[00:29] <pteague> any reason why ubuntu compiles the php gd with the libgd library instead of the using the --with-gd ?  the php gd library has more options built in & is apparently configured a bit better specifically for php
[00:33] <KillMeNow> no idea
=== bnjmn is now known as turbosloth
[00:55] <foob12> clear
[01:21] <ScottK> Generally we try to avoid using embedded copies of libraries.
[01:46] <necro\> drac cards - is there no way for console access via putty/ssh shell?
[02:04]  * genii sips and ponders how non-console ssh access might work
[03:10] <pteague> genii: i think it might be more of an issue of not having sshd installed...  otherwise i'd wonder about non-console ssh access as well >.>
[03:13] <pteague> ScottK: yeah & normally my preference is to extend a library in a new library for something specific, but php team i think has reasons for their own internal library...  they have to support $#!@% windows & keep things consistent between OSs... i believe there's the same issues with dates & locale
[03:23] <Debolaz> pteague: I don't know about that, the PHP developers has a history of making very bad design choices.
[03:24]  * genii thinks about PEAR and shivers
[03:26] <pteague> i know the dates & locale issue was a pain because i specifically remember hearing somebody complaining about certain linux distributions using standard libs rather than compiling the php libs causing issues where the answers didn't match across platforms
[03:28] <pteague> mentioning PEAR is just so... wrong.  PEAR requires no knowledge of how a computer works apparently...  php core requires knowledge of C which i think would make the average PEAR dev's brain explode
[03:29]  * Debolaz hugs perl.
[03:29] <pteague> although some PEAR ideas have been good...  DB, MDB, & MDB2 i would like to think evolved into PDO which php has needed for oh so long
[03:31] <pteague> i think had i gotten started with perl i probably would have used it more
[03:38] <Debolaz> I'm not always overjoyed by the language design of perl, but CPAN makes it far, far superior to other languages in terms of getting stuff done in the real world.
[03:40] <aubre> thoughts on cloning Ubuntu Karmic node controllers? hardware is identical
[03:42] <ScottK> If you're using Karmic for production, my thought it not good.
[03:49] <aubre> it's a proof of concept system
[03:49] <aubre> it won't be a production system
[03:49] <ScottK> OK.  Well if the hardware is identical, it's probably mostly to clone the system.
[03:49] <aubre> I have a node controller that's working and I just want to copy it to another system
[03:50] <aubre> what software should I use to clone the system?
[03:53] <ScottK> One thing to be careful of, if you system has /etc/iftab it has the mac address of the ethernet device in it and will have to be changed on the new system.
[03:53] <ScottK> Probably dd.
[04:02] <aubre> can dd work across a network?
[04:04] <aubre> I figured I'd need something like Ghost or Clonezilla
[04:04] <twb> aubre: no, it works with files and streams.  You can, however, use netcat to run a stream over network ports.
[04:05] <aubre> twb: hmm something to think about
[05:26] <uvirtbot> New bug: #436366 in net-snmp (main) "package libsnmp15 5.4.1~dfsg-12ubuntu3 failed to install/upgrade: " [Undecided,New] https://launchpad.net/bugs/436366
[06:07] <Anirban1987> is there any difference between the terminology nameserver and dns server ?
[06:07] <twb> Anirban1987: a nameserver could be using NIS or LDAP instead of DNS
[06:07] <twb> or that windows protocol... NETBIOS?
[06:08] <Anirban1987> twb: na na , I am on Ubuntu Server 9.04 . asking from its point of view
[06:08] <twb> Anirban1987: however, sites that don't use DNS for hostname resolution tend to be rare these days.
[06:08] <twb> Certainly Ubuntu can provide name services over any of those protocols.
[06:08] <twb> Why do you ask?
[06:09] <Anirban1987> twb: Actually I am planning to use OpenDNS for my server .
[06:11] <Anirban1987> twb: Currently I can see it lists 208.130.152.7 and 209.130.152.8 as DNS
[06:13] <twb> Anirban1987: OK, in /etc/resolv.conf "nameserver" means the DNS name server.
[06:14] <twb> Anirban1987: resolv.conf is where your server acts as a DNS *client*.
[06:14] <Anirban1987> twb: yeah , I was editing that file
[06:14] <twb> Anirban1987: OpenDNS is unrelated to that.  OpenDNS is a service where *your* server can have a name, which *other* people can resolve to your IP address.
[06:15] <twb> And obviously OpenDNS are just one supplier; for there's also groups like dyndns.org and zoneedit.
[06:16] <twb> Or you can just host your own DNS records, which is what my company does (because they're masochistic old-school BUGs).
[06:17] <sub> OpenDNS is not unrelated to resolv.conf
[06:17] <sub> OpenDNS offers a free resolving/recursive DNS service.
[06:18] <twb> sub: well, you know what I meant
[06:19] <twb> There's negligible relation between your DNS client configuration and how your own A records are hosted.
[06:20] <sub> I don't see where Anirban1987 says he was trying to host his own DNS, but I haven't scrolled up yet either :P
[06:20] <twb> sub: using opendns = hosting your a record.
[06:20] <sub> No...
[06:20] <twb> Specifically, OpenDNS are hosting your records -- you aren't hosting them on your own server.
[06:21] <sub> No, OpenDNS's primary thing is "use us for resolution" AFAIK
[06:21] <Debolaz> sub is right.
[06:21] <twb> Oh, I'm sorry.  I must have misunderstood
[06:21] <twb> Anirban1987: ignore most of what I said
[06:21] <sub> twb: I think you may have them confused with ZoneEdit, et al :)
[06:21] <twb> sub: yeah, I did.
[06:22] <twb> I looked at the first para of wikipedia's OpenDNS page and misunderstood it.
[06:22] <sub> no worries
[06:22] <Anirban1987> hmmm.... !
[06:23] <Anirban1987> How is Webmin as control panel for Ubuntu 9.04 ?
[06:24] <mushroomblue> disliked, for some reason.
[06:24] <mushroomblue> apparently, it is not to be discussed in this channel.
[06:25] <twb> This channel doesn't support using webmin.
[06:26] <twb> This is for good reasons: webmin isn't part of Ubuntu, and it has an abysmal design, negligible quality control and input validation, and doesn't try very hard to be secure.
[06:26] <Debolaz> twb: While I'm really not a fan of webmin, are there any good alternatives? Just curious.
[06:27]  * Debolaz has been looking for a way to manage his server park from a central interface and landscape is a bit too anemic.
[06:28] <Anirban1987> Actually I can't afford CPanel and Plesk. Any other free panel ??
[06:28] <twb> Debolaz: unfortunately, nope
[06:30] <twb> Officially I think we recommend ebox, but personally I didn't like the look of it.
[06:30] <twb> Obviously for an intelligent admin, you should be using ssh.
[06:31] <Debolaz> I've been looking a bit at ebox, but it seems the only way I get to try their centralized control interface is to purchase it.
[06:31] <twb> The problem is that you don't want to hand your non-technical customer's staff that much power, when all the really need is a couple of operations like "add new user"
[06:32] <twb> Debolaz: you mean centralized configuration, like a single config interface for your whole cluster or network of servers?
[06:32] <Debolaz> twb: While giving staff members the possibility to perform certain tasks like the one you mentioned is one goal, I also want to reduce the amount of work I have to do. The "ssh" way simply does not scale.
[06:33] <twb> A web interface is not automatically any more scalable than an ssh interface.
