid' does in the mountoptions. Does it mean you cannot change the user/group of a file after it has been created?
[19:36] <RoyK> martian: seems you've already got an interrupt
[19:36] <RoyK> apt-get dist-upgrade just takes you to the new 9.10 (if that was what you had)
[19:36] <RoyK> it doesn't take you to 10.04
[19:36] <RoyK> you'll need a do-release-upgrade command to do that
[19:36] <impi> lol
[19:37] <Fidelix> patdk-wk, RoyK does ManageSieve comes with Dovecot 2?
[19:37] <impi> started 11 am and im done with this bloody php compile now..
[19:37] <impi> what a mess
[19:37] <Fidelix> The wiki says nothing about that.
[19:37]  * RoyK has no idea
[19:37] <martian> Well, as long as apache, mysql, and shibboleth keep chugging away until the end of the day, I'm ok :)
[19:37] <RoyK> martian: they won't be chugging worse with updated packages
[19:38] <Fidelix> Well, did u had to compile sieve when you set ur dovecot?
[19:38] <RoyK> martian: the updates you get with a dist-upgrade are just bug fixes and _maybe_ minor upgrades, nothing major
[19:39] <martian> RoyK: but I'll still need to restart to flush these processes, so I'd rather wait 3 hours
[19:39] <RoyK> martian: no, with 9.10 you won't even get minor upgrades
[19:39] <smoser> Daviey, quilt refresh..
[19:39] <smoser> what do i want to add to that to make it keep 'a/b' style diff output
[19:39] <RoyK> martian: with an LTS, you may get minor upgrades if upgrading from, say, 10.04 to 10.04.1, but only then
[19:39] <smoser> rather than $(basename)
[19:40] <martian> RoyK: and actually, dist-upgrade only has an ldap-utils update for me
[19:40] <Daviey> smoser: ?
[19:40] <RoyK> martian: no kernel stuff?
[19:40] <smoser> never mind. i found it.
[19:40] <Daviey> smoser: wassup?
[19:40] <RoyK> martian: this seems to be something bad in kernel
[19:40] <smoser> so refreshing (i think) a patch, you want: quilt refresh -pab
[19:41] <smoser> hmm.. manybe not
[19:41] <smoser> maybe its just my quilt is different than yours
[19:41] <Fidelix> 
[19:41] <Daviey> smoser: cleaner patches:
[19:41] <Daviey> dave@voodoo:~$ cat /home/dave/.quiltrc
[19:41] <Daviey> QUILT_PATCHES=debian/patches
[19:41] <Daviey> QUILT_DIFF_ARGS="--no-timestamps --no-index -pab"
[19:41] <Daviey> QUILT_REFRESH_ARGS="--no-timestamps --no-index -pab"
[19:41] <smoser> danke
[19:42] <martian> RoyK: afraid not. I do have a 10.10 machine mirroring this one, so I may swap them around at some point if this continues. This is the third time in about 2 months that it's happened.
[19:42] <martian> RoyK: well hey, thanks so much for your help! I really appreciate it!
[19:42] <RoyK> martian: why not 10.04 LTS?
[19:43] <RoyK> I only run LTS for critical servers, not betas
[19:43] <martian> RoyK: oh yeah, got my version numbers confused
[19:43] <RoyK> 10.04 => April 2010
[19:52] <RoyK> martian: I asked around a little, and all I can find is debug calls in kernel, which may not give you much valuable info if you're not a kernel hacker
[19:53] <Fidelix> RoyK, can u please pastebin your main.cf and dovecot.conf?
[19:54] <smoser> Daviey, https://code.launchpad.net/~ubuntu-virt/ubuntu/maverick/eucalyptus/2.0 is updated.
[19:54] <RoyK> Fidelix: I'm not using dovecot anymore
[19:54]  * RoyK switched to Zimbra for less hassle
[19:55] <Fidelix> oh, ok
[19:55] <Fidelix> Should i switch too?
