[01:47] <eqx311> hi
[01:48] <mathiaz> smoser: hi!
[01:48] <mathiaz> smoser: I'm going through the test case "EC2 Single instance"
[01:48] <mathiaz> smoser: at http://testcases.qa.ubuntu.com/System/CloudImages
[01:49] <mathiaz> smoser: what's that actual things I need to do for point 7.?
[01:49] <mathiaz> smoser: "Pass a --user-data script to the new instance" ??
[01:51] <KillMeNow> no idea mathiaz
[01:53] <erichammond> mathiaz: I added that --user-data note so it wouldn't get lost.  I'm about to head to an interview dinner, but I'll be online in a few hours.  If it's urgent you can also call my cell (number sent separately).
[01:53] <mathiaz> erichammond: ok - not urgent.
[01:54] <erichammond> mathiaz: I do think this is important to test because I believe the last public AMI might have a problem with user-data that I was going to test later.
[01:54] <mathiaz> erichammond: just wondering about the actual step required. I'll give it a try
[01:54] <mathiaz> erichammond: agreed.
[01:54] <mathiaz> erichammond: I'll update the test case if I can find something
[01:55] <erichammond> For now, any simple --user-data-file will do as long as it starts with #!/bin/bash and runs a few commands.  Then look in /var/log/syslog and search for "user-data" (I think) to see if it got run.
[01:55] <erichammond> mathiaz: The problem I was having was that sshd never started listening on port 22 when I ran with --user-data-file
[01:56] <erichammond> I don't know if it was the specific script I passed or the fact that I was passing user-data.
[02:01] <smoser> mathiaz, erichammond is right on how to test
[02:01] <mathiaz> smoser: ok - great.
[02:01] <mathiaz> smoser: bug 431103
[02:01] <mathiaz> smoser: ^^ reported from an AMI using ubuntu-bug
[02:01] <smoser> printf "#!/bin/sh\necho [02:02] <smoser> printf '#!/bin/sh\necho [02:02] <mathiaz> smoser: I had to add the ec2-images tag when filing the bug though
[02:02] <smoser> then start instance with --user-data-file /tmp/my.txt
[02:03] <mathiaz> smoser: could you update the test case with the exact steps?
[02:03] <smoser> sure.
[02:14] <centaur5> Can a MySQL and PostgreSQL database be retrieved even if you can't chroot into the Ubuntu install to use mysqldump?
[02:26] <smoser> mathiaz, is that system still up ?
[02:27] <mathiaz> smoser: hm -bug 431103?
[02:27] <mathiaz> smoser: the system nope - I've booted another one
[02:27] <mathiaz> smoser: and it's still the same issue
[02:27] <smoser> seriously?
[02:27] <mathiaz> smoser: going through the syslog shows that the fingerprint is actually printed to syslog
[02:27] <mathiaz> smoser: this is eu-west1 though
[02:28] <smoser> yeah, i noticed that too
[02:28] <mathiaz> smoser: ha - Sep 17 01:21:11 ubuntu rsyslogd-2039: Could no open output file '/dev/xconsole' [try http://www.rsyslog.com/e/2039 ]
[02:29] <smoser> yeah. somethign needs to be told to write to /dev/xvc0
[02:30] <smoser> where di dyou see that message ?
[02:31] <mathiaz> smoser: in /var/log/syslog
[02:31] <mathiaz> smoser: http://paste.ubuntu.com/272455/
[02:32] <mathiaz> smoser: more contextual lines: http://paste.ubuntu.com/272456/
[02:32] <smoser> do you know why it looks there ? as opposed to anywhere else ?
[02:32] <smoser> oh. xconsole.
[02:33] <smoser> as in bootsplash or things such as that ?
[02:33] <mathiaz> smoser: line 14 is the last one printed on the console
[02:33] <mathiaz> smoser: I don't know
[02:33] <mathiaz> smoser: IIRC zul was working on a similar bug in the first AMIs
[02:34] <smoser> yeah. i think we've seen something like that before.
[02:35] <mathiaz> smoser: found it: http://paste.ubuntu.com/272457/
[02:48] <mathiaz> smoser: ok - I've updated bug 431103 with my findings
[02:48] <mathiaz> smoser: it's probably a blocker for release (I've marked as High) however I haven't milestoned it
[02:50] <smoser> i dont think its blocker.
[02:50] <mathiaz> smoser: well - you can't trust if your connecting to the right host
[02:50] <mathiaz> smoser: I didn't mean a blocker for alpha6
[02:50] <mathiaz> smoser: I think it should be fixed for karmic
[02:51] <smoser> absolutely
[02:51] <smoser> have you seen: http://alestic.com/2009/08/ec2-poll-fingerprint
[02:51] <smoser> thats why i didn't think blocker for alpha6
[02:52] <mathiaz> smoser: oh right. not a blocker for alpha6
[02:53] <mathiaz> smoser: but it should probably be milestoned for beta
[02:53] <smoser> yeah.
[02:53] <smoser> done
[02:54] <mathiaz> smoser: can you accept nominations?
[02:54] <mathiaz> smoser: I just did it for the bug
[02:54] <mathiaz> smoser: but I was wondering if you could accept a nomination
[02:57] <smoser> i dont think i can.
[02:59] <ScottK> mathiaz: Is that for 431103?
[02:59] <mathiaz> ScottK: yes
[03:00] <ScottK> Someone milestoned it for beta.
