[02:43] <rockzman> can someone help me to install ubuntu 6.10 server?
[03:06] <mgalvin> rockzman: what do you need help with?
[03:19] <Innatech> asking here because #bind has been silent: can anyone point me towards good docs on split DNS (internal/external) in terms of how to pattern the hostnames/domains and any issues that might arise therefrom in Kerberos, LDAP & Samba? 
[03:34] <mralphabet> I run them as two seperate entities
[03:34] <mralphabet> granted, I don't run kerb / ldap / samba access on both sides of the firewall
[03:35] <Innatech> Yes, that's more or less what I intend. Nothing funky cropped up doing that? 
[03:35] <mralphabet> nope
[03:35] <Innatech> cool, thanks. 
[03:35] <mralphabet> resources that _do_ have holes poked in the firewall for them have two names
[03:36] <Innatech> hrmm. OK, that should be manageable without too big a headache. 
[03:36] <mralphabet> IE internally it resolves both blah.somedomain.com *and* blah.somedomain.local
[03:37] <mralphabet> when dealing with end users who use the resource internally and externally, I refer to everything as blah.somedomain.com
[03:37] <mralphabet> external obviously only has the listing blah.somedomain.com
[03:38] <mralphabet> any services that need inter-server communication uses blah.somedomain.local
[03:39] <mralphabet> if that makes any sense
[03:39] <mralphabet> not saying it's right, but it works for me /shrug
[03:40] <Innatech> I think I get it. That's about what I envisioned. 
[03:41] <Innatech> I'm just not sure why one would want to do it that way, rather than using a subdomain. But I'm probably missing some no-duh big picture issue there. 
[03:42] <mralphabet> I don't think it matters one way or the other, I think it is just preference
[03:43] <Innatech> ah, OK. I felt like something obvious was escaping me. 
[12:01] <davekempe> free tech support from me for the next half hour. ask me anything to do with ubuntu on the server! :)
[12:01] <Kamping_Kaiser> why doesnt my work use ubuntu on its server? :(
[12:03] <davekempe> hehehe. just get them to call me
[12:03] <davekempe> after I get back from holidays
[12:03] <davekempe>  :)
[12:03] <davekempe> no seriously
[12:04] <Kamping_Kaiser> hehe
[12:10] <fabbione> davekempe: i have an HSG80 in multibus failover exporting 2 DX towards 2 switches and there to 1 one machine with 4 FC-HBA lanes.. can i use dm-multipath+dm-roundrobin to load balance the traffic and handle failover?
[12:11] <fabbione> ops
[12:11] <davekempe> fabbione, hrmm
[12:12] <davekempe> on dapper?
[12:12] <fabbione> dapper or feisty
[12:12] <davekempe> multipath is a bit undermaintained for my liking
[12:12] <davekempe> what hbas?
[12:12] <fabbione> the controllers are 2 Emulex LP9000 and 2 qlogic 2400
[12:13] <davekempe> nice
[12:13] <davekempe> dunno about their support in dapper. I would get a more recent kernel myself, built by hand. 
[12:13] <fabbione> of course all cross connected to avoid failures by vendor
[12:13] <davekempe> thats what i did for a similiar situation recently
[12:14] <davekempe> ok
[12:14] <fabbione> you didn't answer my question tho
[12:14] <davekempe> yeah im still thinking
[12:14] <davekempe> honestly, I would just try it and see
[12:15] <davekempe> can i use questions require a certain degree of experimentation in my exp
[12:15] <fabbione> davekempe: i will tell you the answer.. i was just teasing you...
[12:15] <davekempe> yeah i thought so
[12:15] <fabbione> i know the answers to all these questions
[12:15] <davekempe> hey have you seen priomark?
[12:15] <fabbione> but free support was an offer i couldn't resist :)
[12:15] <fabbione> priomark?
