Intro
=====

We will be following the basic install guide at
http://docs.openstack.org/trunk/basic-install/.

Layout
======

We have three networks:

* vmnet10 (NAT, 192.168.0.x) -- management network
* vmnet11 (host-only, 10.10.10.x) -- data network
* vmnet12 (NAT, 192.168.230.x) -- "external"/API network

All networks have x.x.x.1 assigned as the host's IP, and NAT networks have
x.x.x.2 set as the default gateway/NAT box.  For this reason, we will start
all IPs at x.x.x.3 instead of x.x.x.1 (just add 2 to every IP in the guide)

controller.rdo-test
-------------------

* eth0: 192.168.0.3 (mgmt)
* eth1: 129.168.230.7 (ext)

compute.rdo-test
----------------

* eth0: 192.168.0.5 (mgmt)
* eth1: 10.10.10.4 (data)

network.rdo-test
----------------

* eth0: 192.168.0.4 (mgmt)
* eth1: 10.10.10.3 (data)
* eth2: 192.168.230.8 (ext)

Setup
=====

NOTE: make sure that the outside network is reachable
(for example, in our VMWare setup, add `DNS1=192.168.0.2`
and `GATEWAY=192.168.0.2` to /etc/sysconfig/network-scripts/ifcfg-eth0)

Controller Node (controller.rdo-test)
-------------------------------------

1. Add the repositories:
   1. `yum-config-manager --add-repo http://repos.fedorapeople.org/repos/openstack/openstack-trunk/el6-openstack-trunk.repo` (RDO)
   2. `yum install -y http://dl.fedoraproject.org/pub/epel/6Server/x86_64/epel-release-6-8.noarch.rpm` (EPEL)

2. Update to grab the new kernel, and reboot to use it
   `yum -y update && shutdown -r now`

3. Edit the network scripts to contain the correct lines:
   `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth{0,1}`
   1. `ONBOOT=yes`
   2. `NETMASK=255.255.255.0`
   3. `GATEWAY=x.x.x.2` (replace the `x.x.x` with the appropriate prefix)
   4. `BOOTPROTO=none`
   5. `IPADDR=[SEE LAYOUT SECTION]`

4. Edit sysctl.conf to disable route verification
   `$EDITOR /etc/sysctl.conf`
   1. `net.ipv4.conf.all.rp_filter = 0`
   2. `et.ipv4.conf.default.rp_filter = 0`

5. restart networking
   `service network restart`

6. Edit hosts file to contain other hosts
   `$EDITOR /etc/hosts`
   1. `192.168.0.3 controller controller.rdo-test`
   2. `192.168.0.5 compute compute.rdo-test`
   3. `192.168.0.4 network network.rdo-test`

7. Install NTP and configure (?) (skipped for now)
   `yum -y install ntp`

8. Install MySQL
   `yum -y install mysql-server python-mysqldb`

9. Allow connections from anywhere on the network (not needed on RHEL 6.4)
   `sed -i 's/127.0.0.1/0.0.0.0/g' /etc/my.cnf`

10. Run the setup for MySQL
    `/usr/bin/mysql_secure_installation`
    (set root pw to something (we will use 'mysqlpw`), allow root access from non-localhost)

10. Create the DBs, users, and rights:
    1. `mysql -u root --password=mysqlpw`
    2. `CREATE DATABASE nova;
        GRANT ALL PRIVILEGES ON nova.* TO 'nova'@'localhost' \
        IDENTIFIED BY 'password';
        GRANT ALL PRIVILEGES ON nova.* TO 'nova'@'192.168.0.3' \
        IDENTIFIED BY 'password';
        GRANT ALL PRIVILEGES ON nova.* TO 'nova'@'192.168.0.4' \
        IDENTIFIED BY 'password';
        GRANT ALL PRIVILEGES ON nova.* TO 'nova'@'192.168.0.5' \
        IDENTIFIED BY 'password';
        CREATE DATABASE cinder;
        GRANT ALL PRIVILEGES ON cinder.* TO 'cinder'@'localhost' \
        IDENTIFIED BY 'password';
        CREATE DATABASE glance;
        GRANT ALL PRIVILEGES ON glance.* TO 'glance'@'localhost' \
        IDENTIFIED BY 'password';
        CREATE DATABASE keystone;
        GRANT ALL PRIVILEGES ON keystone.* TO 'keystone'@'localhost' \
        IDENTIFIED BY 'password';
        CREATE DATABASE neutron;
        GRANT ALL PRIVILEGES ON neutron.* TO 'neutron'@'localhost' \
        IDENTIFIED BY 'password';
        GRANT ALL PRIVILEGES ON neutron.* TO 'neutron'@'192.168.0.4' \
        IDENTIFIED BY 'password';
        GRANT ALL PRIVILEGES ON neutron.* TO 'neutron'@'192.168.0.5' \
        IDENTIFIED BY 'password';
        FLUSH PRIVILEGES;
        exit;`