[03:00] <mathiaz> smoser: it's all set correcty now
[03:00] <mathiaz> ScottK: yes - I nominated/accepted it for karmic
[03:00] <ScottK> OK
[03:00] <mathiaz> ScottK: and smoser assigned itself and milestoned it for beta
[03:01] <mathiaz> *himself*
[03:01]  * ScottK nods.
[03:08] <szczym> helo all i use ubuntu 8.04.3 server on p4 with 720 mb ram, i have only 2 copy processes over sshd (sshfs) and one cp from local ntfs disk to local ext3 (other disk). all system is very slow, i have 4.05 in top and its not very responsive. the cp process is 500 KB/s but should be 5MB/s. how i could diagnose its problem ?
[03:10] <szczym> in top the moust intensive process is sshd with 0.3 cpu and mount.ntfs with 0.2 cpu. system is using only 200k of swap
[03:11] <smoser> mathiaz, with the data center cloud that i have access to, do i have ability to upload-bundle and such ?
[03:12] <mathiaz> smoser: it should be possible
[03:12] <mathiaz> smoser: I haven't tested it yet
[03:12] <mathiaz> smoser: I was debugging it with Ng - but I didn't get through the whole process
[03:12] <mathiaz> smoser: but the goal is to make it possible
[03:15] <ScottK> No iotop in hardy, so that won't work.
[03:19] <mathiaz> kirkland: right - I've seen the fixes
[03:19] <mathiaz> kirkland: I've noticed some strange things in my seeded node controller install
[03:19] <mathiaz> kirkland: I end up with a cloud controller rather than a node controller
[03:20] <mathiaz> kirkland: it may be related to bad preseeds - I'll investigate later
[03:20] <mathiaz> kirkland: if you could check that both cloud installs are working correctly I can focus on the other tests in the meantime
[03:22] <kirkland> mathiaz: where working correctly means making it through the install, and services running on reboot?
[03:22] <kirkland> mathiaz: i haven't tried to get past that (ie, registration, running an image, etc)
[03:27] <mathiaz> kirkland: yeah - I think that's enough for now
[03:27] <mathiaz> kirkland: making the right packages get installed
[03:46] <eqx311> g'
[04:00] <quizme> libqt4-dbg  <--- where can i get this?  don't i need to modify /etc/apt/sources.list  ?
[04:17] <kirkland> mathiaz: yeah, that's what ttx told me earlier
[04:40] <Bytes> Buenas noches!!!
[05:05] <kirkland> mathiaz: \o/
[05:05] <kirkland> mathiaz: my tests seem to work, save one networking issue on the -nc
[05:05] <kirkland> mathiaz: i'm going to commit a fix for that now
[05:05] <mathiaz> kirkland: cool!
[05:06] <mathiaz> kirkland: by testing, you meant that you've got an UEC cloud up and running on your two computers?
[05:06] <kirkland> mathiaz: on the -nc, i found that i had to change iface eth0 inet from dhcp to manual
[05:06] <kirkland> mathiaz: i have a cc/cloud/storage system
[05:06] <kirkland> mathiaz: and i have an nc
[05:06] <kirkland> mathiaz: both installed from the latest iso
[05:06] <kirkland> mathiaz: all services up and running on boot
[05:06] <mathiaz> kirkland: and you can register/start instances?
[05:06] <kirkland> mathiaz: i have not tried to register
[05:07] <mathiaz> kirkland: oh ok.
[05:07] <kirkland> mathiaz: come one, now, you're asking too much :-)
[05:07] <mathiaz> kirkland: so you've checked that the expected packages are there
[05:07] <mathiaz> kirkland: which is what is written in the test case
[05:07] <mathiaz> kirkland: was this with amd64 or i386?
[05:07] <kirkland> mathiaz: amd64
[05:22] <graphic> Hey I need some help with make files
[05:22] <graphic> Anyone point me in a good direction, I dont know how to get the load_makefile statement to work
[05:40] <Tazm> Hey guys, quick question. Is there any way you can check the raid-5 status of an unbuntu 6.1 server?
[05:41] <ScottK> You mean 6.10?
[05:41] <Tazm> Yeah 6.10 sorry
[05:43] <ScottK> You know 6.10 hasn't been supported in about a year and a half, right?
[05:45] <mathiaz> Tazm: try to look into /proc/mdstat
[05:49] <Tazm> Thanks guys.
[07:02] <kirkland> smoser: manpage search at http://people.canonical.com/~kirkland/search.html is fixed ;-)
[08:08] <artillerytx> Hey guys for some reason my zone files are not updating with my slave server
[08:26] <error404notfound> i have a machine with vpn configured over it, iptables with everything allow and nat for the wan interface. Problem is systems behind that router can't access vpn, not even ping. I found that machines behind the router dont get any route add for the vpn.
[08:30] <twb> Is your server configured to route packets?
[08:33] <soren> Is "the router" and the machine with the vpn connection the same thing?
[08:33] <twb> Good question
[08:33] <error404notfound> twb, ?
[08:33] <error404notfound> soren, yes..
[08:33] <twb> The VPN endpoint must be configured to route packets between the "real" network and the VPN.
[08:33] <soren> Also, why would they get a route for the vpn? Have you configured your dhcp to give one?
[08:33] <error404notfound> soren, i call it router.
