[07:19] <Garo_> Hello. I'm planning to build a grsecurity enabled kernel for ubuntu 10.04 running inside EC2 EBS root instance. So far I haven't found any good pointers how to build an EBS root based ubuntu kernel for the ec2. Are there any good pointers and/or guides on compiling ubuntu for ec2?
[07:20] <flaccid> Garo_: just use pvgrub, you can run any kernel you want
[07:23] <Garo_> flaccid: thanks for the tip, I'll start with reading over this http://backdrift.org/xen-pvgrub-howto
[07:23] <flaccid> the official docs are fine
[07:40] <lethalduck> Hi all, been a while since I've been on irc. I'm playing with the idea of putting a mongodb instance and a node.js server on a single amazon EC2 micro instance. It's just for a self promotional services site. I now it's generally not a good idea to run a database in a DMZ, but I'm aiming to do this cheap, without having to lease multiple instances of EC2 machines. What are your thoughts on the potential security issues around
[07:40] <lethalduck> this idea?
[07:41] <flaccid> lethalduck: just don't open ports to the world
[08:16] <lethalduck> Yeah, common sense stuff.
[08:58] <kim0> Morning everyone
[09:00] <flaccid> evening
[09:04] <lethalduck> night
[13:34]  * Blulix waves
[13:36] <Blulix> All these people and everyone idling? xD
[13:47] <kim0> Blulix: hehe hey o/
[15:08] <koolhead11> kim0, hey
[15:19] <kim0> koolhead11: hey :)
[15:20] <koolhead11> kim0, back to work?
[15:21] <kim0> yep
[15:22] <koolhead11> cool!! saw the updates :D
[15:22] <kim0> Yeah!
[15:22] <kim0> any idea how to merge the docs with the wiki ?
[15:22] <kim0> what did you say the source format was
[15:23] <koolhead11> kim0, as of now its plain html but we will have it in LAteX
[15:25] <koolhead11> kim0, my boss will be the correct person to answer and give more details on same :)
[15:25] <kim0> yeah I'll ping him
[15:34] <koolhead11> hey TeTeT
[18:08] <Blulix> Hey guys, if I have my controller and node controller on the same router which is also connected to the internet, would that be a problem?
[18:09] <Blulix> (using EUC CD Install)
[18:11] <obino> Blulix: nope. It should all be dandy. Are you seeing problems?
[18:12] <Blulix> Sort of yeah, I used automatic network setup in the install for the controller and the Node still cant see the controller on the network.
[18:12] <Blulix> I tried having them on the same machine instead but that just posed more problems.
[18:12] <Blulix> So now I'im  down to trying them both on seperate VMs on my pc to see if i can get it to work at all.
[18:17] <Blulix> I think I'm missing a step, i install the Controller (I use automatic network setup) and call the hostname 'controller', wait for the install to finish and use 'sudo apt-get update' then 'sudo apt-get upgrade'.
[18:18] <Blulix> Then reboot it. Then I install the node on the other system with hostname 'nodecontroller' and when it gets to "Select cloud installation node" it says "No Eucalyptus clouse controller was found on your network".
[18:18] <Blulix> If i type in the controller name it says it finds it but that it doesnt find walrus.
[18:19] <Blulix> even though i chose to install everything (except node controller) onto the controller.
[18:20] <Blulix> ( i chose standard guided partitioning btw, not the one with LVM, does that make a difference? )
[18:45] <obino> Blulix: the autoregistration is based on avahi
[18:45] <obino> are the machines on the same subnet?
[18:45] <Blulix> same subnet?
[18:46] <obino> what's your network setup?
[18:46] <Blulix> They're both connected to the same router via eth0
[18:46] <obino> CLC/CC sits on the router right?
[18:46] <obino> router or switch?
[18:46] <Blulix> Well switch to the router
[18:46] <Blulix> one sec
[18:47] <Blulix> internet <-> router <-> switch <-> Controller(eth0)/NodeController(eth0)
[18:47] <Blulix> though thats basically internet<->router<->Controller(eth0)/NodeController(eth0)  to the computers.
[18:48] <Blulix> The controller has an eth1 if i need it.
[18:48] <Blulix> but atm its unused
[18:50] <obino> and you for the CLC/CC/SC/Walrus installed correctly?
[18:50] <obino> did you get to the webUI?
[18:51] <Blulix> It didnt error when installed so i guess so
[18:51] <Blulix> and the webui hsould come up one sec
[18:51] <Blulix> ah its running vm atm so i dont know
[18:52] <Blulix> and apparnently Walrus didn't install (or so the nodecontroller install says)
[18:52] <Blulix> and offers to be the walrus controller itself
[18:54] <Blulix> Gunna start up the one i made on the hardware and see if the webUI works
[18:55] <Blulix> how can i tell (from the controller side) if they've all isntalled correctly?