[19:56] <fuentesjr> anyone know why someone would want to use ubuntu enterprise cloud vs xen ?
[19:56] <RoyK> Fidelix: that's up to you - zimbra is not in ubuntu, and distributed as a separate system, not suitable to run on anything but a VM or a dedicated system, but it works for me
[19:57] <smoser> Daviey, how do i build that?
[19:58]  * patdk-wk loves dovecot :)
[19:58] <Fidelix> Well, that seems lot of... work.
[19:59] <patdk-wk> ya, zymbra is a, do all my email crap, and don't annoy me setup
[19:59] <kirkland> kees: https://bugs.launchpad.net/bugs/615077
[19:59] <kirkland> kees: is that a dupe of your compiz bug?
[20:01] <kees> kirkland: I don't think so.
[20:01] <kees> kirkland: but maybe... though it's a weird symptom
[20:01] <RoyK> patdk-wk: heh - works for me (tm)
[20:03] <kirkland> kees: k, thanks
[20:04] <patdk-wk> royk, yep :)
[20:04] <patdk-wk> I wonder how well it works under load
[20:04] <patdk-wk> heavy imap load, so dovecot is nice
[20:04] <patdk-wk> from all the smartphones these days
[20:06] <RoyK> patdk-wk: I know the norwegian meteorological office uses it with drbd/heartbeat for quite a few users
[20:09] <Jinxed-> I don't know where else to ask this, so I thought I might ask it here
[20:10] <Jinxed-> I would like to have essentiall a complete video server running on my ubuntu machine 10.04, but im unsure what would be able to do the trick
[20:10] <patdk-wk> depends what you mean by, *video server*
[20:10] <patdk-wk> mythbuntu?
[20:11] <patdk-wk> or a streaming video server, like a website
[20:11] <Jinxed-> I should clairfy that it is all on a private network
[20:11] <Jinxed-> with no need to have access to the outside
[20:11] <patdk-wk> I just use mythtv for all my video needs
[20:11] <Jinxed-> 1.) I want to be able to stream webcam from any remote computer to any other remote computer and have it stored locally on my server. I have bandwidth issues on a few connections, so i would like the server to take the stream save it and rebroadcast it or something similar.
[20:12] <Jinxed-> 2.) I want to be able to stream the videos on my server via VoD to any remote computer on the network with variable quality
[20:12] <patdk-wk> heh, no idea what would do that
[20:12] <Jinxed-> 3.) It would be nice to have some sort of front end for remote users to be able to see what is in the server and be able to select what video that want to watch and the bandwidth
[20:13] <patdk-wk> sounds like you want real-time transcoding of the video stream
[20:13] <patdk-wk> that uses some cpu power
[20:14] <smoser> Daviey,
[20:14] <smoser> $ bzr bd -S
[20:14] <Jinxed-> yeah, that would be ideal
[20:14] <smoser> ...
[20:14] <smoser> bzr: ERROR: Unable to find the needed upstream tarball: eucalyptus_2.0~bzr1219.orig.tar.gz.
[20:14] <Jinxed-> I won't have very many connections at once so I don't think it should be too bad
[20:14] <Jinxed-> but if that doesn't work then the ability to transcode a video on the server to different qualities
[20:15] <Jinxed-> first, and then just stream
[20:15] <patdk-wk> that second way, you can do anything to do that
[20:15] <patdk-wk> as long as you don't do live
[20:16] <patdk-wk> sounds more like you want a security cam type system, with trascode abilities
[20:17] <Jinxed-> patdk-wk, Any idea on what I could use to capture streams to disk and rebroadcast to the original destination
[20:17] <patdk-wk> zoneminder is the security cam package that is pretty nice
[20:17] <patdk-wk> dunno if it will work the way you want though
[20:19] <RoyK> Jinxed-: how many connections? what sort of media? Mbps?
[20:19] <RoyK> Jinxed-: and what bandwidth do you have to the clients?
[20:19] <RoyK> gigE? 100Mbps? a crappy 54Mbps WIFI?