[08:33] <soren> Well in that case, why do you need a specific route
[08:33] <soren> +?
[08:33] <error404notfound> twb, how to do that? i never knew :(
[08:34] <twb> error404notfound: turn routing on in sysctl.conf, and make sure your FORWARD chain isn't too liberal (or too restrictive).
[08:35] <twb> net.ipv4.ip_forward=1 and net.ipv6.conf.all.forwarding=1
[09:01] <error404notfound> i have forwarding allowed
[09:01] <error404notfound> twb is gone :(
[09:07] <_ruben> error404notfound: you should do some traffic sniffing (wireshark/tcpdump) to see where you packets are going
[09:09] <error404notfound> _ruben, hmm, ok..
[09:15] <error404notfound> _ruben, 08:13:21.772241 IP (tos 0xa0, ttl  64, id 0, offset 0, flags [DF], proto: UDP (17), length: 757) 192.168.1.123.2048 > 192.168.56.1.5060: SIP, length: 729	REGISTER sip:192.168.56.1 SIP/2.0
[09:15] <error404notfound> that tells me that snom voip phones behind router are sending requests for register, but they can't reach 56.1 vpn ip, i have a tcpdump there as well
[09:17] <_ruben> what kind of vpn is this?
[09:18] <_ruben> and on which interface did you run tcpdump?
[09:23] <error404notfound> _ruben, openvpn
[09:24] <error404notfound> _ruben, on all
[09:25] <error404notfound> i will be back in a while
[09:31] <error404notfound> _ruben, i am back
[10:23] <ttx> soren: I've been trying to rebuild eucalyptus from the bzr tree you helped me to setup yesterday... But my pbuilder is stopping asking me questions about already-applied patches: http://pastebin.ubuntu.com/272655/
[10:23] <buckethead> Hey guys, I can't get the apparmor module to load. Anybody know what I might be missing?
[10:30] <soren> ttx: Just accept the defaults.
[10:30] <ttx> soren: regular buildds apparently don't care.
[10:30] <soren> ttx: Right. Their stdin is /dev/null, so patch accepts the defaults for you.
[10:30] <ttx> soren: I figured that :) Just wondering if that's something we should fix
[10:31] <soren> ttx: Feel free :)
[10:31] <soren> ttx: Just move the patch thing inside the REBUILD_STUBS conditional, and you should be good.
[10:32] <ttx> soren: ok
[10:32] <_ruben> error404notfound: did you run tcpdump on the internal (lan) interface or on the tun/tap (openvpn) interface?
[10:33] <error404notfound> _ruben, i just ran: tcpdump -v | grep ".123"
[10:35] <error404notfound> _ruben, on tun0 e.g. vpn interface: 09:35:13.834107 IP (tos 0xa0, ttl  63, id 0, offset 0, flags [DF], proto: UDP (17), length: 757) 192.168.1.123.2049 > 192.168.56.1.5060: SIP, length: 729
[10:36] <error404notfound> so request is going fine from here..
[10:36] <_ruben> but nothing on the tun interface on the remote box? .. then next would be to check if you see the encapsulated vpn traffic on your wan interface
[10:38] <error404notfound> okay...
[10:40] <error404notfound> _ruben, how do i check for vpn traffic?
[10:41] <_ruben> error404notfound: tcpdump -nvi *yourwaninterface* port *portconfiguredinopenvpn*
[10:41] <error404notfound> _ruben, i do get "11:39:56.229785 IP (tos 0x10, ttl 64, id 43253, offset 0, flags [none], proto UDP (17), length 75) 192.168.56.1.7410 > 192.168.56.6.iax: UDP, length 47" on remote hosts tun interface
[10:42] <_ruben> looks NAT'ed
[10:43] <_ruben> wait .. which ip is trying to contact which ? which is client and which is server? .. because first snippet shows 192.168.1.123 -> 192.168.56.1 and second shows 192.168.56.1 -> 192.168.56.6
[10:44] <error404notfound> _ruben, okay, i see a lot of openvpn packets being sent to remote host from the router
[10:45] <error404notfound> first was from the local router, second one was from remote router
[10:45] <error404notfound> local router vpn ip is 56.22, and remote host's is 56.1
[10:45] <error404notfound> no vpn ip of local router is 192.168.1.1, and that snom voip phone has 1.123
[10:47] <error404notfound> on remote host i get: Sep 17 11:44:47 server ovpn-iaxvpn[7093]: snom-home/202.35.120.118:2049 MULTI: bad source address from client [192.168.1.123], packet dropped
[10:47] <error404notfound> where 1.123 is the ip assign by local router to my snom voip phone
[10:47] <error404notfound> nat is not working good on vpn packets?
[10:51] <_ruben> NAT shouldnt be used at all with VPNs .. that's pretty much the whole idea of VPNs
[10:51] <_ruben> well .. one of several ideas of VPNs
[10:51] <error404notfound> _ruben, so i should remove that line of "MASQUERADE" from my firewall script?
[10:52] <_ruben> error404notfound: probably not, but you should exempt your VPN traffic from it
[10:53] <error404notfound> _ruben, iptables -nvL doest print anything..
[10:53] <error404notfound> _ruben, like? sorry, i am total noob in this case...
[10:53] <_ruben> because its in the nat table and not filter .. use iptables-save to see full output
[10:53] <error404notfound> _ruben, redirect traffic from certain ports to tun0?