[18:56] <obino> for the front-end you can check if the webUI is coming up
[18:56] <obino> then check under configuration if walrus, sc and cc are registered
[18:56] <Blulix> apparently the web ui isnt coming up
[18:57] <Blulix> wait what
[18:57] <Blulix> i put in the local IP without the port 8443
[18:57] <Blulix> and it said "Ubuntu enterprise cloud - redirecting"
[18:57] <Blulix> then goes the to local IP but with the port
[18:57] <Blulix> and nothing comes up
[18:57] <Blulix> so the redirect works, but the ui itself doesnt :s
[18:58] <Blulix> Oh
[18:58] <Blulix> and now it works... the WebUI is up now o.O
[18:59] <Blulix> It asks for cloud Host IP i have no idea if its need the LAN IP or WAN IP
[19:10] <Blulix> So with the WebUI up do I do anything?
[19:16] <Blulix> The DNS config is set to localhost, shouldnt i put that as the router gateway?
[19:18] <Blulix> The ubuntu tutorial conviniently forgot to tell me how to set up walrus and cluster on the webUI x.x
[19:22] <obino> did you install walrus?
[19:22] <Blulix> I ticked it on install yes.
[19:22] <Blulix> Now I@ve been into the WebUI I registered Walrus with host ip as the controllers LAN IP.
[19:23] <skizza1> anyone need hosting?
[19:23] <obino> it should be the 'public ip' name
[19:23] <Blulix> My internet IP?
[19:23] <skizza1> Anyone need hosting?
[19:24] <Blulix> Should that go for the controller IP is asks for too?
[19:24] <obino> Blulix: do you have a home router?
[19:24] <Blulix> Yes
[19:24] <Blulix> That's what I'm using
[19:24] <obino> k, then not the public IP, just use the IP you have
[19:24] <skizza1> obino
[19:24] <skizza1> Blulix
[19:24] <Blulix> i told the controller to start broadcasting walrus and it says it's already doing it :s
[19:24] <Blulix> Yeah?
[19:25] <skizza1> need any hosting?
[19:25] <Blulix> Uh not really i guess x3 I like hosting my own stuff personally.
[19:29] <smoser> oneiric builds on EC2 now!
[19:29] <smoser> $ lsb_release -a
[19:29] <smoser> No LSB modules are available.
[19:29] <smoser> Distributor ID: Ubuntu
[19:29] <smoser> Description:    Ubuntu oneiric (development branch)
[19:29] <smoser> Release:        11.10
[19:29] <smoser> Codename:       oneiric
[19:29] <smoser> $ cat /etc/cloud/build.info
[19:29] <smoser> build_name: server
[19:29] <smoser> serial: 20110503.1
[19:30] <Blulix> I am so confused
[19:31] <Blulix> The ubuntu tutorial puts the 'install node controllers' which is supposed to automatically connect to the controller BEFORE setting everything up in the WebUI... o.O
[19:33] <obino> Blulix: yes
[19:34] <obino> you can check if the nc is been polled looking at the log file
[19:34] <obino> on the NC machine, look at /var/log/eucalyptus/nc.log
[19:34] <obino> you can use tail -f nc.log and see if it is been polled
[19:38] <Blulix> tail: cannot open '/var/log/eucalyptus/nc.log' for reading: No such file of directory
[19:38] <Blulix> (just incase i forgot to mention, Walrus CC and NC are all on the same machine)
[19:39] <Blulix> woos bad that
[19:39] <Blulix> not NC
[19:40] <Blulix> Everything BUT NC is on the controller macine xD
[19:40] <Blulix> So Cloud controller, walrus, cluster controller and storage controller.
[19:40] <Blulix> Also I don't get what 'polled' is.
[19:43] <obino> the CC polls regurlarly each NC
[19:43] <Blulix> ah
[19:43] <Blulix> um
[19:43] <obino> you should see both in the cc.log and nc.log the results of it
[19:44] <obino> you can see errors on the CC side or NC side or no errors at all :)
[19:44] <Blulix> wolud running the NC on a virtual machine from my PC on the same network work for a test?
[19:44] <obino> NC will need to be on hardware
[19:44] <obino> it will need to interact with the hypervisor
[19:44] <obino> to start/stop instances
[19:44] <Blulix> Hypervisor?
[19:44] <obino> kvm
[19:45] <Blulix> So I could run the...Cloud/Cluster/Walrus Controller on a VM?
[19:45] <obino> I discourage that
[19:46] <obino> it is possible in theory but debugging is harded
[19:46] <Blulix> This is jsut for testing atm mind.