[20:19] <patdk-wk> t1 :)
[20:19] <Jinxed-> My slowest connection is abou 250kbs
[20:20] <RoyK> Jinxed-: then you'll realtime transcoding, which is _NOT_ cheap
[20:20] <Jinxed-> I can stream video at about 320x240 with 400kbs bitrate with 15 fps
[20:20] <Jinxed-> yeah
[20:20] <RoyK> Jinxed-: transcode the videos to smaller ones, and let the users choose between them
[20:21] <RoyK> Jinxed-: there are methods of allowing users to choose dynamically based on bandwidth, but I don't know any OSS that allows that
[20:21] <RoyK> Jinxed-: look into ffmpeg and the rest of the transcoding software available
[20:22] <Jinxed-> Do you know of anything that could take a stream from a remote user save it to disk and rebroadcast it?
[20:22] <RoyK> Jinxed-: still, VoD over 256kbps will suck rather hard
[20:22] <Jinxed-> haha yeah
[20:22] <Jinxed-> not every connection is that bad
[20:22] <hggdh> Daviey: bloody thing got completely lost (I tried to just upgrade the packages). I am going for a full reinstall now :-(
[20:22] <Jinxed-> some will be standard 100 Mbs
[20:22] <RoyK> Jinxed-: I think Måns Rullgård (an ffmpeg developer) has developed something like that - we used to work together some years back
[20:23] <RoyK> Jinxed-: Måns also wrote an addon to ffmpeg to allow parallel transcoding, which is rather neat
[20:24] <Jinxed-> interesting
[20:24] <RoyK> as in thread one splits the stream into i-frame (b-frame (p-frame ...) ...) and sends each i-end stream to a thread
[20:25] <RoyK> it's in ffmpeg, but the ffmpeg syntax is rather cryptic
[20:25] <Jinxed-> hmm
[20:25] <RoyK> it makes a crapy perl scrip look like a children's book
[20:26] <Fidelix> RoyK, how do i test if Sieve is working?
[20:27] <RoyK> Fidelix: sending an email containing the matches?
[20:28] <Jinxed-> hmm what about this... is there a way to setup something so I just send the stream to a server which could make it available for other who wanted to see it
[20:28] <Jinxed-> ...
[20:28] <Jinxed-> hmm
[20:28] <RoyK> Jinxed-: still, if you need to serve VoD over 256kbps, spending all that CPU power on live transcoding seems a waste
[20:29] <RoyK> Jinxed-: even VNC can do that
[20:29] <RoyK> by multicast or unicast
[20:29] <RoyK> Jinxed-: what sort of network topology are you on?
[20:29] <RoyK> is it a LAN, a WAN or an internet site?
[20:29] <Jinxed-> lan
[20:30] <RoyK> Jinxed-: wtf has only a 256kbps on a LAN?
[20:30] <Jinxed-> form of wireless
[20:31] <RoyK> what sort of wireless? even 802.11 is basically 2Mbps
[20:31] <RoyK> and that's like eight years old
[20:31] <RoyK> 802.11b, that is
[20:32] <RoyK> 802.11g is (theoretically) 54Mbps, which should suffice
[20:32] <Jinxed-> radio
[20:32] <RoyK> so not LAN
[20:32] <RoyK> what sort of radio links are these?