[10:55] <_ruben> no, dont touch it at all, just accept it .. assuming you want 192.168.1.123 to able to talk with 192.168.56.6 'directly' and your openvpn config allows that
[10:56] <_ruben> basically, firewalling and vpns are very noob-friendly in my experience
[10:56] <error404notfound> _ruben, have i show you my iptables rules script?
[10:57] <error404notfound> _ruben, http://pastie.org/619956
[10:57] <error404notfound> tells about my interfaces on router and iptables rules and policies
[11:27] <oversize> hello, where can i modify the command that starts dnsmasq with libvirtd? I would like to change the ip range of the dhcp that the libvirt guests use by default but cant find where. htop shows the command with --dhcp-range but as said, where can i modify it?  thx
[11:29]  * soren lunches
[11:51] <maxagaz> how to check which application use which port of the network ?
[11:52] <oversize> maxagaz: try lsof -i
[11:53] <oversize> use sudo to see all daemons
[12:19] <maxagaz> oversize, thanks :)
[12:59] <jdstrand> oversize: see the net-define, net-dumpxml and net-destroy commands for virsh (see man 1 virsh). I also see a 'net-edit' command on karmic, but am not sure when it was introduced
[13:01] <acalvo> hi, anyone using backuppc? (#backuppc is almost dead...)
[13:15] <soren> ttx: My eucalyptus install picked a heck of a time to blow up :(
[13:16] <ttx> soren: beh
[13:16] <ttx> soren: any sign of potential recovery ?
[13:16]  * soren tries wiping it from the face of the Earth again and starts over
[13:21] <ttx> soren: in the meantime I'll ask smoser to do the UEC repack dumbtest, and zul to run the EC2 multipleinstances one. That should get us covered.
[13:21] <soren> ttx: /me hopes so
[13:42] <ttx> zul: could you take care of running the EC2/multiinstance tests from http://iso.qa.ubuntu.com/qatracker/build/all/all ?
[13:42] <zul> ttx: sure
[13:43] <ttx> zul: unless you prefer to test rebundling of the UEC images to EC2, your call.
[13:43] <zul> ttx: heh ill do the multiinstance tests
[13:46] <smoser> ttx, rebundling to ec2 ?
[13:46] <ttx> smoser: see "EC2" at the bottom of http://testcases.qa.ubuntu.com/System/CloudImages
[13:47] <ttx> smoser: I know it must look a lot like what you originally did to get those EC2 deliverables.
[13:48] <smoser> ttx, i can do that test. thats fine.
[13:48] <ttx> smoser: we might have trouble (again) completing the UECimage/UEC test, so having a +1 in that UECimage/EC2 checkbox might help.
[13:50] <ttx> smoser: thx
[13:53] <zul> hah...it would help if I was using the right key
[13:53] <smoser> yeah. fwiw last night i tried to interface with the canonical-cloud instance
[13:54] <smoser> the intent was to upload those to there
[13:54] <smoser> but it seems ill at the moment
[13:54] <ttx> smoser: and soren's setup also blew up, he's working on fixing it.
[13:55] <soren> I think I'm almost there.
[13:55] <ttx> smoser: and the Eucalyptoids are sleeping.
[13:55]  * soren found a few /rather/ annoying issues.
[13:56] <zul> smoser: did the libc6 stuff get fixed?
[13:56] <smoser> yes, but you still see the message if thats what the next question is :)
[13:56] <smoser> the message comes from the glibc in the initramfs
[13:57] <zul> ok
[13:57] <smoser> i'm not 100% certain on that, but definitely all programs that you run in user space are getting the correct libdl and libc
[13:59]  * soren shouts and screams
[13:59] <zul> smoser: also the regular boot messages (ie syslog starting) after that message is not coming up either
[13:59] <soren> smoser: Have you used euca-bundle-kernel?
[14:00] <soren> Err..
[14:00] <soren> euca-bundle-image, I mean.
[14:00] <smoser> soren, i have. i think.
[14:00] <soren> What did you pass as the user id?
[14:01] <soren> I don't see where Eucalyptus exposes my user id.
[14:01] <soren> Oh!
[14:01] <soren> Found it!
[14:01] <smoser> hm...yeah. ec2
[14:02] <soren> hm?
[14:03] <smoser> i'm guessing thats like the aws account number, right?
[14:04] <soren> Right.
[14:05] <soren> It's listed in eucarc.
[14:05] <soren> And nowhere else, as far as I can see.
[14:05] <soren> and only in one of the alias wrappers around ec2-foo-bar
[14:44] <kwork> hello, i have ubuntu on my hdd, now i accuired new pc with new hdd, so i want to transfer my old system to new box
[14:45] <kwork> any ideas
[14:45] <kwork> my first idea was to use dd, but im not sure can i do it, when the size of disks isnt the same
[14:46] <giovani> kwork: the cleanest method would be to do a fresh install on the new hard drive -- but, there are ways of expanding partitions once you copy them over
[14:47] <kwork> im kinda lazy to install all the packages i have installed at the moment
[14:47] <giovani> you could export a list of installed packages, and have them all reinstalled in one shot
[14:48] <giovani> like I said, you can dd the drive over to the new one, and then work to expand the partition(s) -- it's just not always very clean
[15:02] <garymc> i know this may not be the right channel, but If I DMZ my home router to my XBOX360 my gaming stuff works super quick. More so than just with the specified ports opened. What i wanna know is, is it risky to leave my XBOX360 on DMZ. Cos I thought it wouldnt matter so much
[15:07] <soren> ttx: Ok, I've managed to bundle, upload, and register kernel, ramdisk, and image. I can't boot an instance, though.