[06:34] <Debolaz> But doing things manually through the command line will never scale, whileas a management interface designed for the task can.
[06:34] <twb> Debolaz: my question is: are you trying to have a single administrative interface to manage a cluster or network of servers?
[06:35] <Debolaz> That is what many fail to understand when they disregard any non-"hardcore-1337-admin" solutions.
[06:35] <Debolaz> Now, I don't really care if the interface is web based or curses based. When I say web based, that's not really the important aspect.
[06:35] <Debolaz> twb: Yes.
[06:35] <pwnguin> of course ssh doesn't scale
[06:35] <pwnguin> thats why we have clusterssh
[06:35] <twb> Debolaz: then you should look at chef, puppet or cfengine.
[06:36] <twb> pwnguin: clusterssh is really the wrong solution, particularly if your nodes are heterogeneous.
[06:36] <pwnguin> twb: yea, more of a joke answer
[06:36]  * Debolaz keeps hearing cfengine mentioned.
[06:36] <twb> Debolaz: since they, unlike webmin (or ebox, I guess), are actually intended to address that issue.
[06:36] <pwnguin> cfengine is the classic solution
[06:36] <twb> I wouldn't even bother mentioning cfengine if I didn't have reservations about puppet's implementation
[06:37] <twb> Particularly, puppet requires ruby everywhere, generally wants you to install the latest unstable version of the puppets and puppetmaster, and the wire protocol between them has no version negotiation step -- it just assumes you have the same version of puppet installed everywhere.
[06:38] <pwnguin> twb: so you're saying puppet is built to solve the problems it created
[06:39] <twb> pwnguin: I'm saying that it's a bit too "alpha" for me to put manage an international network with it.
[06:39] <twb> Not that doing it by hand is really any better...
[06:39] <pwnguin> anyways, these things do require you to be able to script everything.
[06:39] <twb> pwnguin: meh, I'm not worried about that.
[06:40] <twb> I would definitely trust puppet/chef/cfengine MORE than landscape, since at least I can RTFS the former.
[06:41] <pwnguin> people use rhn
[06:41] <pwnguin> nobody seems to cry out
[06:42] <twb> People also use RHEL, and I can't stand THAT either
[06:42] <pwnguin> we have rhel at work
[06:42] <pwnguin> i'm not clear why
[06:42] <twb> Hysterical raisins
[06:43] <pwnguin> heh
[06:43] <pwnguin> not far off the truth
[06:43] <twb> Particularly if you deployed it around 2004, when it was the obvious choice
[06:43] <pwnguin> oh no, this stuff goes back to 2001
[06:43] <pwnguin> at least
[06:43] <twb> I mean, unless you are in Europe, SuSE is automatically out, and Debian is out because they can't even commit to a release date.
[06:44] <pwnguin> heh
[06:44] <twb> What else is left?  Just proprietary stuff like Solaris
[06:44] <pwnguin> well, i work for a college
[06:44] <pwnguin> Debian isn't automatically out
[06:44] <twb> Oh man, I was called in to look at the server of a "college" (read: dorm house).
[06:44] <twb> It had Gentoo on it.
[06:45] <pwnguin> heh yea
[06:45] <pwnguin> my CS dept decided to switch from debian to gentoo
[06:45] <twb> And its disks (one of which had died) had LVM, md RAID *and* EVMS configured on it
[06:45] <twb> Fucking idiots
[06:45] <twb> Gentoo is not for production any more than LFS or Debian/unstable is
[06:45] <pwnguin> because the student admins knew gentoo and claimed it could do the stuff they needed that debian couldn't
[06:46] <pwnguin> anyways, the college i work for is a bit enterprisey
[06:47] <pwnguin> lots of rhel
[06:47] <pwnguin> but stuff like rhel4
[06:47] <twb> A lot of our customers still run RHEL4 and Knoppix.
[06:47] <twb> It's really horrible
[06:48] <mushroomblue> don't hate the distro.
[06:48] <pwnguin> apparently someone used ubuntu for a project, and decided it was too different and hard
[06:48] <mushroomblue> hate the admins.
[06:48] <mushroomblue> a good gentoo admin is a valuable tool.
[06:48] <pwnguin> sudo
[06:48] <twb> mushroomblue: well, I hate yum, for example.
[06:48] <mushroomblue> as do I.
[06:48] <pwnguin> a good gentoo admin would be an equally effective ubuntu/debian admin
[06:49] <twb> I don't care if a developer wants to run gentoo on his workstation, but I do not want the fileserver or the mailserver to be running gentoo.
[06:49] <twb> ESPECIALLY if I am asked to babysit that server.
[06:49] <mushroomblue> now, sure.
[06:49] <twb> But I'm also not convinced that Gentoo has the same kind of rigorous Q/A that Debian provides.
[06:49] <mushroomblue> because gentoo is broken horribly.
[06:50] <mushroomblue> but make no mistake, when gentoo was popular, and not being ruined by seemant, it made ubuntu look like a bunch of hacks.
[06:50] <mushroomblue> </rant>
[06:50] <twb> IMO Ubuntu is still a bunch of hacks.
[06:50] <twb> They take Debian, run it into the ground, then install the latest GNOME and call it a release.
[06:50] <pwnguin> im not even sure what that means but
[06:51] <twb> That's basically my view of Ubuntu
[06:52] <twb> Ubuntu is on some of my servers because 1) it's more like Debian than the alternatives; and 2) Debian didn't have a solid release schedule.
[06:52] <pwnguin> I don't see what's wrong with putting the latest GNOME release in the hands of users
[06:52] <twb> pwnguin: I'm speaking of servers.  I don't really give a shit what users put on their workstations.
[06:52] <twb> Because essentially I'm not responsible for that end
[06:52] <pwnguin> then why complain about running debian into the ground
[06:53] <twb> pwnguin: because running it into the ground applies to servers, too.
[06:54] <twb> A more polite way of putting it would be to say that Ubuntu has, especially historically, primarily been a desktop distro and not worried too much if it shipped with server-related bits in a bit of a mess.
[06:54] <pwnguin> on the other hand, with about a month left, ubuntu+1 is looking kinda wobbly
[06:55] <twb> For example, old versions of Ubuntu didn't officially support package upgrades unless you used the GUI update tool
[06:55] <twb> afk beer
[06:55] <pwnguin> so who runs ubuntu+1 on a server?
[07:02] <twb> pwnguin: not me!  I would only track LTS except for scratch servers.
[08:38] <mattcen> Hey all. Just asked this question on #ubuntu, with no joy. *May* get more joy here: I'm looking for some information on how to use 'uvesafb' on Jaunty, so that my TTY will support 1680x1050 (my monitor's native resolution). Anybody here able to help? I fear that perhaps given that servers are usually headless, this is a niche area.
[08:42] <qman__> mattcen, getting that to work properly is a complete crapshoot
[08:42] <qman__> anything other than the standard 4:3 resolutions is really difficult to get working
[08:43] <mattcen> qman__: I suspected as much. Perhaps not worth spending time on then. Thanks for the reply.
[08:43] <mattcen> I know a friend at work who has done it; he uses Debian, and has almost given up on X completely. Might ask him how he went about it. Knowing him he probably messed with it for ages.
[08:44] <qman__> you could at least set it to 1024x768 or 1280x960 to get more on the screen
[08:44] <mattcen> Yeah I've done that.