[12:16] <davekempe> sorry I am not as familiar with your hardware as I would like to be :)
[12:16] <davekempe> http://www.ipacs-benchmark.org/index.php?s=download&unterseite=priomark
[12:16] <davekempe> I can send you the paper if you want to read more - I bought the paper for it yesterday
[12:17] <fabbione> no i didn't see it
[12:17] <davekempe> I am looking for an effective way to test my AoE SAN across different archs/distros
[12:17] <fabbione> AoE? brrrrrrrrr....
[12:17] <fabbione> it depends what kind of tests you want to perform
[12:17] <fabbione> stability? performance? redundancy?
[12:17] <fabbione> failover?
[12:18] <davekempe> bonnie++ seems to give me strange results
[12:18] <fabbione> i use dbench
[12:18] <davekempe> more like what the hit in performance
[12:18] <davekempe> over local disk etc
[12:18] <davekempe> ok I will check it out
[12:18] <fabbione> it's somewhere in universe
[12:18] <davekempe> going back onsite tomorrow to play with it
[12:18] <fabbione> well clearly the first hit you get is the network traffic
[12:18] <fabbione> you should really separate the bits in the setup
[12:19] <fabbione> first benchmark the network of the server
[12:19] <fabbione> because you might have a 10Gbit Ethernet
[12:19] <davekempe> yeah i ahve been testing them all seperatly as i build it
[12:19] <fabbione> but a broken driver that push 10kbit
[12:19] <davekempe> yeah i noticed a real difference on the areca raid card between dapper and fiesty
[12:19] <davekempe> 20% performance increase
[12:20] <fabbione> i would also perform tests from one machine only towards the SAN
[12:20] <fabbione> and see if you can saturate
[12:20] <fabbione> if you can't it's pointless to go with another node on it
[12:20] <fabbione> and if you plug more than one node.. then you need to decide how you want to test the access to AoE
[12:21] <fabbione> 2 different partitions from SAN to 2 machines?
[12:21] <fabbione> or one partition using a clustered FS?
[12:21] <fabbione> it's a pain to track all the options
[12:21] <davekempe> no clustered fs
[12:21] <davekempe> yeah its a minefield
[12:21] <davekempe> aiming for xen dom0's booting the domU off their own raided aoe slices
[12:22] <davekempe> btw - you have any idea how i can tell if I have jumbo frames enabled?
[12:22] <davekempe> not that i have googled that yet
[12:22] <fabbione> no i don't know.. i would have to goggle it too
[12:23] <fabbione> but it also depends from the hw
[12:23] <fabbione> some cards have limited MTU in hw
[12:23] <davekempe> ahh
[12:23] <davekempe> ok cool
[12:23] <fabbione> to offload ipv4 checksum calculaton down to the chipset
[12:23] <fabbione> that's something i am sure about :)
[12:23] <davekempe> thanks for the free tech support :)
[12:25] <fabbione> why your name keeps ringing a bell in my head but i can't associate to something specific?
[12:26] <davekempe> appamour?
[12:26] <fabbione> no ablo francese :)
[12:28] <fabbione> davekempe: i might be just on crack
[12:28] <ivoks> je ne parle pa fracise :)
[12:28] <ivoks> pas
 appamour?
[12:28] <ivoks> or whatever :)
[12:28] <ivoks> has anyone had anything to do with chillispot?
[12:28] <davekempe> i was pushing various canonical peeps about it. now its getting into fiesty i am happy. big gratz all round
[12:30] <fabbione> food time
[12:30] <fabbione> bbl
[12:31] <davekempe> ivoks, thats a wireless captive portal right?
[12:32] <ivoks> yes...
[12:32] <ivoks> i have problems with it; it wouldn't connect to my radius server
[12:35] <davekempe> i haven't used it before, but I still may be able to help you
[12:35] <davekempe> what have to you checked - does it have logs? tethereal/tshark? strace?