[15:07] <zul> hah eucalyptus has surpassed mysql in the bug count
[15:07] <soren> How about php5?
[15:08] <soren> I thought that was worse.
[15:08] <ttx> zul: it's still far from beating samba yet.
[15:08] <zul> soren: nope php5 is under control ;)
[15:08] <zul> ttx: its getting there
[15:19]  * soren is fetching the euca2ools bzr branch
[15:19] <soren> 23 MB so far...
[15:19] <soren> ?!?
[15:21] <ttx> smoser: how is the UEC rebundling doing ?
[15:21] <smoser> rebundled, uploaded, registered, waiting on exit from 'pending'
[15:21] <smoser> for both x86_64 and i386
[15:31] <Ng> soren: kirkland: have either of you guys successfully brought up eucalyptus in karmic yet?
[15:31] <Ng> EtienneG and I have both failed so far ;)
[15:43] <kirkland> Ng: define "brought up"
[15:43] <kirkland> Ng: i have installed the cc and the nc
[15:43] <kirkland> Ng: and the services run
[15:43] <kirkland> Ng: i have not registered the nc with the cc
[15:43] <kirkland> Ng: nor run images
[15:44] <Ng> kirkland: ok. I guess I mean "seen it actually work" :)
[15:44] <soren> I'm failing at running instances.
[15:45] <soren> I'm getting authentication failures.
[15:46] <soren> com.eucalyptus.ws.AuthenticationException: User authentication failed.
[15:46] <soren> Specifically.
[15:47] <jpds> http://www.datacenterknowledge.com/archives/2009/09/14/video-data-center-floods-in-istanbul/
[15:48] <ttx> soren: I want to get rid of eucalyptus-javadeps in karmic (was only needed for euca-1.5) -- any objection ?
[15:49] <soren> ttx: No, the sooner the better.
[15:49] <ttx> soren: the process is to subscribe ubuntu-archive to the bug, right ?
[15:49] <soren> ttx: I was /just/ discussing this with Ng.
[15:49] <soren> ttx: Yes.
[15:49] <soren> ttx: And we should make eucalyptus-cloud conflict with it (or Breaks:? I'm not sure) as well.
[15:50] <Ng> \o/
[15:54] <smoser> ttx, where it that i log that i ran that test ? the rebundle-to-ec2 ?
[15:54] <smoser> is it one of http://iso.qa.ubuntu.com/qatracker/test/3007 ?
[15:55] <ttx> http://iso.qa.ubuntu.com/qatracker/result/3006/358
[15:55] <ttx> http://iso.qa.ubuntu.com/qatracker/result/3007/359
[15:55] <ttx> The naming is incorrect
[15:55] <ttx> should be a "rebundle to EC2" test
[15:56] <ttx> not "Install (EC2 Single Instance)"
[15:57] <smoser> ok. i marked those 2 as passed. (which they did)
[15:57] <smoser> also, i attached step by step instructions to the test doc
[15:57] <ttx> smoser: I think the rebundle (step 8) in the EC2 single instance test should be a separate test
[15:57] <ttx> ttx: as well as the user-data test.
[15:58] <ttx> ttx: yes, you're right.
[15:58]  * ttx disconnects dual personality mode for a bit.
[15:59] <smoser> well, rebundling it is essentially exactly what i did
[15:59] <smoser> other than i didn't run all of the tests
[15:59] <smoser> i agree, it should be a separate test
[15:59] <ttx> except that the rebundle in the first case is from EC2, not starting from the img.tar.gz
[16:00] <ttx> smoser: do you have the Alpha6 HEADER on standby ?
[16:11] <soren> jdstrand: https://bugs.edge.launchpad.net/ubuntu/+source/eucalyptus/+bug/431090   Something you could look at?
[16:12] <jdstrand> soren: sure thing
[16:16] <ttx> soren: did you tag your recently-filed new eucalyptus bugs ?
[16:16] <smoser> ttx, i do have it ready
[16:17] <soren> ttx: I believe so.
[16:17] <ttx> smoser: so you are ready to trigger, right
[16:17] <smoser> safety is off
[16:37] <mcupples> I'm running the following script as root and it gives me a value for the variable CONSUMPTIONS.  When I run it as a daemon (user: nagios) the variable is not set.  http://pastebin.com/d7bbd9b7d
[16:38] <samuel> hi, looking for a recommendation for what kind of setup to do for SSO (Single Sign On Setup)
[16:38] <samuel> i'm setting up a 80% Linux Desktop & Server network - with a few potential windows clients
[16:39] <samuel> do i use Kerberos? NIS, LDAP, Samba+Winbind?
[16:40] <soren> ttx: I've milestoned my favourite Eucalyptus bugs now.
[16:59] <ink> mcupples:  you have to "source" the script; in bash you'd do `. blah.sh`
[16:59] <mcupples> ink: I'm sorry, I don't follow.  Lemme see...
[17:00] <jjohansen> well its time for the EC2 kernel status meeting
[17:00] <jjohansen> smoser, erichammond: you ready
[17:00] <smoser> ready
[17:01] <erichammond> jjohansen: as I'll ever be.