[08:44] <qman__> but anything else is like voodoo magic, it's possible, but highly dependent on your hardware, and might require compiling your own kernel
[08:45] <mattcen> ... Which is where things start getting messy etc.
[08:48] <mattcen> qman__: Thanks again! I shall go back to my Apache peformance testing (uni assignment) :-)!
[09:57] <artillerytx> Hey guys for some reason i keep getting domain.com/AXFR/IN denied in my daemon.log
[09:58] <artillerytx> looks like it can't transfer the zone records
[10:02] <_ruben> from trusted sources or not ? :)
[10:03] <artillerytx> Im trying to forward to my everydns.net slave dns
[10:03] <artillerytx> so i've added for allowed transfers both name servers IP address and then the axfr.everydns.net IP as well
[10:04] <_ruben> those denied messages in the log tend to list the ip address as well
[10:05] <artillerytx> yeah its not any of the name servers IP addreses
[10:06] <artillerytx> wait wait
[10:06] <artillerytx> its the IP for the axfr.everydns.net
[10:06] <artillerytx> so now that i've added that should work fine
=== georg_ is now known as kwork
[11:08] <soren> smoser: http://pastebin.com/m249f845b
[11:08] <soren> smoser: http://lists.xensource.com/archives/html/xen-changelog/2007-02/msg00075.html
[12:05] <erichammond> soren: Here is one way to help figure out what times are good for the individuals in a group of people: http://www.doodle.com
[12:07] <erichammond> It does get a bit unwieldy if you want folks to specify every single hour for every day in a week.
[12:07] <erichammond> Let me know if you'd like help figuring it out or want me to set up an example.
[12:31] <Jeeves_> Anyone here who works for Canonical? Something is broken on your internets.
[12:45] <soren> Jeeves_: Elaborate, please.
[12:46] <soren> erichammond: Neat. I'm just not sure how to map the whole DST farce in it.
[12:47] <soren> erichammond: We switch DST at different times, and it's going to be tricky to get things adjusted in the transition periods.
[12:48] <soren> Jeeves_: I can see http://www.ubuntu.com/, I can see http://launchpad.net/, I can log into machines in the DC..
[12:58] <Jeeves_> soren: pm
[13:14] <zul> morning
[13:19] <ScottK> soren: Good luck picking a time that makes everyone happy.  This is never easy.
[13:22] <soren> ScottK: I'm not trying to make everyone happy. That will inevitably fail. :) I'm trying to make noone (particularly myself) completely miserable.
[13:22] <soren> The current meeting time fails spectacularly at that.
[13:31] <kwork> by defaylt is postfix chrooted in ubuntu 8.04 ?
[13:31] <kwork> default*
[13:32] <soren> kwork: Yes.
[13:33] <kwork> so if i set up some script in aliases i should give path relevant to chroot ?
[13:34] <soren> I'm not sure, really.
[13:40] <Jad> hello
[13:40] <Jad> ubuntu server is like windows azure? that is its a cloud OS also?
[13:41] <soren> It is nothing like Azure. Several cloud providers offer Ubuntu on their platform (e.g. Amazon and Rackspace).
[13:42] <Jad> ok, but i mean in concept
[13:43] <Jad> i mean your app runs on the server or a bunch of cloud connected servers
[13:43] <Jad> so your app is hosted elsewhere
[13:43] <Jad> isn't this singularity?
[13:43] <soren> Ubuntu is an operating system.
[13:44] <Jad> yes
[13:45] <soren> What do you mean "yes"?
[13:46] <soren> Ubuntu is a operating system. You can install it on your laptop and your servers. It runs applications and services.
[13:46] <soren> Some cloud providers offer Ubuntu as the base operating system in their clouds.
[13:46] <Jad> i mean i know what you mean, though i don't get much this cloud thing. does the server host processes that you run? for example if you run a game on your pc, will it run on the cloud connected to the pc, that is a ubuntu cloud for example?
[13:47] <_ruben> plain ubuntu server has nothing to do with clouds
[13:47] <soren> Most people do not have a cloud connected to their PC.
[13:47] <soren> (whatever that means)
[13:47] <soren> Do you know what an operating system is?
[13:48] <Jad> sure
[13:48] <soren> I'm not convinced. Your questions suggest otherwise.
[13:48] <Jad> the OS thats connected to the ubuntu server is ubuntu or windows
[13:48] <soren> What?
[13:48] <soren> And OS is not connected to a server. That makes no sense.
[13:48] <Jad> that is like a winxp connects to windows server via domain controller
[13:48] <soren> ?!?
[13:48] <soren> No.
[13:49] <Jad> in xp i can log to a domain on a win server, doesn't ubuntu server let me log in to it the same way xp does?
[13:49] <soren> Are you attempting to answer my question or are you asking something completely different?
[13:50] <soren> In either case, this is very confusing.
[13:52] <Jad> sorry for confusing you, 1 sec
[13:53] <Jad> what i mean is i heard that ubuntu server can support roaming and domain accounts. also i know in xp you can log in as a user on a server's domain. maybe ubuntu server has the same thing also
[13:53] <soren> Ubuntu is an operating system. It runs on a variety of different hardware, though usually on i386 and x86_64 based laptops, desktops, and servers.
[13:53] <soren> What does that have to do with cloud computing or Azure?
[13:54] <_ruben> Jad: ubuntu (server) has ldap support, which is similar to windows' active directory
[13:55] <Pici> 'It has support' meaning there are packages that you can install and then configure that will add that functionality.
[13:57] <soren> Pici: Who are you talking to?
[13:59] <Pici> soren: just clarifying _ruben's statement for Jad.
[14:00] <soren> Pici: Ok.
[14:00] <soren> This whole conversation is confusing :)
[14:00] <Pici> I agree.
[14:04] <Jad> sorry was away, i see what you mean soren. i'll look at the ubuntu cloud computing, i guess windows cloud computing isn't for public yet
[14:08] <_ruben> Jad: perhaps you could try to explain to us what your actual goals are? :)
[14:09] <Jad> i want to build a game, can it run on multiple servers thereby sharing cpu load among the cloud? do i have to make the load management of cpu myself or the cloud of ubuntu has an app for it?
[14:11] <soren> What kind of game is this?
[14:12] <Jad> MMO using unreal or most likely cryengine
[14:12] <soren> I mean.. I think very, very few games will need to run on more than one machine.
[14:12] <Jad> better editor imho
[14:12] <Jad> ok
[14:12] <soren> And no, Ubuntu doesn't offer this. You have to "cloudify" your application yourself.
[14:13] <Jad> so ubuntu offers to cloudify virtual machines for exmaple?
[14:13] <_ruben> beowulf cluster comes to mind, which is aimed at distributed computing, never worked with it though
[14:15] <Jad> maybe it clouds diskspace
[14:27] <giovani> game servers generally aren't so cpu intensive that you'd need to cluster it
[14:28] <giovani> since most of what they do are exchange coordinates, etc between players
[14:28] <giovani> not render
[14:30] <_ruben> WoW has rather largish server farm though ;)
=== cyphermo1 is now known as cyphermox
[14:32] <giovani> _ruben: I'm sure -- however, that's an obviously specialized case
[14:32] <gamla_kossan> anyone familiar with syslog-ng in here?
[14:32] <giovani> there are only a handful of games of that scale
[14:32] <gamla_kossan> having a problem, can't get it to start wiht my new conf..