[12:35] <[miles] > morning guys
[12:36] <[miles] > I really do think we should try to get the ubuntu team to fix up the saslauthd and postfix problem
[12:36] <[miles] > as its bloddy annoying
[12:37] <ivoks> davekempe: yes it has, but nothing meaningful in them
[12:37] <ivoks> davekempe: radius doesn't record connection being made at all
[12:37] <davekempe> tried tshark? packet sniffers never lie
[12:38] <[miles] > having the socket outside of the postfix jail is a nag
[12:38] <davekempe> miles - yeah its a bit of trap for new players
[12:38] <[miles] > davekempe: I submited a patch
[12:38] <[miles] > davekempe: and opened a bug
[12:38] <davekempe> overall the setup is still way easier than other platforms.... :)
[12:39] <[miles] > davekempe: yeah, but messing around with the init script etc
[12:39] <[miles] > davekempe: is still a fuckin pain
[12:39] <davekempe> i agree
[12:39] <davekempe> got any response on your bug?
[12:46] <[miles] > yeah, I submitted it a few months ago, but it got checked the other day I think
[12:47] <[miles] > https://launchpad.net/bugs/79371
[12:47] <davekempe> damn searching for bugs on launchpad is annoying
[12:49] <davekempe> what about a debian bug? upstream might have more weight on this package
[12:56] <[miles] > no idea
[12:56] <[miles] > I don't touch debian
[12:57] <[miles] > not involved with it in any way shape or form
[12:57] <[miles] > davekempe: your british I guess by your name?
[12:57] <davekempe> lodging the same but against the debian package might be tackling the problem from two angles :) or maybe just bad form :(
[12:57] <davekempe> nope im Aussie
[12:57] <[miles] > ah jeje
[12:57] <[miles] > nice!
[12:57] <[miles] > better
[12:58] <[miles] > the brits are a bunch of twats
[12:58] <davekempe> ill take your word for it :)
[12:58] <[miles] > but live and work in Barcelona 5 years
[12:59] <davekempe> wtfs a scouser?
[12:59] <[miles] > someone from Liverpool
[01:00] <davekempe> i see :)
[04:41] <rambo3> what are the basic meta-packeges  for basuc system in server ?
[04:47] <Nafallo> ubuntu-{minimal,standard}
[04:49] <rambo3> ok thanks 
[04:58] <JakeX> hey can someone help me with an issue regarding samba? I just need to upgrade to latest version on a breezy dist.. http://www.ubuntu.com/support/communitysupport
[04:59] <JakeX> i seem to be limited to samba 3.0.14 :(
[04:59] <JakeX> oops.. wrong given a sec ago, heres the right one: http://ubuntuforums.org/showthread.php?t=401200
[05:02] <shawarma> 3.0.14 is the newest available on Breezy.
[05:02] <shawarma> And if I remember correctly, Breezy will be unsupported in 10 days.
[05:02] <JakeX> hmm well i have a production server..
[05:03] <shawarma> JakeX: https://lists.ubuntu.com/archives/ubuntu-announce/2007-March/000099.html
[05:03] <JakeX> and due to the MacOSX tiger upgrade I'm forced into upgrading
[05:04] <shawarma> Which version do you need?
[05:04] <JakeX> hmm so do a dapper upgrade from breezy -> dapper -> edgy
[05:05] <shawarma> Do you really need Edgy?
[05:05] <JakeX> i found that 3.0.14 causes problems with macs.. and other users mentioned 3.0.20 as having a fix
[05:05] <shawarma> Dapper is supported for 5 years.
[05:05] <JakeX> no.. i don't really need an upgrade.. but I need samba fixed :|
[05:05] <JakeX> well i can do dapper i guess..
[05:05] <JakeX> but my chief concern is .. a broken os.. due to upgrade.. i've read some people having issues..
[05:05] <JakeX> and since its a server install i didn't wanna play with it :P
[05:06] <JakeX> but it looks like thats my best option.. plus support for 5 years
[05:06] <shawarma> JakeX: Precisely.
[05:07] <JakeX> process for the upgrade would be simply -> swap in dapper source.list and apt-get update > dist-upgrade ?