[17:01] <jjohansen> alright, lets start then
[17:01] <mcupples> ink: Do you mean for me to change this line to something else?  In my paths?  CONSUMPTIONS=`/usr/sbin/rabbitmqctl list_queues | /bin/grep consumptions | /usr/bin/awk '{print $2}'`
[17:02] <jjohansen> So I don't have much to update from yesterday
[17:02] <jjohansen> I started looking at i386 configs to get the more virtual like config working but didn't get very far on it
[17:03] <smoser> i dont have any progress either.
[17:03] <ink> mcupples:  can the nagios users run /usr/sbin/rabbitmqctl?
[17:03] <smoser> erichammond, without releasing to much information, i'll just comment that the casual browsing user might see images labeled alpha6 in available images on ec2.
[17:04] <smoser> and we've been working on testing those.
[17:04] <ink> su to nagios and try out `/usr/sbin/rabbitmqctl list_queues`
[17:04] <jjohansen> smoser: ohh!, I will have to test that out.
[17:04] <jdstrand> soren: re bug #431090
[17:04] <mcupples> ink: This happens when I try that: sudo -u nagios /usr/sbin/rabbitmqctl list_queues | /bin/grep consumptions | /usr/bin/awk '{print $2}'
[17:04] <mcupples> Kernel pid terminated (application_controller) ({application_start_failure,kernel,{shutdown,{kernel,start,[normal,[]]}}})
[17:04] <soren> jdstrand: shoot
[17:05] <mathiaz> samuel: have a look at https://help.ubuntu.com/9.04/serverguide/C/network-authentication.html
[17:05] <erichammond> smoser: sweet
[17:05] <jcastro> https://wiki.ubuntu.com/Testing/EC2
[17:05] <jdstrand> soren: are they using a patched libvirt? how did they specify the radically different location for the log (and I assume pid and monitor)
[17:05] <smoser> i really hate that you can't rename files on ec2
[17:05] <jcastro> hear ye, hear ye, feedback on EC2 testing plan wanted!
[17:05] <smoser> err... rename images.
[17:05] <jcastro> (and good morning everyone)
[17:05] <smoser> as such, if i want to test am image (and record results against that ami) i have to publish it with its final name
[17:06] <erichammond> smoser: yeah
[17:06] <smoser> (granted, i could one-by-one add launch permissions)
[17:06] <mcupples> ink: So I'd imagine no?  It doesn't have permission?  How would I go about giving the nagios user permission to run that?
[17:06] <erichammond> smoser: You can ... right
[17:07] <smoser> i guess once we (if we ever) get to nightly pushes, then i would feel ok telling people to test against the nighly and then just promoting it to a new name
[17:08] <soren> jdstrand:         <serial type="file">
[17:08] <soren>             <source path='BASEPATH/console.log'/>
[17:08] <soren>             <target port='1'/>
[17:08] <soren>         </serial>
[17:08] <soren> Like so.
[17:08] <soren> ...where BASEPATH is something exotic.
[17:08] <jdstrand> soren: ah, ok. so the pidfile and monitor file should be fine
[17:09] <ink> mcupples:  you could change permissions, but that may not work entirely
[17:09] <ink> another option is to allow the nagios user to run that command w/o a password by editing /etc/sudoers
[17:09] <erichammond> smoser: Sounds reasonable since it would be the same image.  I can't think offhand how that could go wrong.
[17:10] <smoser> jdstrand, sorry to ignore you, but rushing to get final pieces of this release in place.
[17:11]  * jdstrand didn't feel ignored at all
[17:14] <erichammond> jjohansen, smoser: anything else?
[17:14] <jjohansen> nope
[17:15] <smoser> no.
[17:15] <jjohansen> oh 1 more thing
[17:15] <smoser> we've got a busy beta coming up though
[17:15] <zul> stupid internets
[17:15] <jjohansen> though the meetings are canceled for the next week, I will be around mon and tues
[17:15] <ttx> soren, smoser: I'll be afk for the next hours, checking backlog regularly though.
[17:16] <jjohansen> so we might just want to do an informal update tuesday when we are all online at somepoint
[17:16] <jjohansen> smoser, erichammond: thanks, lets call it meeting adjourned
[17:17] <erichammond> thanks
[17:20] <soren> ttx: Alright.
[17:21] <ttx> soren: UEC image test should be completed in ~1h
[17:22] <soren> Awesome.
[17:24] <smoser> so if i want to push something to launchpad
[17:24] <smoser> am i right that the new repo has to be lp:~smoser/some-existing-project-here ?
[17:24] <smoser> i can't just push to lp:~smoser/my-test-stuff
[17:27] <smoser> i see https://help.launchpad.net/NonProjectBranches
[17:29] <jcastro> you need to register a branch first
[17:29] <jcastro> in the webui, you can't just push to an arbritary url
[17:29] <smoser> for anyone interested, i just pushed to https://code.launchpad.net/~smoser/ubuntu-on-ec2/ami-pages . i was annoyed by my un-organized managing of ami pages and HEADER pages for uec-images.ubuntu.com
[17:30]  * soren calls it a day as well
[17:30] <jdstrand> soren: the proper fix for bug #431090 is to get the path to the serial console from xml and the libvirt-<uuid>.files dynamically. I should be able to do this in my resubmission to upstream. In the meantime, as soon as Daniel responds to my requests, I can provide a workaround and upload it today
[17:31] <jdstrand> soren: are you interested in a quick workaround to be removed at some later date when the proper fix is available?