[14:32] <giovani> gamla_kossan: somewhat -- I've used it -- just ask
[14:32] <gamla_kossan> giovani: thanks - can you spot something odd with this snippet?
[14:33] <gamla_kossan> http://fpaste.org/QKee/
[14:40] <giovani> gamla_kossan: nothing stands right out -- there's one guy in #syslog-ng
[14:40] <giovani> might want to ask him
[14:40] <Jad> does ubuntu offer to cloudify its cpu time so that different apps running can run on different cpus of different servers? for example maybe when a new process on ubuntu cloud comes to run, ubuntu will see which server has the least cpu usage and make the process run on that server?
[14:40] <gamla_kossan> giovani: thanks =)
[14:41] <giovani> gamla_kossan: or post to the mailing list for syslog-ng
[14:41] <gamla_kossan> good idea!
[14:42] <Jad> or is it merely clouding disk space among virtual machines on ubuntu?
[14:47] <soren> Jad: Neither.
[14:47] <soren> Jad: You have to do all of that yourself. I already told you.
[14:48] <soren> Jad: You have to make sure it scales across numerous servers both in terms of processing and storage.
[14:48] <soren> Ubuntu is not going to do that for you.
=== dingy_ is now known as dingy
=== MianoSM2 is now known as MianoSM
[15:00] <huats> hi soren
[15:00] <huats> how are you ?
[15:00] <soren> Hello.
[15:00] <soren> Not on the phone :)
[15:00] <huats> great :)
[15:01] <huats> I have a question for you :)
[15:01] <huats> of course you can RTFM me :)
[15:01] <huats> I haven't been able to find the definition of the virbr0 interface
[15:01] <huats> I have understand its usage
[15:01] <huats> understood
[15:02] <huats> but it is not mention on the help.ubuntu.com (there are only references to br0)
[15:02] <Jad> soren, ok. but what does ubuntu cloud then if not disk space or it doesn't cloud anything. i have to do it all myself?
[15:02] <Jad> does it have a cloud mangement app for example
[15:02] <huats> Jad: from my understanding (but I might be right)
[15:03] <huats> ubuntu is not going to change your application
[15:03] <huats> it will create the cloud infrastructure
[15:03] <huats> but your applications needs to be able to take advantage of it
[15:04] <Jad> ok cool, what applications already take advantage of it, vmm esxi for example?
[15:06] <soren> huats: virbr0 is created by libvirt. You can see its definition by running "virsh net-dumpxml default".
[15:06] <huats> soren: ok
[15:06] <Jad> sorry i meant vmware esxi
[15:06] <huats> thanks soren
[15:07] <soren> huats: sure thing.
[15:07] <huats> soren: and the virbr0 is used for the nat in the vm right ?
[15:07] <soren> Jad: No. VMWare is a hypervisor.
[15:07] <soren> huats: Well.. Yes, sort of.
[15:08] <soren> huats: Think of it as the switch all your VM's are plugged into.
[15:08] <huats> soren: ok
[15:29] <seyDoggy_> hey guys, can I ask some painfully newby-ish q's
[15:40] <joe-mac> anybody here with an 8.04 HTTPS box?
[15:41] <joe-mac> i think i'm seeing a regression of a memory leak, and if someone else can independently confirm/deny i can either move the ticket along probably or take a different avenue
[15:46] <giovani> this relates to https?
[15:47] <Pici> Apache? What release of Ubuntu?
[15:49] <pmatulis> he said 8.04
[15:49] <atomic_1> erm, why is iotop NOT in the hardy server repo ?
[15:51] <heath|work> I need to check passwords in my web app against an ubuntu LDAP server. The passwords are returned like {crypt}<some letters>. Is there a salt or something some where? What alg is being used?
[15:51] <Pici> pmatulis: oops :)
[15:52] <giovani> heath|work: man crypt()
[15:52] <giovani> or man 3 crypt rather
[15:53] <gopp> anyone of a command line version of google desktop for ubuntu
[15:53] <joe-mac> giovani: bug report is here ls -s $file | cut -d' ' -f1
[15:53] <pmatulis> atomic_1: obviously no one put in the effort
[15:53] <joe-mac> wrong paste
[15:53] <joe-mac> https://bugs.launchpad.net/ubuntu/+source/apache2/+bug/422138
[15:53] <uvirtbot> Launchpad bug 422138 in apache2 "Slow memory leak, seen on two machines, appears to be dupe of 224945 even after -updates" [Undecided,Incomplete]
[15:54] <joe-mac> it only relates to HTTP, appears to be a regression of #224945, but i just need some confirmation that i am either insane, or right
[15:54] <joe-mac> HTTPS***
[15:55] <soren> atomic_1: Because it didn't exist when Hardy was released.
[15:56] <atomic_1> oh, i see
[15:56] <atomic_1> thanks
[15:56] <atomic_1> guess ill just compile it
[15:58] <heath|work> giovani, ok I get what it is doing, but what sets the 2 char string?
[16:04] <heath|work> ahh nvm. You just try and rebind with the users full dn... d'uh
[16:06] <uvirtbot> New bug: #429749 in eucalyptus (main) "After installation, euca-run-instances fail unless you run euca-describe-groups first" [Undecided,Fix released] https://launchpad.net/bugs/429749
[16:13] <orudie> is there a way i could set up a radio stream ?
[16:13] <orudie> for shoutcast
[16:19] <garymc> yes use shoutcast
[16:21] <garymc> orudie :  Either record some mp3 files and place them on the server and get shoutcast to stream them, or use a third party peice of software called SAM broadcaster or winamp to stream live to the server and allow people to log onto your server ip with winamp or other win media player
[16:46] <heath|work> I'm trying to use LDAP in our web app. I would like to store the allowed modules and other ACL info in the LDAP server. Does anyone have any reading they can suggest to me for such a thing?
[16:46] <teddymills> Anyone have a URL that shows how to install and configure mdadm from a single drive Ubuntu 8042 server edition ?
[16:46] <uvirtbot> Launchpad bug 8042 in nautilus "Text file with execute permission are opened in gedit" [Medium,Fix released] https://launchpad.net/bugs/8042
=== orudie_ is now known as orudie
[16:51] <teddymills> I do not see how this can be done. Existing filesytems are EXT3 and RAID1 filesystems are 'fd' linux raid auto.
[16:55] <JanC> teddymills: what do you mean exactly?
[16:56] <JanC> mdadm is for RAID, and RAID uses more than one drive by default?  ;)
[16:57] <teddymills> If your running an exisitng single drive server Ubuntu 8042. Then want ot make it a RAID1..The partition types are EXT3. RAID1 are 'fd' linux raidauto.
[16:57] <uvirtbot> Launchpad bug 8042 in nautilus "Text file with execute permission are opened in gedit" [Medium,Fix released] https://launchpad.net/bugs/8042
[16:58] <JanC> teddymills: put in the points in the version in the future, so that uvirtbot doesn't think it's a bug #  ;)
[16:58] <teddymills> If you change them to 'fd' you will probably wipe the entire server..If thats the case I may as well reinstall a new server. (not possible in this case since data must be retained)
[16:59] <JanC> so, you want to add an extra disk, and create a RAID with the existing one and the new one?
[16:59] <teddymills> yes
[16:59] <JanC> that's possible, but it might be faster to do a new install  ;)
[17:00] <teddymills> what about using g4L..and making a disk image of the single server..It would be nice if g4L can write to external SATA/USB drives.