[05:07] <shawarma> You might be able to backport Samba to your Breezy install, but as I mentioned before, it's unsupported as of the 13th .
[05:07] <JakeX> ya
[05:07] <JakeX> wasn't aware of that..
[05:08] <Nafallo> JakeX: you're not concerned that breezy reached EOL? :-)
[05:08] <JakeX> not really.. it is an internal samba server..
[05:09] <JakeX> iptables + samba + mysql & tomcat for internal app processing..
[05:09] <JakeX> in reality.. i couldn't care less except for samba problem with mac osx :(
[05:51] <chandu_> hi
[05:51] <chandu_> Is there any utility to mount the gzip compressed image without gunziping it 
[05:59] <mralphabet> hi chandu, glad you hung around for an answer!
[06:14] <fdoving> total of 4 min. patience!. :)
[07:20] <mralphabet> so . .. how would one revert from feisty to edgy
[07:32] <mralphabet> ivoks: you wouldn't happen to know how to downgrade to edgy from feisty would you?
[07:32] <ivoks> sure i would :)
[07:33] <mralphabet> Oh? Please, share ;)
[07:33] <ivoks> but, it's not that flawless like upgrade
[07:33] <mralphabet> The feisty + vmware server issues are driving me crazy
[07:33] <ivoks> in /etc/apt/preferences
[07:33] <ivoks> you have to add:
[07:33] <ivoks> Package: *
[07:34] <ivoks> Pin: release a=edgy
[07:34] <ivoks> Pin-Priority: 1001
[07:34] <ivoks> and you should have edgy repos in sources.list
[07:34] <ivoks> simple apt-get  dist-upgrade should start your longest night in your life :)
[07:35] <mralphabet> hmm, it seems I do not have /etc/apt/preferences
[07:36] <ivoks> of course you don't
[07:36] <mralphabet> make a new file?
[07:37] <ivoks> yes
[07:39] <mralphabet> deb http://us.archive.ubuntu.com/ubuntu/ edgy main restricted universe multiverse
[07:39] <mralphabet> deb http://us.archive.ubuntu.com/ubuntu/ edgy main restricted universe multiversdeb-src http://us.archive.ubuntu.com/ubuntu/ edgy main restricted universe multiverse
[07:39] <mralphabet> deb http://us.archive.ubuntu.com/ubuntu/ edgy-proposed main restricted universe multiverse
[07:39] <mralphabet> look right?
[07:40] <ivoks> this is support for server
[07:40] <ivoks> not 'how do ubuntu repos look like
[07:40] <ivoks> :)
[07:40] <mralphabet> heh
[07:40] <mralphabet> good point
[07:40] <ivoks> we have to have a clear separation; we don't want this to become support for everything
[07:41] <mralphabet> I understand, no worries
[07:41] <mralphabet> 0 upgraded, 13 newly installed, 928 downgraded, 86 to remove and 0 not upgraded.
[07:42] <mralphabet> we'll just let that run for a while
[07:43] <mralphabet> ivoks: thanks for the help
[07:43] <ivoks> good luck :)
[08:36] <mralphabet> If at all in doubt, answer Yes. If you know exactly what you are doing, and are prepared                                                                                                                                       
[08:36] <mralphabet>  to hose your system, then answer No
[08:36] <mralphabet> 
[08:37] <mralphabet> certainly an eyeopener
[08:53] <jronnblom> mralphabet: what problems do you have with feisty and vmware server? Im running feisty server with vmware server 1.02 and I haven't run into any problems yet.
[08:57] <mralphabet> jronnblom: for whatever reason, I can not get the host and guests machines to do anything more then ping
[08:58] <mralphabet> ubuntu host with xp and vista guests, guests can not browse network (host is network browsing master) and can not access host samba shares and can not ssh to host
[08:59] <mralphabet> guests can see other machines on the network, other machines can see guests
[08:59] <jronnblom> ah, I remember your problem with ping
[09:00] <jronnblom> what h/w are you on?