11. Install Qpid
    `yum install -y qpid-cpp-{client,server} python-qpid`

12. Disable auth in Qpid
    `$EDITOR /etc/qpidd.conf`
    1. `auth=no`

13. Restart Qpid
    `service qpidd restart`

14. Install Keystone
    `yum install -y openstack-keystone python-keystone python-keystoneclient`

15. Configure Keystone
    `$EDITOR /etc/keystone/keystone.conf`
    1. `admin_token = password`
    2. `bind_host = 0.0.0.0`
    3. `public_port = 5000`
    4. `admin_port = 35357`
    5. `compute_port = 8774`
    6. `verbose = True`
    7. `debug = True`
    8. `log_file = keystone.log`
    9. `log_dir = /var/log/keystone`
    10. `log_config = /etc/keystone/logging.conf`
    11. `connection = mysql://keystone:password@localhost/keystone`

16. Create the SSL key
    `keystone-manage pki_setup && chown -R keystone:keystone /etc/keystone/*`

17. (Re)start keystone and create the databases
    `service openstack-keystone restart && keystone-manage db_sync`

18. Create novarc and source it
    `$EDITOR ~/novarc && source novarc`
    `export OS_TENANT_NAME=admin
     export OS_USERNAME=admin
     export OS_PASSWORD=password
     export OS_AUTH_URL="http://192.168.0.3:5000/v2.0/"
     export SERVICE_ENDPOINT="http://192.168.0.3:35357/v2.0"
     export SERVICE TOKEN=password`

19. Create keystone users and tennants
    NOTE: it is reccomended to use the following code to extract ids:
    `get_id () { echo `$@ | awk '/ id / { print $4 }'`; }`, so the commands
    can be run as `ENV_VAR=$(get_id command)`
    1. (admin tenant) `keystone tenant-create --name admin`
    2. (admin user) `keystone user-create --name admin --pass password --email admin@example.com`
    3. (admin role) `keystone role-create --name admin`
    4. (all together now) `keystone user-role-add --user_id $ADMIN_USER --role_id $ADMIN_ROLE --tenant_id $ADMIN_TENANT`
    5. (service tenant) `keystone tenant-create --name service`
    6. (member role) `keystone role-create --name=Member`
    7. (demo tenant) `keystone tenant-create --name=demo`
    8. (demo user) `keystone user-create --name demo --pass password --email demo@example.com`
    9. (all together now) `keystone user-role-add --user_id $DEMO_USER --role_id $MEMBER_ROLE --tenant_id $DEMO_TENANT`
    10. (all together now) `keystone user-role-add --user_id $ADMIN_USER --role_id $ADMIN_ROLE --tenant_id $DEMO_TENANT`
    11. (service roles) `SVC=(nova glanc swift neutron cinder);
                         for svc in ${SVC[@]}; do
                           export ${svc}_USER=$(get_id keystone user-create --name=$svc --pass=password --tenant-id $SERVICE_TENANT --email=${svc}@example.com);
                           keystone user-role-add --tenant-id $SERVICE_TENANT --user-id $(eval "echo \$${svc}_USER") --role-id $ADMIN_ROLE;
                         done`
    12. (reseller role) `keystone role-create --name=ResellerAdmin`
    13. (all together now) `keystone user-role-add --tenant-id $SERVICE_TENANT --user-id $nova_USER --role-id $RESELLER_ROLE`

20. Create the keystone services
    NOTE: descriptions can be added if desired with `--description`
    NOTE: ec2 was skipped, but can be added with a name and type of 'ec2'
    NOTE: get_id should again be used here
    1. (nova) `keystone service-create --name nova --type compute`
    2. (cinder) `keystone service-create --name cinder --type volume`
    3. (glance) `keystone service-create --name glance --type image`
    4. (swift) `keystone service-create --name swift --type object-store`
    5. (keystone) `keystone service-create --name keystone --type identity`
    6. (neutron) `keystone service-create --name neutron --type network`