[17:01] <JanC> the trick you can use is to use the new disk to create a "broken" RAID1 (a mirror that's lacking 1 disk)
[17:01] <JanC> then copy everything from the old drive to the new one
[17:01] <JanC> and then add the old drive as the second mirror drive
[17:05] <teddymills> is there anyway to install mdadm/raid1 on an existing server, without putting the server back together again from backups?
[17:06] <JanC> that's what I just explained how to do  ;)
[17:06] <zer0her0> anyone run a server on Amazon's EC2?
[17:06] <teddymills> from what Jac was saying, the answer is yes, but the procedure is basically as bad as reinstalling the server from scratch
[17:08] <JanC> teddymills: yeah, if you can make an image on an external disk, doing a new install might be faster...
[17:10] <JanC> teddymills: on the occasion I did it, I didn't have a backup drive large enough around, so...
[17:10] <joe-mac> teddymills: there is no 'easy' or streamlined way to do it
[17:10] <joe-mac> if you have another server, to minimize downtime you can set that up first, do an rsync, then cutover
[17:15] <teddymills> thx guys..knowing what does not work is important than knowing what does work  <--I just made that up. if no one else said it, then you heard it here first :)
[17:16] <ttx> zul: please sync with smoser for ec2-init changes review + sponsoring
[17:17] <smoser> ttx, i'll tag the bug as request for sponsor
[17:17] <JanC> joe-mac: actually, the way I did it you have no more downtime than a normal reboot  ;)
[17:17] <smoser> it is ready to go. and soren agrees to everything.
[17:21] <seyDoggy_> installing ubuntu server onto PPC mac and it's stuck at 83% yaboot install. any ideas? can I recover from this?
[17:22] <JanC> yaboot is the bootloader?
[17:23] <seyDoggy_> JanC: yep
[17:24] <JanC> I suppose you can always (re)install it later?  (no experience with PPC/yaboot)
[17:24] <seyDoggy_> JanC: what can I do at the moment though? kill the whole process?
[17:26] <JanC> maybe first try to find out what's up, but otherwise, not sure there is another option  ;)
[17:28] <seyDoggy_> JanC: so how can I diagnose?
[17:28] <uvirtbot> New bug: #433885 in mysql-dfsg-5.0 (main) "installation of mysql-server and phpmyadmin failed while system under stress" [Undecided,Incomplete] https://launchpad.net/bugs/433885
[17:29] <JanC> I'd first look at dmesg & logfiles, maybe top, ...
[17:31] <seyDoggy_> JanC: from cmd line?
[17:32] <seyDoggy_> JanC: dmesg just dumped out a pile of jebberish (to me anyway)
[17:33] <JanC> seyDoggy_: does it have something about errors at the end?
[17:34] <joe-mac> JanC: and a cutover is < a reboot?
[17:34] <joe-mac> if you have another server, of course
[17:35] <JanC> joe-mac: of course, just meaning my way is not too bad, especially compared to a complete reinstall
[17:35] <JanC> without another server to take over
[17:36] <joe-mac> yea, i actually didn't see that above
=== turbosloth is now known as bnjmn
[17:38] <seyDoggy_> JanC: it's hard for me to say. further up the list is "Oops: kernel access of bad area..."
[17:39] <JanC> seyDoggy_: bad disk area?
[17:41] <seyDoggy_> JanC: just says "...bad area, sig:11 [#1]"
[17:43] <seyDoggy_> i'm just going to kill it and try again.
[17:52] <uvirtbot> New bug: #412664 in eucalyptus (main) "very hard to firewall eucalyptus securely" [Undecided,New] https://launchpad.net/bugs/412664
[17:56] <Deeps> hi guys, i was wondering if there was an easy way to make a screen session that's attachable by any user, not just the one that created it?
[17:56] <smoser> there is.
[17:57] <Deeps> i could use an intermediary account to make the screen, have that account in a specific group, then have a script that can be called to alter the user's pty perms to include the group that the intermediary account is on to have perms to read/write/execute
[17:57] <Deeps> but that seems a bit of a hack
[17:57] <Deeps> smoser: any pointers?
[17:57] <smoser> http://aperiodic.net/screen/multiuser
[17:58] <smoser> there are some issues
[17:58] <Deeps> yep, i see at the bottom
[17:59] <Deeps> i'll give that a go, cheers
[17:59] <smoser> there was another url i saw once that i follwed and had it working for me, but its been long ago
[18:00] <Deeps> i realise this next question is completely outside the remit of this channel, as X is involved, however, i've recently discovered that my new fileserver box is actually quiet enough to sit by the TV and be used as a mediastation of sorts too
[18:01] <Deeps> what i'm in the process of doing is using ssh/screen/mplayer with X and a minimal window manager (wm2) to control the X output
[18:01] <Deeps> hense wanting multiuser screen, so my flatmates can easily ssh in and pause/rewind/close something that i've started, without needing root
[18:02] <Deeps> wondering if anyone had any better suggestions for a lightweight media system- not too inclined to go the whole hog and get kubuntu-desktop/linuxmce going on
[18:02] <Deeps> as this machine's primary role is to be a fileserver, despite being vastly overpowered for the job
[18:26] <Keizer>  http://2.bp.blogspot.com/_rbeqbj-n1Z0/Skeak1qBGyI/AAAAAAAAAeM/3k_ntrDWmOw/s1600-h/ubuntu.png
[18:26] <genii> Keizer: No need for trolling
[18:29] <uvirtbot> New bug: #436447 in eucalyptus (main) "UI freeze exception for Eucalyptus 1.6 console" [Wishlist,Triaged] https://launchpad.net/bugs/436447
[18:31] <seyDoggy_> I am having no luck here. I keep getting to different stages in this ppc install then I have to kill it for one reason or another.
[18:35] <joe-mac> lmao Keizer
[18:36] <Deeps> lol Keizer
[18:36] <seyDoggy_> ok in guided partitioning, I want to use the whole disk. What is LVM for?
[18:38] <joe-mac> LVM is for logical volumes
[18:38] <joe-mac> chances are unless you'll be adding disks to this at a later date, you don't want the administration overhead, since you don't kbnow what it is
[18:38] <seyDoggy_> joe-mac: yeah this is just for LAMP
[18:39] <joe-mac> https://bugs.launchpad.net/ubuntu/+source/apache2/+bug/422138    <--- anybody got an HTTPS 8.04 system they want to test this out on?
[18:39] <uvirtbot> Launchpad bug 422138 in apache2 "Slow memory leak, seen on two machines, appears to be dupe of 224945 even after -updates" [Undecided,Incomplete]
[18:39] <seyDoggy_> basic web serving, no need for LVM me thinks?
[18:39] <joe-mac> seyDoggy_: probably not
[18:40] <seyDoggy_> joe-mac: cool tnx
[18:40] <Deeps> smoser: cheers, got that working now :)
[18:46] <uvirtbot> New bug: #436199 in eucalyptus (main) "display admin URL on boot" [Wishlist,Triaged] https://launchpad.net/bugs/436199
[18:47] <uvirtbot> New bug: #430877 in eucalyptus "describe volumes and describe snapshots times out occasionally" [Medium,Confirmed] https://launchpad.net/bugs/430877
[19:06] <seyDoggy_> so far so good, I might have a web server by the days end. lol
[19:13] <seyDoggy_> "Finishing the installation" nice!