[09:00] <mralphabet> it's a white box, asus board w/2.8ghz intel p4
[09:01] <mralphabet> edgy + vmware server 1.0.1 worked
[09:01] <jronnblom> and I assume that you tried 1.02 with feisty? 
[09:01] <theacolyte> yeah ive ran feisty in vmware server 1.02 myself just fine
[09:02] <mralphabet> jronnblom: correct
[09:03] <mralphabet> theacolyte: "feisty in vmware"? feisty as guest?
[09:04] <jronnblom> have you tried with etherreal/tcpdump and see whats on the network when you try to ping the host or guest?
[09:04] <jronnblom> what NIC is on the asus board?
[09:05] <theacolyte> feisty IN vmware server, actually, sorry just reread it all
[09:05] <theacolyte> not as host
[09:06] <mralphabet> jronnblom: RealTek RTL8139
[09:07] <daq4th> fabbione: ?
[09:08] <fabbione> daq4th: sorry.. it was the wrong nick/tab completion
[09:09] <daq4th> ;-)
[09:09] <mralphabet> jronnblom: when I ssh from guest to host I get a timeout error in auth.log
[09:10] <jronnblom> mralphabet: I have only Intel or Broadcom in my servers and they don't seem to have a problem with feisty and vmware-server
[09:11] <jronnblom> I would try and replace the realtek card if possible
[09:11] <theacolyte> broadcom makes some really good chips these days
[09:11] <theacolyte> not too hot on realtek either
[09:13] <mralphabet> I believe I have a 3c905b or c laying around
[09:15] <theacolyte> ! now that's a good card. I still use my DEC Tulip card..
[09:15] <jronnblom> hmm, it used to be a good card but its very old nowdays.
[09:42] <J_P> hi all
[09:56] <pursuantirc> first time ir
[09:56] <pursuantirc> irc
[09:56] <pursuantirc> question on linux
[09:56] <pursuantirc> ubuntu rocks, by the way, and have used the desktop software.  I am interested in a server.
[09:56] <pursuantirc> are there gui tools for the server?
[09:58] <theacolyte> no
[09:58] <theacolyte> not unless you installed them
[10:09] <pursuantirc> thanks
[11:47] <Aw0L> I have a basic dns server setup with dnsmasq - is there a way to make what I have in /etc/hosts to take precedence over the real IP of a site?
[11:47] <Aw0L> like, if I want to make google.com point to a local IP for instance
[11:52] <shawarma> dnsmasq already serves your /etc/hosts via dns.
[11:53] <Aw0L> shawarma, true, but if I add an entry for a domain that already exists, it points to that domain instead of what's in my /etc/hosts
[11:53] <shawarma> Aw0L: That sounds broken. Why should it try to resolve something it finds in its hosts? Odd.
[11:53] <Aw0L> that's what I'm curious about
[11:53] <shawarma> Aw0L: Have you tried putting something like "www.google.com." in the hosts file? Note the final dot.
[11:54] <Aw0L> what does the final dot do?
[11:54] <shawarma> It's kind of like the leading / of a path.
[11:54] <Aw0L> yeah, but why is it necessary?
[11:54] <Aw0L> wait
[11:54] <Aw0L> maybe I should clarify
[11:54] <shawarma> Well, it shouldn't be, but it might help.
[11:55] <Aw0L> on the dns server itself, if I add an entry for google in my /etc/hosts file and type "ping google.com" it pings google
[11:55] <shawarma> Since you tell it that "www.google.com" has this address and not "www.google.com.your.own.domain".
[11:55] <Aw0L> on another computer that uses my dns server as it's dns server - it just goes to google instead
[11:56] <shawarma> Aw0L: eh?
[11:56] <Innatech> anyone have any ideas about why SSH port forwarding suddenly stopped working for connections to local ports on the target host? As in, I SSH to my office LAN, port 22 is forwarded to a machine running the SSH server, and login is normal. Ports tunneled from my home machine to other hosts on the office LAN work normally, but connections tunneled to the machine running the SSH server die. 