[20:32] <Jinxed-> wan
[20:34] <RoyK> I think it's doable, but the best guess is to start with a bandwidth test, and then choose the desired media file
[20:34] <Jinxed-> well the bandwidth test i did
[20:34] <RoyK> other methods will involve rewriting the media player and potentially the server
[20:34] <Jinxed-> gave pings with about
[20:34] <Jinxed-> 400 ms
[20:34] <Jinxed-> average response
[20:35] <RoyK> round-trip time doesn't say anything about bandwidth
[20:35] <RoyK> only about latency
[20:37] <RoyK> Jinxed-: write a small client-side thing that checks the bandwidth to the server, it can be a java applet or something like that. when you know the downstream speed, you know at what speed it can stream. Then choose the media file made for that connection
[20:37] <Jinxed-> well tests show about 180-200 kBs
[20:38] <Jinxed-> or about 1.5 megabits/sec
[20:38] <RoyK> then transcode movies into separate files, depending on the link quality available
[20:38] <RoyK> the transcoding should not be done in realtime, because that'll take a truckload of CPUs
[20:39] <RoyK> say, have the full MPEG2-TS (probably around 7Mbps) available, and add another 4Mbps, and others with 2Mbps, 1Mbps, 512kbps and so on
[20:40] <RoyK> to alternate between them in realtime, you'll need to write your own player, though, unless mplayer or vlc or someone has written a hack for this already
[20:41] <Jinxed-> hmmm
[20:41] <RoyK> you want lots of buffering on the client side, and preferably a client that can switch to another server when the one he's on goes down, that can usually be done with DNS RR
[20:41] <Jinxed-> what about recording a stream to disk
[20:41] <RoyK> that's the simple part
[20:42] <Jinxed-> that is the part im the most hung up on right now
[20:42] <Jinxed-> how I could take a stream from one remote site to another remote site and record it
[20:42] <RoyK> Jinxed-: I may be wrong, but it seems you're just touching the parts of VoD...
[20:42] <Jinxed-> to the server
[20:42] <RoyK> Jinxed-: from where do you get the video stream?
[20:43] <RoyK> a TCP or UDP stream? from DVB?
[20:44] <Jinxed-> Say I have a webcam attached to a laptop at two remote sites A and B, and I want to stream from A to B but I also want that stream recorded on my server, but because of how slow the connection is from A to the server I only want to have to tramsit it once and have the server record it while it transmits it to B
[20:44] <Jinxed-> it would have to be udp... tcp would be to slow i imagine
[20:44] <Jinxed-> for my slowest link
[20:45] <RoyK> if you're doing site-to-site webcam, you'll need to sniff the traffic and dump whatever you get
[20:45] <RoyK> otherwise, you can setup a server that handle the two and allow them  to connect to that server, which can do the copying
[20:45] <Jinxed--> you don't know of anysoftware that could take/rebroadcast
[20:46] <Jinxed--> that latter idea sounds good
[20:46] <Jinxed--> great actually
[20:46] <RoyK> I really don't know - there's a lot of webcam servers out there, and I guess most of them can store the streams
[20:47] <RoyK> but then, if it's private data, it won't be very nice to your users
[20:50] <ball> In Ubuntu Server, how do I list the partitions on a disk?
[20:50] <RoyK> ball: cat /proc/partitions
[20:51] <RoyK> ball: or just fdisk ......
[20:51] <ball> RoyK: Thanks
[20:56] <ball> RoyK: Thanks, that helped me find and mount a partition that I wanted to check prior to wiping a disk.
[20:57] <RoyK> that's a good start :)
[21:07] <Fidelix> Guys, postfix or smtpd or something is bouncing all emails back to the sender.
[21:07] <Fidelix> Its getting really annoing. Any ideas on what can i do about that?
[21:07] <ball> I'll be back later.
[21:08] <Fidelix> I've created a couple of days before a thread on UF but no response atm
[21:08] <Fidelix> http://ubuntuforums.org/showthread.php?t=1549288
[21:11] <Jinxed--> RoyK, do you know how much it would cost to get someone to help me with my system (even rough estimate)/a good place to look for help
[21:13] <RoyK> Jinxed--: I think you should define your needs better first - read up on VoD a little more - no offence, really, but you're not asking to specific questions
[21:14] <Jinxed--> well im a little flexible on what exactly i want based on how much easier/harder it would be to implement
[21:16] <RoyK> Jinxed--: can you try to simplify what you want or need?