[19:13] <seyDoggy_> "BONG"
[19:13] <seyDoggy_> welcome to yaboot
[19:14] <seyDoggy_> loading
[19:14] <seyDoggy_> ... :S
[19:15] <seyDoggy_> crap pram battery is dead. clock won't set. will this be an issue?
[19:16] <seyDoggy_> ubuntu login: yeah!
[19:17] <seyDoggy_> adam@ubuntu:~$
[19:17] <seyDoggy_> yeah
[19:17] <seyDoggy_> well that was fun... now what
[19:29] <zul> smoser: whats the bug number for the console bug again?
[19:29] <smoser> bug 431103
[19:29] <uvirtbot> Launchpad bug 431103 in linux-ec2 "ssh host key fingerprint no longer available in the console log" [Undecided,New] https://launchpad.net/bugs/431103
[19:29]  * smoser thanks awesomebar
[19:30] <zul> smoser: meric buckets
[19:30] <zul> er...merci buckets :)
[19:32] <uvirtbot> New bug: #429631 in eucalyptus (main) "Invalid S3_URL value in eucarc (dup-of: 429734)" [Low,Confirmed] https://launchpad.net/bugs/429631
[19:33] <aubre> isn't it merci bukkits? :P
[19:34] <axisys> I noticed none of my syslogd-listfiles rotated since Aug 16 .. is there a way I can find out why and how do I rotate them now ?
[19:35] <zul> smoser: http://people.canonical.com/~chucks/0001-UBUNTU-SAUCE-ec2-Default-domU-console-to-tty.patch
[19:36] <uvirtbot> New bug: #429781 in eucalyptus "snapshots created in 1.5 do not get imported to 1.6" [Wishlist,Incomplete] https://launchpad.net/bugs/429781
[19:38] <smoser> thanks zul
[19:38] <axisys> none of my logs have been rotated for more than a month.. here is an snippet http://pastebin.com/d34cc17cf
[19:39] <axisys> anyidea why ? and how do I rotate them now?
[19:41] <axisys> am I hitting a bug?
[19:42] <disown> Hi. I am looking into different solutions of building a elastic computing cluster. I want to acheive something similar to a diskless PXE boot cluster, without having to own the hardware. Ec2 look good, but I wonder if it is "elastic" enough to fix a PXE-like scenario. Is adding servers completely automated, i.e will hostname etc be automatically updated when you spawn a new instance, or do...
[19:42] <disown> ...you need to edit that manually after launching a new instance? Thanks
[19:51] <axisys> how do I find all files in /etc dir that is not owned by root ?
[19:54] <seyDoggy_> ok, so I have a web server now... got DNS, LAMP and ssh installed... any good docs on where I go from here?
[20:05] <genii> seyDoggy_: https://help.ubuntu.com/9.04/serverguide/C/
[20:12] <seyDoggy_> genii: tnx
[20:15] <genii> seyDoggy_: You're welcome
[20:20] <ruben23> hi
[20:20] <ruben23> hi..are there application on linux that i can screen view the monitor of a multiple linux desktop screen on a server like computer
[20:21] <ruben23> what i mean is an ubuntu desktop network
[20:22] <ruben23> i have 25 ubuntu desktop client on my network and want to monitor its screen desktop real time on a single server client
[20:23] <uvirtbot> New bug: #377364 in eucalyptus (main) "package eucalyptus-nc 1.5~bzr266-0ubuntu2 failed to install/upgrade: subprocess post-installation script returned error exit status 1" [High,Fix released] https://launchpad.net/bugs/377364
[20:26] <seyDoggy_> ugh... I just simply want to serve the files I ftp'd? How do I alias the ftp directory to the www directory?
[20:27] <uvirtbot> New bug: #399042 in eucalyptus (main) "package eucalyptus-nc 1.5~bzr266-0ubuntu2 failed to install/upgrade: subprocess post-installation script returned error exit status 1 (dup-of: 377364)" [Undecided,New] https://launchpad.net/bugs/399042
[20:28] <uvirtbot> New bug: #399041 in eucalyptus (main) "package eucalyptus-nc 1.5~bzr266-0ubuntu2 failed to install/upgrade: subprocess post-installation script returned error exit status 1 (dup-of: 377364)" [Undecided,New] https://launchpad.net/bugs/399041
[20:36] <ruben23> hi anyone have idea.?
[20:40] <joe-mac> vnc?
[20:41] <uvirtbot> New bug: #436835 in vm-builder (universe) "[PATCH] fix failure: TypeError: not enough arguments for format string" [Undecided,New] https://launchpad.net/bugs/436835
[20:43] <ruben23> joe-mac:you used vnc..?
[20:44] <CocaCola77> Is there a way to test that my old greybox pc is stable enough to run a set and forget ubuntu server?
[20:44] <joe-mac> ruben23: i used to use ti when i did desktop leenuckz support every now and then when i couldn't fix something via SSH, so kinda rare. i sort of forget how the authentication worked, there was some hackery involved iirc
[20:44] <joe-mac> CocaCola77: what kind of server?
[20:45] <ruben23> joe-mac: probbaly this is just a LAN remote desktop viewing
[20:45] <CocaCola77> joe-mac, ubuntu + samba (working as file server and nt4 like domain controller)
[20:45] <ruben23> but ist multiple-real time view scrren
[20:45] <ruben23> screen
[20:46] <joe-mac> ruben23: i've seen a tool like that for RDP on windows, but never saw anything like that for linux. what exactly do you need tio accomplish?
[20:46] <joe-mac> CocaCola77: only way you can tell is to put a load on it
[20:46] <joe-mac> cpuburn etc
[20:47] <CocaCola77> joe-mac, so just cpu, ram and hdd tests could tell me?
[20:47] <joe-mac> that's all you can really do aside from a pilot of the server
[20:48] <ruben23>  joe-mac: sample- i have 25 users- ubuntu desktop client and there is another client (servers as server) whihc can remote screen on all 25 desktop view its desktop screen or can open it at the same time, maybe just view purpose.
[20:48] <joe-mac> i don't understand what or even how you would view 24 desktops at once
[20:49] <ruben23>  joe-mac: maybe one at a time
[20:49] <ruben23> or 2 desktop at a time
[20:49] <ruben23> is it possible, its for monitoring purposes
[20:51] <joe-mac> then use vnc
[20:51] <joe-mac> only problem is, you'll want to tunnel it for security, requies a little hacking
[20:52] <uvirtbot> New bug: #418409 in eucalyptus (main) "package eucalyptus-cloud 1.6~bzr452-0ubuntu3 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [High,Incomplete] https://launchpad.net/bugs/418409
[20:53] <ruben23> joe-mac: no need for tunnel coz im just on the same network
[20:53] <ruben23> local network
[20:56] <ruben23> joe-mac: thanks for this
[20:57] <ruben23> anyone have used tight vnc...?
[20:58] <ruben23> for ubuntu desktop
[20:58] <seyDoggy_> do you guys use pico to edit config files? is there something else?
[20:58] <Hypnoz> vim
[20:58] <boshhead> seyDoggy_: nano would be the editor you're looking for if you're looking for something like pico
[20:58] <boshhead> seyDoggy_: but I use vim
[20:59] <Hypnoz> ruben23: vnc is kinda flaky on ubuntu-desktop, not the easiest thing to get working
[20:59] <ruben23> Hypnoz:..why..?