21. Create keystone endpoints
    NOTE: ec2 was skipped here
    1. `create_endpoint() { eval "keystone endpoint-create --region RegionOne --service-id \$${1}_SERVICE --publicurl '$2' --adminurl '${3:-$2}' --internalurl '${4:-$2}'"; }`
    1. (nova) `create_endpoint NOVA 'http://192.168.0.3:8774/v2/$(tenant_id)s'`
    2. (cinder) `create_endpoint CINDER 'http://192.168.0.3:8776/v1/$(tenant_id)s'`
    3. (glance) `create_endpoint GLANCE 'http://192.168.0.3:9292/v2'`
    4. (swift) `create_endpoint SWIFT 'http://192.168.0.3:8080/v1/AUTH_$(tenant_ids)' 'http://192.168.0.3:8080/v1'`
    5. (keystone) `create_endpoint KEYSTONE 'http://192.168.0.3:5000/v2.0' 'http://192.168.0.3:35357/v2.0'`
    6. (neutron) `create_endpoint NEUTRON 'http://192.168.0.3:9696'`

22. Install Glance
    `yum install -y openstack-glance python-glanceclient`

23. Configure Glance
    1. `$EDITOR /etc/glance/glance-api.conf`
        1. `sql_connection = mysql://glance:password@localhost/glance`
        2. `admin_tenant_name = service`
        3. `admin_user = glance`
        4. `admin_password = password`
        5. `notifier_strategy = qpid`

    2. `$EDITOR /etc/glance/glance-registry.conf`
        1. `sql_connection = mysql://glance:password@localhost/glance`
        2. `admin_tenant_name = service`
        3. `admin_user = glance`
        4. `admin_password = password`

24. Restart Glance
    `service openstack-glance-api restart && service openstack-glance-registry restart`

    NOTE: this may complain about cinder being missing.  Simply install the
    cinder packages now to fix this problem.

25. Install Glance dbs
    `glance-manage db_sync`

    NOTE: this may complain about not being able to import the correct
    exception.  This is due to an issue in the version of migrate that it
    requests.  Modify the file `glance/db/sqlalchemy/migration.py` to say
    `from migrate.versioning import exceptions as versioning_exceptions`
    (the modified part is the `.versioning` part)

26. Install Nova
    `yum install -y openstack-nova openstack-nova-{api,cert,common,conductor,scheduler} python-nova python-novaclient python-nova-adminclient`

27. Configure Nova
    1. `$EDITOR /etc/nova/api-paste.ini` (under `filter:authtoken`)
        1. `admin_tenant_name = service`
        2. `admin_user = nova`
        3. `admin_password = password`
    2. `$EDITOR /etc/nova/nova.conf`
        1. `sql_connection = mysql://nova:password@192.168.0.3/nova`
        2. `admin_tenant_name = service`
        3. `admin_user = nova`
        4. `admin_password = password`

        and insert (under DEFAULT)

        5. `verbose = True
            scheduler_driver=nova.scheduler.filter_scheduler.FilterScheduler

            # nova-api #
            cc_host=192.168.0.3
            auth_strategy=keystone
            nova_url=http://192.168.0.1:8774/v1.1/
            api_paste_config=/etc/nova/api-paste.ini
            allow_admin_api=true
            dmz_cidr=169.254.169.254/32
            metadata_host=192.168.0.3
            metadata_listen=0.0.0.0
            enabled_apis=osapi_compute,metadata

            # Networking #
            network_api_class=nova.network.neutronv2.api.API
            neutron_url=http://192.168.0.3:9696
            neutron_auth_strategy=keystone
            neutron_admin_tenant_name=service
            neutron_admin_username=neutron
            neutron_admin_password=password
            neutron_admin_auth_url=http://192.168.0.3:35357/v2.0
            libvirt_vif_driver=nova.virt.libvirt.vif.LibvirtHybridOVSBridgeDriver
            linuxnet_interface_driver=nova.network.linux_net.LinuxOVSInterfaceDriver
            firewall_driver=nova.virt.firewall.NoopFirewallDriver
            security_group_api=neutron

            # Cinder #
            volume_api_class=nova.volume.cinder.API

            # Glance #
            glance_api_servers=192.168.0.3:9292
            image_service=nova.image.glance.GlanceImageService

