Skip to content

Latest commit

 

History

History
133 lines (103 loc) · 5.5 KB

notes.md

File metadata and controls

133 lines (103 loc) · 5.5 KB

VM (on Ubuntu)

Overview:

  • Setup a VM that is your bootstrap node
  • Setup N VMs to pretend to be your cluster
  • Go to town testing deployments on them.
  • For troubleshooting see troubleshooting.md

Configuration

The bootstrap instance expects to run with its eth0 connected to the outside world, via whatever IP range you choose to setup. You can run NAT, or not, as you choose. This is how we connect to it to run scripts etc - though you can equally log in on its console if you like.

As we have not yet brought quantum up, we're using flat networking, with a single broadcast domain which all the bare metal machines are connected to.

According, the eth1 of your bootstrap instance should be connected to your bare metal cloud LAN. The bootstrap VM uses the rfc5735 TEST-NET-1 range - 192.0.2.0/24 for bringing up nodes, and does its own DHCP etc, so do not connect it to a network shared with other DHCP servers or the like. The instructions in this document create a bridge device ('ooodemo') on your machine to emulate this with virtual machine 'bare metal' nodes.

Detailed instructions

  • add a bridge to your own machine called ooodemo (this emulate the physical network of a cloud). Add this to /etc/network/interfaces:

      auto ooodemo
  iface ooodemo inet manual
      bridge_ports none

  • Exclude that bridge from dnsmasq:

  • add except-interface=ooodemo to your dnsmasq setup (e.g. in a file in /etc/dnsmasq.d/foo)

  • Activate these changes:

      sudo ifup ooodemo
  sudo service libvirt-bin restart

  • Create your bootstrap VM:

    sudo ubuntu-vm-builder kvm precise --rootsize=16384 --swapsize=0 -m 1024
    --proxy=$http_proxy --ssh-user-key=/home/robertc/.ssh/authorized_keys
    --user=stack --name=stack --pass=stack -ai386 --hostname=bootstrap
    --mirror=http://nz.archive.ubuntu.com/ubuntu --domain=
    --libvirt qemu:///system

  • Locate it in kvm - it will be called 'bootstrap', and then..

  • Create a second NIC for the virtual machine, telling it to use the ooodemo shared network device rather than the default NAT setup. This places eth1 of the virtual machine on your 'cloud' network.

  • install openssh if you like, do not install other network services such as DNS.

  • If you installed ssh you probably want to ssh-copy-id your ssh public key in at this point.

  • If you don't copy your SSH id in, you will still need to ensure that /home/stack/.ssh/authorized_keys on your bootstrap node has some kind of public SSH key in it

  • After the base installation, configure eth1 for communicating with your baremetal nodes - put this at the end of your /etc/network/interfaces. The iptables commands expose the metadata service needed by the nodes as they boot and permit baremetal nodes to access the rest of the world by routing via the bootstrap node (libvirt rejects traffic from unknown ip addresses, meaning that using the default libvirt nat environment requires the MASQUERADE for the bare metal nodes unless you reconfigure libvirt as well). Alternatively you can create a second dedicated bridge on your actual machine, put an ip address on it and configure eth0 in the bootstrap node manually or by manually configuring dnsmasq (or similar) on your machine to serve on that bridge.

         auto eth1
         iface eth1 inet static
         address 192.0.2.1
         netmask 255.255.255.0
         up iptables -t nat -A PREROUTING -d 169.254.169.254 -p tcp -m tcp --dport 80 -j REDIRECT --to-port 8775
         up iptables -t nat -A POSTROUTING  -s 192.0.2.0/24 -o eth0 -j MASQUERADE
         up ip addr add 192.0.2.33 dev eth1
  • Create your 'baremetal' nodes.
  • using KVM create however many hardware notes your emulated cloud will have, ensuring that for each one you select ooodemo as the network device.
    • Give them no less than 1GB of disk each, we suggest 2GB.
    • 512MB of memory is probably enough.
    • Tell KVM you will install the virtual machine via PXE, as that will avoid asking you for a disk image or installation media.
    • A nice trick is to make one then to clone it N-1 times, after powering it off.
    • Specify the MAC address for each node. You need to feed this info to the populate-nova-bm-db.sh script later on.
    • Create 2 network cards for each VM: nova baremetal requires 2 NICs.
  • Configure your bootstrap VM:

  • Setup a network proxy if you have one (e.g. 192.168.2.1 port 8080):

         echo << EOF >> ~/.profile
     export no_proxy=192.0.2.1
     export http_proxy=http://192.168.2.1:8080/
     EOF
  • Prep bootstrap VM - do all this as the "stack" user:

  • install git

         sudo apt-get install git

  • clone our demo environment into /home/stack/demo

         cd ~
     git clone git://github.com/tripleo/demo.git

  • if you have varied from the defaults described here, edit the demo environment as needed (see localrc and scripts/defaults).

  • now make all the magic happen (this will use sudo, so don't just wander off and ignore it :))

         ~/demo/scripts/demo

  • Inform nova about your baremetal nodes

         scripts/populate-nova-bm-db.sh -i "xx:xx:xx:xx:xx:xx" -j "yy:yy:yy:yy:yy:yy" add

  • if all goes well, you should be able to run this to start a node now:

      source ~/devstack/openrc
  # flavor 6 is i386, which will work on 64-bit hardware.
  # use 7 for amd64.
  nova boot --flavor 6 --image bare_metal --key_name default bmtest