[21:00] <Hypnoz> took me a lot of google searching just to kind of get it working right
[21:00] <ruben23> Hypnoz:its just a plain vnc...? or any particular version.
[21:00] <Hypnoz> for me it was partly because the login screen doesn't support vnc, so you have to be logged in. And the primary desktop that shows on your monitor isn't what vnc shows by default
[21:00] <Hypnoz> it creates its own virtual desktop
[21:01] <ruben23> ow
[21:01] <ruben23> problem-im deploying it for a particular requirements
[21:01] <Hypnoz> actually i think the desktop comes with a version of vnc already installed that you can use, somewhere in settings > remote desktop
[21:01] <Hypnoz> if you enable remote desktop, it starts a vnc server
[21:02] <ruben23> ok, but how do i connect to it..?using vnc client or vnc server..?
[21:02] <qman__> seyDoggy_, I also use vim
[21:02] <qman__> surprised no emacs people are here though
[21:03] <Hypnoz> ruben23: check out System > Preferences > Remote Desktop
[21:03] <Hypnoz> that might work well enough for you to use
[21:03] <ruben23> ok ill try it thanks
[21:03] <Hypnoz> and there is a vnc viewer already built into ubuntu desktop, otherwise they are free to download for windows clients
[21:04] <ruben23> actually little confuse between ubuntu / windows desktop on my requirements.
[21:04] <ruben23> local network desktop screen monitoring
[21:05] <seyDoggy_> scuze my ignorance but how does one use vim from the command line?
[21:06] <seyDoggy_> im admin through ssh
[21:06] <qman__> vi
[21:06] <uvirtbot> New bug: #423308 in eucalyptus (multiverse) "ec2-describe-instances sometimes shows the wrong IP" [Medium,Confirmed] https://launchpad.net/bugs/423308
[21:06] <qman__> if you're using hardy or older, you might want to install the more complete version of vim though
[21:06] <qman__> the default version doesn't let you use the arrow keys properly
[21:06] <qman__> while in insert mode
[21:07] <qman__> the newer releases include a better default vim though
[21:07] <qman__> if you've never used vim/vi before, you're likely going to be very confused
[21:08] <seyDoggy_> oh geeze vim crashed the file
[21:08] <seyDoggy_> how do I get out of vim?
[21:09] <joe-mac> hit esc a bunch of times
[21:09] <joe-mac> then hit :q!
[21:09] <qman__> press escape, :wq, enter
[21:09] <qman__> err
[21:09] <qman__> q!
[21:09] <qman__> my bad
[21:09] <qman__> wq is save
[21:10] <seyDoggy_> oops, crash caused a swap file to be created
[21:10] <seyDoggy_> better get the full version of vim
[21:11] <qman__> in hardy and older the package is  vim-full
[21:12] <seyDoggy_> oh, I just checked out "vim". what's "hardy"? I installed lts 8.1... is that hardy?
[21:12] <genii> 8.04
[21:13] <qman__> hardy is 8.04 lts
[21:13] <qman__> I don't know about the 8.10 package
[21:13] <genii> 8.10 is Intrepid
[21:13] <genii> (and not long-term-support)
[21:13] <seyDoggy_> i'll check out vim-full and see what happens
[21:13] <qman__> I do know that jaunty (9.04) has a better vim by default
[21:14] <seyDoggy_> still vi from cmd line for full vim?
[21:14] <qman__> yes
[21:14] <qman__> installing vim-full adds more features to vim
[21:15] <seyDoggy_> tnx
[21:15] <qman__> one of them is better arrow keys support
[21:15] <qman__> among many others
[21:15] <seyDoggy_> can you tell I am not a linux guy ;)
[21:15] <qman__> you should definitely look up a crash course on vim, though
[21:16] <qman__> the manual is massive and has way too much information just to learn how to use it
[21:16] <seyDoggy_> eek maybe I'll stick with pico/nano
[21:17] <qman__> it takes effort but it does have useful features
[21:17] <qman__> so if you plan on spending a lot of time editing files, it's worth looking into
[21:18] <qman__> one of the features I use most is regex string replacement
[21:20] <seyDoggy_> i come from the textmate camp. don't know if I'm ready to wrap my head around a whole new way of doing things.
[21:32] <uvirtbot> New bug: #411692 in eucalyptus (main) "stopping eucalyptus may lose the database" [Critical,Incomplete] https://launchpad.net/bugs/411692
[21:37] <seyDoggy_> i am having a heck of a time getting apache to serve anything other the /var/www/
[21:38] <kirkland> zul: ping
[21:38] <kirkland> zul: https://bugs.edge.launchpad.net/ubuntu/+source/eucalyptus/+bug/436876
[21:38] <uvirtbot> Launchpad bug 436876 in eucalyptus "Unzip should be considered" [Undecided,New]
[21:38] <kirkland> zul: what system do you want unzip on?
[21:38] <zul> kirkland: the node
[21:38] <kirkland> zul: the node?
[21:38] <kirkland> zul: hmm, i was thinking the cc
[21:38] <kirkland> zul: why the node?
[21:39] <kirkland> zul: to unzip your credentials?
[21:39] <zul> isnt that were you run the eucatools
[21:39] <zul> yes to unzip your credentials
[21:40] <zul> since zip is already installed by default iirc
[21:40] <kirkland> zul: you really shouldn't need to do anything on the -nc
[21:40] <kirkland> zul: ever, really
[21:40] <zul> ah ok
[21:40] <kirkland> zul: -cc, yeah, perhaps
[21:40] <kirkland> zul: cool
[21:42] <uvirtbot> New bug: #409760 in eucalyptus (main) "fresh install of eucalyptus-cloud on karmic fails" [High,Fix released] https://launchpad.net/bugs/409760
[21:51] <uvirtbot> New bug: #436876 in eucalyptus (main) "Unzip should be considered" [Undecided,New] https://launchpad.net/bugs/436876
[21:57] <uvirtbot> New bug: #385660 in eucalyptus "In MANGED* modes, cannot access public IPs of VMs from the machine on which the CC is running" [Low,Fix committed] https://launchpad.net/bugs/385660
[21:58] <joe-mac> jeez eucalyptus is getting a lot of qa today eh?
[21:58] <genii> I was just thinking the same
[21:58] <joe-mac> a colleague brought this stuff up to me the other day
[21:58] <joe-mac> never heard of it before maybe two days ago...
[22:16] <uvirtbot> New bug: #436041 in samba (main) "Access Denied but TSC confirms it should not be." [Undecided,Incomplete] https://launchpad.net/bugs/436041
=== vraa__ is now known as vraa
[22:27] <vraa> hello everyone, i finally am at my workstation, i was having an issue with my computer, it has 4 on board LAN ports (but only 1 is plugged in) during boot time it hangs at the "configuring network interfaces" portion, i press ctrl_alt+del and it skips it and boots up and things work fine though, how can i resolve so i can boot smoothly without user intervention
[22:28] <vraa> or perhaps what direction should i head to, it runs ubuntu but i also have openssh server isntalled so i can control it using putty from my laptop
[22:32] <joe-mac> vraa: what is in your /etc/network/interfaces file?