            # novnc #
            novnc_enable=true
            novncproxy_base_url=http://192.168.230.7:6080/vnc_auto.html
            vncserver_proxyclient_address=192.168.0.3
            vncserver_listen=0.0.0.0`


28. Create the Nova DBs
    `nova-manage db sync`

29. Restart Nova Services
    `for s in openstack-nova-{api,cert,consoleauth,scheduler,conductor}; do service $s restart; service $s status; done`

30. Install Cinder
    `yum install -y openstack-cinder python-cinderclient python-cinder`

31. Start iSCSI
    `service iscsid start && service iscsi start`

32. Configure Cinder
    1. `$EDITOR /etc/cinder/cinder.conf`
        1. `sql_connection = mysql://cinder:password@localhost/cinder`
        2. `admin_tenant_name = service`
        3. `admin_user = cinder`
        4. `admin_password = password`
        5. `auth_strategy = keystone` (under the default section)
    2. `$EDITOR /etc/cinder/api-paste.ini` (under filter:authtoken)
        1. `admin_tenant_name = service`
        2. `admin_user = cinder`
        3. `admin_password = password`

33. Create Cinder Volumes
    1. `fdisk /dev/sdb`
    2. `pvcreate /dev/sdb1`
    3. `vgcreate cinder-volumes /dev/sdb1`

34. Restart Cinder services
    `for s in openstack-cinder-{api,scheduler,volume}; do service $s restart; service $s status; done`

35. Install Neutron
    `yum install -y openstack-neutron python-neutron python-neutronclient openstack-neutron-openvswitch`

36. Configure Neutron
    1. `$EDITOR /etc/neutron/neutron.conf`
        1. `core_plugin = neutron.plugins.openvswitch.ovs_neutron_plugin.OVSNeutronPluginV2`
        2. `auth_strategy = keystone`
        3. `rpc_backend=neutron.openstack.common.rpc.impl_qpid`
        4. `qpid_hostname = localhost`
        5. `qpid_port = 5672`
    2. `$EDITOR /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini`
        1. `[database]
            connection = mysql://neutron:password@localhost/ovs_neutron`
        2. `[OVS]
            tenant_network_type = gre
            tunnel_id_ranges = 1:1000
            enable_tunneling = True`
    3. `$EDITOR /etc/neutron/api-paste.ini`
        1. `admin_tenant_name = service`
        2. `admin_user = neutron`
        3. `admin_password = password`

32. Link `plugin.ini` to `plugins/openvswitch/ovs_neutron_plugin.ini`
    `ln -s /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini plugin.ini`

33. Start Neutron Server
    `service neutron-server start`

34. Install Horizon
    `yum install -y openstack-dashboard memcached python-memcached`

35. Start httpd
    `service httpd start`

    NOTE: horizon seems to have an auth error here
    TODO(sross): iptables seems to block outside connections.  Figure out why this is
    TODO(sross): debug auth error

36. Set Qpid as the default messaging system (and restart services after)
    `$EDITOR /etc/{nova,neutron,...}`
    1. `rpc_backend=nova.rpc.impl_qpid`
    2. `qpid_hostname=controller`


Network Node (network.rdo-test)
-------------------------------

NOTE: remember that eth1 is the external network, NOT eth2 (unlike the basic
install guide)

1. Add the repositories:
   1. `yum-config-manager --add-repo http://repos.fedorapeople.org/repos/openstack/openstack-trunk/el6-openstack-trunk.repo` (RDO)
   2. `yum install -y http://dl.fedoraproject.org/pub/epel/6Server/x86_64/epel-release-6-8.noarch.rpm` (EPEL)

2. Update to grab the new kernel, and reboot to use it
   `yum -y update && shutdown -r now`

3. Edit the network scripts to contain the correct lines:
    1. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth0`
       1. `ONBOOT=yes`
       2. `NETMASK=255.255.255.0`
       3. `GATEWAY=x.x.x.2` (replace the `x.x.x` with the appropriate prefix)
       4. `BOOTPROTO=none`
       5. `IPADDR=[SEE LAYOUT SECTION]`
       6. `DNS1=x.x.x.2`
    2. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth1`
       1. `ONBOOT=yes`
       2. `BOOTPROTO=none`
       3. `IPADDR=0.0.0.0`
       4. `ETHTOOL_OPTS="promisc"`
    3. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth2`
       1. `ONBOOT=yes`
       2. `NETMASK=255.255.255.0`
       4. `BOOTPROTO=none`
       5. `IPADDR=[SEE LAYOUT SE