[17:31] <smoser> thanks jcastro
[17:39] <slicslak> i'm updating an old box and i can't seem to update or install any packages.  when i update it erros with lines like: Failed to fetch http://security.ubuntu.com/ubuntu/dists/gutsy-security/main/binary-i386/Packages.gz  404 Not Found
[17:40] <smoser> jcastro, ideally my new branch wouldn't be tagged as "junk" just because its nto a fully hosted launchpad project
[17:40] <aubre> in Karmic Eucalyptus 1.6 when I try to use the web interface to specify a custom location for my bukkits what should I use in the "Walrus host" field?
[17:40] <aubre> when I try to save the defaults it says "Failed to save! Check hostname" the default host name is "host"
[17:40] <giovani> slicslak: isn't gutsy EOLed?
[17:40] <aubre> I tried localhost, it didn't seem to like that
[17:41] <jdstrand> slicslak: gutsy is end of lifed (EOL)
[17:41] <jdstrand> slicslak: you need to adjust your sources.list to point to http://old-releases.ubuntu.com/ubuntu/
[17:42] <jdstrand> slicslak: but gutsy does not receive security support anymore, so you are highly encouraged to upgrade to a supported release (you can upgrade to hardy, the LTS release, directly from gutsy)
[17:45] <slicslak> bleh, yeah, ok i'll upgrade.  thanks.
[17:51] <jcastro> smoser: unfortunately all branches not assigned to a project get tagged as "Junk", there's an ongoing discussion/bug about that. :-/
[17:56] <mathiaz> soren: smoser: have you heard about http://libcloud.org/?
[17:58] <erichammond> jcastro: How does http://testcases.qa.ubuntu.com/System/CloudImages relate?  Should it be replaced with a pointer to https://wiki.ubuntu.com/Testing/EC2 ?
[18:00] <jcastro> erichammond: I am not sure, let me ask someone on the QA team
[18:00] <jcastro> actually, yes it should, on it
[18:01] <smoser> if its not hosted on launchpad, it must be junk
[18:02] <smoser> mathiaz, i hadn't seen that, but was actually going to do some digging for ec2 interface libraries/toolkits today
[18:02] <smoser> (assuming i got all the other stuff done that i had to do, which is a bit like assuming that the grass is blue)
[18:02] <mathiaz> erichammond: the first section is redundant with the EC2/Testcase page
[18:03] <mathiaz> erichammond: the second section (UEC Images) isn't.
[18:09] <aubre> hmm, when I do a sudo euca_conf -addnode 192.168.44.2, and it succeeds via scp, and I do a sudo euca_conf -addnode 192.168.44.2 my node doesn't show up
[18:13] <zul> ttx: our version of samba is FTBFS investigating
[18:13] <nijaba> smoser: It seems that you have completely removed instructions for vmbuilder from https://help.ubuntu.com/community/Eucalyptus?action=recall&rev=66#STEP%205:%20Obtain%20a%20Virtual%20Machine%20(VM)%20Image
[18:14] <smoser> that was by design
[18:14] <smoser> but i'm not set on it.
[18:14] <nijaba> smoser: I understand why, but still think that the instruction were somehow interesting.  Could we add it back at the end of the doc as an optional procedure?
[18:14] <smoser> you're surely welcome to do that. i just wanted the easiest possible route through.
[18:15] <smoser> i also think then that you should like to http://open.eucalyptus.com/wiki/EucalyptusUserImageCreatorGuide_v1.5
[18:15] <nijaba> smoser: understood, and I am going to leave the "main route" as is
[18:15] <smoser> using existing images is a lot easier (at least for getting off the ground)
[18:16] <SyL> is there a way to force a package to be removed if the "stop" action keeps failing?
[18:18] <aubre> I want to restart my eucalyptus install from scratch. Is there a way to do it which removes the /etc/eucalyptus directory structure?
[18:18] <aubre> purge didn't work
[18:19] <SyL> aubre: I had to remove them by hand
[18:19] <aubre> SyL: ok thanks
[18:19] <SyL>  /etc/eucalyptus /var/log/eucalyptus & /var/lib/eucalyptus too
[18:19] <aubre> SyL: tyvm
[18:19] <SyL> the /var/lib/ one will have your keys so you'll have to remove them from node too
[18:20] <aubre> ok thanls
[18:24] <aubre> its not wanting to work, guess I need to reinstall the os
[18:25] <aubre> when I try to reinstall, it says "can't find eucalyptus configuration"
[18:25] <aubre> and I don't understand why it is looking for it when I removed the packages
[18:27] <aubre> I guess with version 1.6 it is best, if you are going to have separate partition to host your walrus / block storage mount it under /var
[18:27] <aubre> because it won't let me configure it otherwise
[18:30] <nijaba> smoser: done.  Let me know what you think
[18:30] <erichammond> Congrats, all: Here's an AMI built on the Karmic (Alpha 4) AMI: http://developer.amazonwebservices.com/connect/entry.jspa?externalID=2803&categoryID=208
[18:39] <smoser> nijaba, looks fine.