[19:46] <obino> harder
[19:46] <obino> if you are going to test the full functionality, I would recommend to to use 2 machines
[19:46] <Blulix> I'm in a bind and I have a PC with windows on which uses a RAID0 that linux refuses to install to an ext partition on.
[19:46] <obino> do you have 2 machines?
[19:46] <obino> got it
[19:47] <Blulix> I do yes but as I said, linux cant partition open space on my RAID0
[19:47] <Blulix> No idea why
[19:47] <Blulix> it sees it as nvidia/mapper/<load of garbled text, probably an ID>
[19:48] <Blulix> I just wanna get the whole procedure down for when I install the real cluster x.x
[19:48]  * Blulix wishes it could all be done on multiple VM's on Virtualbox
[19:48] <obino> well, have you read our documentation?
[19:48] <obino> you can put the NC on the bare hardware
[19:49] <obino> then you can put the front end on the VM
[19:49] <Blulix> I'm afraid all I have read so far is the official ubuntu tutorial which does not help.
[19:49] <obino> remember that this is not fully supported
[19:49] <obino> go on open.eucalyptus.com
[19:49] <obino> then disable the autoregistration
[19:49] <Blulix> the front end being the Cloud/cluster/walrus controller?
[19:49] <obino> yes
[19:49] <obino> actually tell you what
[19:50] <obino> as first try no need to disable autoregistration
[19:50] <obino> make sure that the VM is brdiged on your network
[19:50] <Blulix> Sorry I cant keep up I'm on open.eucalyptus.com on my PC and now I'm confused
[19:50] <obino> so that is reachable by the NC
[19:50] <obino> I wanted to give you where the full docs are
[19:50] <Blulix> All my VM's are connected to my network by default via a virtual adaptor on my pc.
[19:50] <obino> in case you need them
[19:50] <Blulix> oh ok
[19:51] <Blulix> I wish the server had a WUBI installer xD This'd be sorted already
[19:52] <obino> there was cloud on a stick
[19:52] <obino> dustin kirkland produced it
[19:52] <obino> same thing as a live-cd ideas
[19:52] <Blulix> Ooo
[19:53] <obino> limited functionality
[19:53] <obino> if you just need to play as a user have you tried the ECC?
[19:54] <Blulix> Nope no idea what that is
[19:54] <Blulix> VirtualBox has 'bridged adaptor feature, am i useing that?
[19:54] <obino> check the testdrive on open.eucalyptus.com
[19:57] <Blulix> mkay
[19:57] <Blulix> Will runnign the bridged adaptor give my Vm it's own IP? x3
[19:59] <Blulix> Why does the install have Walrus Storage Service and Storage Controller both installed? They sound like they do the same thing.
[20:00] <obino> they don't
[20:00] <obino> walrus is like S3
[20:00] <obino> and the SC provides EBS
[20:01] <obino> so you need one SC per cluster
[20:01] <obino> while you have one walrus per cloud
[20:01] <Blulix> Ah oki
[20:01] <Blulix> Well i dont know what S3 and EBS are but i follow
[20:03] <obino> S3 is a get/put storage
[20:03] <obino> works putting and getting full 'buckets'
[20:04] <obino> EBS instread exports block devices  to running instances
[20:04] <obino> instances will see them as new disk drives
[20:04] <Blulix> ah oki
[20:05] <Blulix> Oh nice, bridging the VM  does give it it's own IP ^^
[20:05] <Blulix> Wish i knew that earlier
[20:14] <Blulix> Oh um
[20:14] <Blulix> It always asks for a Mail anem and later on a domain name.
[20:14] <Blulix> No idea what to put there
[20:26] <Blulix> oh you have got to be joking...
[20:27] <Blulix> I think my FrontEnd and NC were different ubuntu versions >.>
[20:28]  * Blulix facepaws
[20:28] <Blulix> facepalms* even
[21:07] <Blulix> and NOW it sees it
[21:07] <Blulix> ... darn
[21:07] <Blulix> sorry x.x thanks for helping me though~
[22:03] <vorlov> hello im looking for some help with amazon ec2 help? and latest ubuntu natty release
[22:09] <vorlov> im having problems with natty ami - reboots fine, but after first shutdown it never recovers
[22:09] <vorlov> on an m1.large
[22:09] <vorlov> here's the syslogs http://pastebin.com/Acv87Cdr
[22:29] <hallyn> smoser: i'm trying to create an ec2 image using vmbuilder, and it's getting hung up on dependencies for linux-image-virtual.  Does that sound at all familiar to you?
[22:33] <hallyn> zul: ^ actually you did the vmbuilder-ec2 prototype, so maybe i should be asking you
[22:34] <hallyn> I see, I shouldn't install a kernel pkg.  but i want to.
[23:50] <Kyle__> How does one go about building a windows 7 image for UEC?