[22:33] <vraa> http://pastebin.com/d633d03e
[22:34] <uvirtbot> New bug: #364806 in eucalyptus "When maximum volume size is exceeded, the system should report a more helpful error" [Wishlist,Confirmed] https://launchpad.net/bugs/364806
[22:34] <uvirtbot> New bug: #364902 in eucalyptus (main) "eucalyptus-cloud should depend on mail-transport-agent" [Medium,Fix committed] https://launchpad.net/bugs/364902
[22:34] <uvirtbot> New bug: #365156 in eucalyptus (main) "-cloud and -cc init scripts don't set JAVA_HOME" [Medium,Invalid] https://launchpad.net/bugs/365156
[22:35] <vraa> on the ubuntu desktop, if i right click network manager icon and go to edit connections, Auto eth2 doesnt show up, but auto eth3, auto eth0, and auto eth1 do show up.
[22:36] <vraa> when i got o System -> Administration -> Network Tools -> i can see eth2 listed under the network devices section and it works fine (pulls IP using dhcp, it's my web server it works great, it's just doesn't restart / boot w/o user intervention)
[22:37] <uvirtbot> New bug: #357999 in eucalyptus "configureNetwork is slow when instance IPs are not in DNS" [Low,Fix committed] https://launchpad.net/bugs/357999
[22:41] <uvirtbot> New bug: #436896 in image-store-proxy (main) "Signature tests are disabled" [High,Confirmed] https://launchpad.net/bugs/436896
[22:46] <uvirtbot> New bug: #338866 in eucalyptus (main) "eucalyptus-cc and nc init scripts do not clear /dev/shm before 'start' or after 'stop' " [Undecided,Fix released] https://launchpad.net/bugs/338866
=== ktk_ is now known as KurtKraut
[23:06] <bjaspan> We are using the official Ubuntu 8.04 EC2 image.  We've noticed that the 32-bit image seems to crash and reboot early in the machine's life a fair percentage of the time (perhaps 10-20%).  We have not seen this with 64-bit images.
[23:06] <bjaspan> Is it possible to get a crash dump of 8.04 on ec2? How else can we debug the cause?
[23:11] <erichammond> bjaspan: The official Ubuntu 8.04 EC2 image is not recommended for use.
[23:12] <bjaspan> Oh?
[23:12] <erichammond> bjaspan: It has a number of serious bugs filed against it and the fixes have not been released as a new image.
[23:13] <bjaspan> I know about the "no ssh access" bug (it sure is annoying).
[23:14] <erichammond> bjaspan: Are you saying that you've been using that image for a while and the reboot problem just started?
[23:14] <bjaspan> We've been using m1.larges for a while.  We just started testing m1.smalls, and the crash/reboot problem started immediately.
[23:15] <erichammond> I see.  As far as I know that's a problem which others have not reported.
[23:16] <erichammond> Are you running setup code when the instance starts?
[23:17] <bjaspan> Yes, lots of it.
[23:17] <erichammond> Have you been able to reproduce the reboot problem just running the image without any of your setup code running?
[23:17] <bjaspan> Well, we're running puppet, which installs a bunch of packages, files, cron jobs, etc.
[23:18] <bjaspan> We haven't specifically tried.  I'm perfectly prepared to believe that something we're running is triggering a crash... but that doesn't help much unless I can figure out what it is. :-)
[23:18] <bjaspan> We're not tweaking the kernel or anything.
[23:18] <erichammond> My assumption would be that something you're doing is interacting with a problem in the image/kernel and causing the reboot, but it sounds like it would be difficult for somebody else to reproduce unless you can reduce it to a simple test case.
[23:19] <erichammond> The fact that it doesn't happen all the time will make it even more difficult for you to isolate, especially if it relates to startup timing :-\
[23:20] <bjaspan> No kidding.
[23:20] <bjaspan> That's why I asked about getting a crash dump.
[23:21] <erichammond> Would it be possible to send the puppet output to a log file to help track down where the reboot happens?
[23:22] <bjaspan> BTW, what is your role w.r.t. Ubuntu on EC2? I see your name associated with Alestic, but don't know what that is.
[23:22] <bjaspan> Sure, we can get the puppet output.  The machine reboots, but when it comes back the root fs is still intact.
[23:22] <bjaspan> Unfortunately puppet only tells you what it just finished doing, not what it is currently doing.  I was thinking of hacking it to log what it is about to execute.
[23:23] <bjaspan> e.g. the last time I looked, the final thing puppet did before crashing was install emacs-nox, which seems an unlikely culprit.
[23:24] <bjaspan> Hmmm.  Puppet has a debug mode we do not have enabled; that might help.
[23:26] <erichammond> I'm just a guy who's been using Ubuntu on EC2 for a long time and decided to publish the images and build script I was using and they sort of became popular.  Now, Canonical has noticed EC2, so I'm working with them to migrate stuff over to the official releases.
[23:27] <erichammond> I assume you've checked syslog for clues about the crash?
[23:28] <erichammond> Isn't a reboot sort of a rare failure mode?  I would have thought that a crash/terminate would be more likely.
[23:28] <bjaspan> Nothing helpful in syslog I could see.  Just puppet happily chugging away on its initial run, then "syslogd restart".  last sometimes reports "crash".
[23:28] <bjaspan> e.g.
[23:28] <bjaspan> root     pts/0        ip67-154-105-210 Tue Sep 22 19:34 - crash  (00:04)
[23:28] <bjaspan> reboot   system boot  2.6.24-6-xen     Tue Sep 22 19:33 - 22:28 (3+02:55)
[23:28] <bjaspan> reboot   system boot  2.6.24-6-xen     Tue Sep 22 19:30 - 19:32  (00:02)
[23:28] <bjaspan> wtmp begins Tue Sep 22 19:30:17 2009
[23:32] <erichammond> bjaspan: It might be worth opening a bug, though I'm not sure how much progress can be made without a reproducible test case.
[23:32] <erichammond> https://bugs.launchpad.net/ubuntu/+filebug?no-redirect&field.tags=ec2-images
[23:33] <erichammond> Include the AMI id and as much info as you can.
[23:33] <bjaspan> will do, thanks
[23:33] <erichammond> You might want to click "this bug affects me too" on https://bugs.launchpad.net/ubuntu/+bug/429120
[23:33] <uvirtbot> Launchpad bug 429120 in ubuntu "ec2: Build new Ubuntu 8.04 Hardy AMIs" [Medium,New]
[23:34] <erichammond> Now I'll get back to work and stop scrolling away your original question so somebody else can pitch in: "Is it possible to get a crash dump of 8.04 on ec2?"
[23:36] <bjaspan> Thanks!  BTW, in case this rings any bells: We're mounting a FUSE filesystem with glusterfs.
[23:41] <zul> hey jjohansen
[23:41] <jjohansen> hi zul
[23:48] <zul> mmmmmm....ice cream with sprinkles
[23:50] <Hypnoz> can someone help me change where dhcp logs. its flooding my /var/log/messages
[23:51] <Hypnoz> every BMC adapter on every server in my dc is requesting a dhcp address cause thats how its setup by default in the bios, even though we didn't wire up the BMC controllers
[23:52] <Hypnoz> From dhcpd.conf
[23:52] <Hypnoz> # Use this to send dhcp log messages to a different log file (you also
[23:52] <Hypnoz> # have to hack syslog.conf to complete the redirection).
[23:52] <Hypnoz> log-facility local7;
[23:53] <Hypnoz> anyone know what I do to syslog.conf to complete this?
[23:56] <uvirtbot> New bug: #436932 in eucalyptus (main) "euca_conf should assume --local-sync if the given IP is local" [Medium,Triaged] https://launchpad.net/bugs/436932