4. Edit `/etc/sysctl.conf`
   `$EDITOR /etc/sysctl.conf`
    1. `net.ipv4.ip_forward=1`
    2. `net.ipv4.conf.all.rp_filter = 0`
    3. `net.ipv4.conf.default.rp_filter = 0`

5. Restart networking
   `service network restart`

6. Edit hosts file to contain other hosts
   `$EDITOR /etc/hosts`
   1. `192.168.0.3 controller controller.rdo-test`
   2. `192.168.0.5 compute compute.rdo-test`
   3. `192.168.0.4 network network.rdo-test`

7. Install and configure NTP (skipped for now)

8. Install Open-vSwitch
   `yum install -y openstack-neutron openstack-neutron-openvswitch`

9. Start Open vSwitch
   `service openvswitch start`

10. Create Virtual Bridging
    1. `ovs-vsctl add-br br-int`
    2. `ovs-vsctl add-br br-ex`
    3. `ovs-vsctl add-port br-ex eth1`
    4. `ip link set up br-ex`

11. Set external bridge IP to whatever eth1's IP was supposed to be
    1. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-br-ex`
       `DEVICE=br-ex
        TYPE=Bridge
        ONBOOT=no
        NM_CONTROLLED=no
        BOOTPROTO=none
        IPADDR=192.168.230.8
        NETMASK=255.255.255.0
        DNS1=192.168.230.2
        GATEWAY=192.168.230.2`
    2. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth1`
        1. `BRIDGE=br-ex`
        2. remove the `IPADDR` line

12. Restart networking
    `service network restart`

??. do the ENABLE A SIMPLE NAT step?

13. Configure Neutron
    1. `$EDITOR /etc/neutron/l3_agent.ini`
        1. `use_namespaces = False`
        2. `debug = True`
        3. `interface_driver = neutron.agent.linux.interface.OVSInterfaceDriver`
    2. `$EDITOR /etc/neutron/neutron.conf`
        1. `auth_host = 192.168.0.3`
        2. `admin_tenant_name = service`
        3. `admin_user = neutron`
        4. `admin_password = password`
        5. `auth_url = http://192.168.0.3:35357/v2.0`
        6. `auth_strategy = keystone`
        7. `core_plugin = neutron.plugins.openvswitch.ovs_neutron_plugin.OVSNeutronPluginV2`
        8. `verbose = True`
        9. `rpc_backend = neutron.openstack.common.rpc.impl_qpid
        10. `qpid_hostname = 192.168.0.3`
    3. `$EDITOR /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini`
        1. `[database]
            sql_connection = mysql://neutron:password@10.10.10.10/neutron`
        2. `[ovs]
            tenant_network_type = gre
            tunnel_id_ranges = 1:1000
            enable_tunneling = True
            integration_bridge = br-int
            tunnel_bridge = br-tun
            local_ip = 10.10.10.3
            firewall_driver = neutron.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver`
    4. `$EDITOR /etc/neutron/dhcp_agent.ini`
        1. `use_namespaces = False`
        2. `interface_driver = neutron.agent.linux.interface.OVSInterfaceDriver`
        3. `dhcp_driver = neutron.agent.linux.dhcp.Dnsmasq`

14. Restart the Neutron services
    `for s in neutron-{openvswitch-agent,dhcp-agent,l3-agent}; do service $s restart; service $s status; done`