[21:18] <Jinxed--> Simply put the ability to watch videos on demand on the network that are stored on my ubuntu machines (given my slow connection) and the ability to stream from any connected device and have those streams stored on my server (again a solution that can work with my slow connection)
[22:20] <tucemiux> how should I handle updates?  "no automatic updates", "install security updates automatically" or "Manage system with landscape"??
[22:21] <guntbert> tucemiux: landscape is paid
[22:22] <tucemiux> guntbert, thanks, i'll skip that one then
[22:23] <guntbert> tucemiux: will you be visiting the server regularly?
[22:24] <tucemiux> guntbert, the server is on my home ESXi box
[22:26] <guntbert> tucemiux: I didn't mean physically but logically - if you hav a server that is once started and the left alone I would choose "install security updates...", if not the "no automatic..." and sudo aptitude update; sudo aptitude safe-upgrade every once in while
[22:30] <rww> I do what guntbert said, with the exception that "every once in a while" is "whenever ubuntu-security-announce emails me"
[22:31] <Iceman_B> okay, when I connect additional harddrives into a system, do they magically become visible after a reboot? if not, how do I go about it?
[22:31] <tucemiux> ok guntbert thanks
[22:32] <markatto> what do I need to do to enable mod_rewrite? the apache conf files seem to be split all over the place
[22:32] <guntbert> rww: point well taken :-) (my ubuntu "servers" are only play things :-))
[22:33] <markatto> am I supposed to copy/symlink the files from mods-available to mods-enabled?
[22:33] <markatto> that looks like how it works, but it all seems very over-engineered and confusing
[22:35] <rww> markatto: sudo a2enmod rewrite; sudo service apache2 restart
[22:35] <rww> (a2dismod is the opposite of a2enmod; there are also similar commands for sites)
[22:40] <markatto> rww: do I need to enable it for the folder specifically as well? it doesn't seem to be working.
[22:52] <rww> markatto: I'm not sure what you mean?
[23:02] <markatto> rww: never mind I got it
[23:05] <Kaelten> anyone know how to tell apt-get to automattically delete packages after installing them?
[23:05] <Kaelten> (the cache files that is)
[23:15] <Daviey> hggdh: Any news? :)_
[23:18] <hggdh> Daviey: yes. Good and bad. Which one first?
[23:20] <hggdh> Daviey: good: I can install (it it keeps running) CLC+Walrus
[23:20] <hggdh> Daviey: bad: the bloody (CC+SC) fail
[23:20] <hggdh> so, no cigar
[23:21] <hggdh> Daviey: I am now trying to install eucalyptus-cloud on the (CC+SC) to see if it runs
[23:22] <hggdh> (i.e., leaving eucalyptus-cloud down, does CC+SC start?
[23:28] <hggdh> Daviey: yes, it works! So I am now installing the NCs
[23:29] <Daviey> hggdh: Hmm.. ok - not the results i anticipated
[23:29] <Daviey> hggdh: Hmm, ok - was this a fresh install or an upgrade?
[23:29] <Daviey> I want to try and establish exactly what elements aren't working :/
[23:46] <hggdh> Daviey: a fresh install. I tried an upgrade, and thing went completely south
[23:47] <hggdh> Daviey: I am now on topo3 (CLC+Walrus), (CC+SC), (NC)+
[23:48] <Daviey> Hmm.. ok.. Would you mind doing a mini truth table?  I'm starting to get a little confused :S
[23:51] <hggdh> :-)
[23:51] <hggdh> awwwww c'mon! just a little? I myself am going bonkers
[23:53] <hggdh> Daviey: I *think* this is some sort of check done in the code for "required" components. I am *guessing* they check on something unique to the components -- like a specific .jar
[23:54] <hggdh> Daviey: this guess comes from the fact that installing eucalyptus-cloud -- and leaving it stopped -- appears to have done the trick on the (CC+SC) machine
[23:55] <hggdh> Daviey: of course, after that many tries I am not sure I am able to think anymore, but it stands to (my doubtful) reason
[23:55] <hggdh> Daviey: I will be more sure in a few, I am finishing the NC installs