[18:39] <smoser> but your ping was ill timed, and got you copied on a response i was writing to someone about that very page
[18:41] <nijaba> :(
[18:42] <nijaba> smoser: ah, ok, got you, now that I have received your reply :)
[18:45] <smoser> ill timed for you (as it got you more mail in your inbox. i wouldn't hvae otherwise thoguht to copy you)
[19:29] <mcupples> Is there a way, in a shell script, to verify that a certain string is present in some output?  CURL=`curl google.com | grep "UA-7277777-1"`
[19:29] <mcupples> I would like to verify that the grep returned a result or did not find it.
[19:30] <mcupples> ink: I added him to sudoers, made sure he didn't have to have a tty and made sure he didn't need a password and it finally worked.  :)
[19:32] <Hypnoz> mcupples, that string should return a 0 or 1 right
[19:32] <mcupples> Hypnoz: No, it returns the line of the curl with that string on it.
[19:32] <Hypnoz> try grep -c
[19:32] <Hypnoz> then test for 0 or 1
[19:33] <mcupples> Hah.  Thanks, Hypnoz!
[19:45] <rickspencer3> hi kkelly18
[19:56] <roodyk> hello i have problem. i make in limit.conf nproc limit 4 for user. after i save user cannot connect to ssh. help please
[20:04] <ink> mcupples:  good deal
[20:15] <roodyk> hello i have problem. i make in limit.conf nproc limit 4 for user. after i save user cannot connect to ssh. help please
[20:30] <cocoa117>  is there a way to disable user login from tty/console?
[20:31] <henkjan> change his shell to /bin/false
[20:32] <cocoa117> henkjan, thanx
[20:45] <orudie> is anyone familiar with pathauto module? how do i make it by default not to do automatic alias?
[20:49] <aubre> I'm running Karmic and Eucalyptus 1.6 - when I try to upload a bundle I get the following error : http://paste.ubuntu.com/273061/
[20:50] <aubre> I think it is walrus related
[20:53] <orudie> sorry wrong chan
[21:02] <Yakuza> 	
[21:02] <Yakuza> hello!! I need a server  mail for my company .... your could show me one?
[21:16] <ikonia> Yakuza: exim, postfix both good, sendmail if you fancy a bigger challange
[21:16] <ScottK> Postfix is the one that's primarily supported here and in the Ubuntu Server Guide.
[21:23] <roodyk> hello i have problem. i make in limit.conf nproc limit 4 for user. after i save user cannot connect to ssh. help please
[21:34] <ink> roodyk:  i don't know what limit.conf is, but on a blank login I use 3 processes already; I can imagine that things in the profile would push that over 4 during login
[21:34] <ink> which could then cause the login to die
[21:36] <roodyk> may be u can tell me best way to limit process for shell users?
[21:41] <ink> roodyk:  probably pick a larger number; like 50
[22:07] <batfastad> Hi everyone. I'm about to start experimenting with ESXi and VMware Server and it seems like JeOS is the ideal candidate for a basic guest installation
[22:07] <batfastad> I've checked through here http://cdimage.ubuntu.com/jeos/ and there;s only 32bit versions
[22:07] <batfastad> But where can I get hold a 64bit version of JeOS 8.04 LTS?
[22:31] <JanC> batfastad: a minimal install of the 64-bit server version should work too...
[22:50] <Tim__> I need help setting up the internet
[22:50] <Tim__> I am connected with an ethernet cable, however I am not receiving internet... I accidentally bypassed the auto setup while installing
[22:51] <batfastad> JanC: Right ok. I know in later versions of the Server CD there's an installation option of "install a minimal virtual machine"... is that what you mean for the 64bit 8.04 server CD?
[22:51] <Tim__> Because I forgot to plug in the ethernet cable... :facepalm:
[22:51] <viezerd> batfastad: I dont think it exist as the linux-image-virtual is only available in 32 bit for hardy --> http://packages.ubuntu.com/search?keywords=linux-image-virtual
[22:52] <viezerd> intrepid and later it will give you 64 bit
[22:53] <batfastad> viezerd: ah right. Because I know JeOS uses the tuned -virtual kernel rather than the regular -server. Might have to go for a full 8.04 server installation in that case
[22:54] <viezerd> do you *really* need the 64 bit ? cause I have the 32 bit jeos hardy running in vmware esxi also
[22:54] <viezerd> and it is really OK
[22:54] <JanC> batfastad: you can always change to the -virtual kernels (unless none exists for hardy?)
[22:54] <Tim__> Anyone know?
[22:55] <Tim__> How to set up an internet connection w/ Ubuntu Server 9.04?
[22:55] <batfastad> I'm looking to allocate 6GB memory to a Zimbra mail VM so I thought that would need 64bit
[22:56] <JanC> hm, apparently only for i386 on hardy...
[22:59] <Tim__> Does *anyone* know how to set up internet on Ubuntu Server 9.04??
[23:01] <Tim__> Screw it I'll just reinstall v.v
[23:02] <SyL> Tim__: what are you trying to do?
[23:13] <Tim__> SyL: I was trying to get my internet working... I got it now
[23:15] <SyL> Tim__: ok
[23:17] <Tim__> Okay, so I just did "sudo aptitude install x-window-system xubuntu-desktop"
[23:17] <Tim__> I was wondering if xubuntu comes with a login manager or if I will have to download and install GDM
[23:17] <Tim__> And I would also like to keep it from starting automatically, I want to have to enter into it from command line