[11:56] <shawarma> Aw0L: Your DNS server pings google even thought it's overridden in the hosts file?
[11:57] <Aw0L> not quite
[11:57] <shawarma> Innatech: Define "die".
[11:57] <Aw0L> from the dns server, if I ping google, it returns what's in my /etc/hosts
[11:57] <shawarma> Aw0L: Ah. That's not waht you said. :-)
[11:57] <shawarma> Aw0L: "and type "ping google.com" it pings google"..
[11:57] <Aw0L> if I ping google from another box that has my dns server in /etc/resolv.conf, it pings google's real IP
[11:57] <shawarma> Aw0L: Right, ok.
[11:58] <shawarma> Aw0L: Could you try adding the final dot and SIGHUP'ing dnsmasq?
[11:58] <Innatech> shawarma: I've only monkeyed around with HTTP traffic thus far, but either a generic "server not found" or a  "connection was reset" error page. 
[11:59] <shawarma> Innatech: "Server not found". Which browseR?
[11:59] <Innatech> Shawarma: Firefox. 
[11:59] <Innatech> IE gave a similar error, I forget the specific language. 
[12:00] <shawarma> Innatech: Ok. I just seem to remember something about IE saying that whenever anything at all went wrong. Very confusing.
[12:00] <Innatech> Yes, IE does do that. Which sucks. 
[12:00] <shawarma> Very much so.
[12:00] <Innatech> Anyway, its a client side not found. Not a 404 from the server. 
[12:01] <shawarma> Especially when you're dealing with people who actually know a little bit about what they're talking about, but still not quite, and they tell you that "requests to blablabla gives me a 404" and after hours of debugging it turns out that they're acutally seeing that no good generic error which doesn't say a thing about 404. 
[12:01] <shawarma> Not that I'm bitter or anything..
[12:02] <Innatech> Now I'm at the office and can verify that the httpd is couldn't reach is indeed running. (whew). So now I want to get the SSH tunnel straightened out so I don't have similar heart attacks (thinking my webapp is down when I try to use it remotely.) 
[12:02] <shawarma> Innatech: Well, I see no reason why this should start failing spontaneously. Try with telnet or netcat or something.
[12:03] <Innatech> yeah....telnet failed silently, and I was on XP hosts @ school when it happened, so no netcat there. 
[12:03] <shawarma> Innatech: And you're sure about your command line being right?
[12:03] <shawarma> I have to ask.. :-)
[12:03] <shawarma> Aw0L: Any luck?
[12:04] <Innatech> I was using PuTTy, actually. So unless there
[12:04] <Innatech> 's a bug in the latest build...
[12:04] <Aw0L> shawarma, no
[12:04] <shawarma> Aw0L: The final dot didn't help?
[12:04] <Aw0L> basically, it get it to do what I want I'd have to setup a zone for that domain, which I don't want to do 
[12:04] <Aw0L> and no, it didn't :(
[12:04] <shawarma> Aw0L: I think I have a dnsmasq running on my router. I can try a few things. hang on.
[12:05] <Aw0L> ok, thanks
[12:05] <Innatech> hrm. I'm going to try it from another company's LAN, down the hall. (Oh, the suspsense!) 
[12:06] <shawarma> Aw0L: Hmmm. it works just fine here.
[12:06] <Aw0L> really?
[12:06] <Aw0L> are you pinging from the router though?
[12:06] <Aw0L> or from another machine?
[12:06] <shawarma> Aw0L: No.
[12:06] <shawarma> Another machine.
[12:06] <shawarma> Although..
[12:07] <shawarma> Ah, I think I know what your problem is.
[12:07] <shawarma> $ host www.google.com
[12:07] <shawarma> www.google.com is an alias for www.l.google.com.
[12:07] <shawarma> www.l.google.com has address 209.85.129.99
[12:07] <shawarma> www.l.google.com has address 209.85.129.104
[12:07] <shawarma> www.l.google.com has address 209.85.129.147