15. Copy the 'novarc' file from above and source it
    `scp root@192.168.0.3:~/novarc . && source novarc`

16. Initialize the networks and routers
    NOTE: once again, `get_id` may be useful here
    NOTE: you may wish to install `python-keystoneclient` to make getting
          IDs easier
    1. (create the internal network) `neutron net-create --tenant-id $DEMO_TENANT demo-net --provider:network_type gre --provider:segmentation_id 1`
    2. (create the internal network's subnet) `neutron subnet-create --tenant-id $DEMO_TENANT --ip_version 4 --name demo-subnet $TENANT_NET_ID 10.5.5.0/24 --gateway 10.5.5.1 --dns_nameservers list=true 192.168.230.2`
    3. (create the router) `neutron router-create --tenant-id $DEMO_TENANT demo-router`
    4. (attach the internal network's subnet to the router) `neutron router-interface-add $ROUTER_ID $TENANT_SUBNET_ID`
    5. (create the external network) `neutron net-create ext-net -- --router:external=True --provider:network_type gre --provider:segmentation_id 2`
    6. (create the external network's subnet) `neutron subnet-create --ip_version 4 --allocation-pool start=192.168.230.100,end=192.168.230.200 --gateway=192.168.230.2 $EXT_NET_ID 192.168.230.0/24 -- --enable_dhcp=False`
    7. (connect the router to the external network) `neutron router-gateway-set $ROUTER_ID $EXT_NET_ID`
    8. flush ip addr for ext bridge and add ip with mask of 24?

17. Configure Neutron L3
    `$EDITOR /etc/neutron/l3_agent.ini`
    1. `gateway_external_network_id = [value of $EXT_NET_ID]`
    2. `router_id = [value of $ROUTER_ID]`

18. Restart Neutron L3
    `service neutron-l3-agent restart`

Compute Node (compute.rdo-test)
-------------------------------

NOTE: remember that eth1 is the external network, NOT eth2 (unlike the basic
install guide)

1. Add the repositories:
   1. `yum-config-manager --add-repo http://repos.fedorapeople.org/repos/openstack/openstack-trunk/el6-openstack-trunk.repo` (RDO)
   2. `yum install -y http://dl.fedoraproject.org/pub/epel/6Server/x86_64/epel-release-6-8.noarch.rpm` (EPEL)

2. Update to grab the new kernel, and reboot to use it
   `yum -y update && shutdown -r now`

3. Edit the network scripts to contain the correct lines:
   1. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth0`
       1. `ONBOOT=yes`
       2. `NETMASK=255.255.255.0`
       3. `GATEWAY=x.x.x.2` (replace the `x.x.x` with the appropriate prefix)
       4. `BOOTPROTO=none`
       5. `IPADDR=192.168.0.5`
   2. `$EDITOR /etc/sysconfig/network-scripts/ifcfg-eth1`
       1. `ONBOOT=yes`
       2. `NETMASK=255.255.255.0`
       4. `BOOTPROTO=none`
       5. `IPADDR=10.10.10.4`

4. Edit sysctl.conf to disable route verification
   `$EDITOR /etc/sysctl.conf`
   1. `net.ipv4.conf.all.rp_filter = 0`
   2. `net.ipv4.conf.default.rp_filter = 0`

5. Restart networking
   `service network restart`

6. Edit hosts file to contain other hosts
   `$EDITOR /etc/hosts`
   1. `192.168.0.3 controller controller.rdo-test`
   2. `192.168.0.5 compute compute.rdo-test`
   3. `192.168.0.4 network network.rdo-test`

7. Install and configure NTP (skipped)

8. Install Nova
    NOTE: this also installs libvirt if you don't already have it, as well as kvm/qemu
   `yum install -y openstack-nova python-novaclient`

9. Configure Nova
    1. `$EDITOR /etc/nova/api-paste.ini`
        1. `auth_host = 192.168.0.3`
        2. `admin_tenant_name = service`
        3. `admin_user = nova`
        4. `admin_password = password`
    2. `$EDITOR /etc/nova/nova.conf`
        1. `verbose = True`
        2. `qpid_hostname = 192.168.0.3`
        4. `rpc_backend = qpid`
        5. `auth_strategy = keystone`
        7. `auth_host = 192.168.0.3`
        8. `admin_tenant_name = service`
        9. `admin_user = nova`
        10. `admin_password = password`
        11. `sql_connection = mysql://nova:password@192.168.0.3/nova1
        (just follow the guide for the rest of the conf, substituting in
        192.168.0.3 or 192.168.230.7 as necessary, instead of 10.x.x.x)

10. Restart Nova and Libvirt
    `service libvirtd restart && service openstack-nova-compute restart`

11. Install Openvswitch and Neutron
    `yum install -y openvswitch openstack-neutron openstack-neutron-openvswitch`

12. Start Open vSwitch and Add the internal bridge
    `service openvswitch start && ovs-vsctl add-br br-int`

13. Configure Neutron
    1. `$EDITOR /etc/neutron/neutron.conf`
        1. `debug = True`
        2. `verbose = True`
        3. `rpc_backend = neutron.openstack.common.rpc.impl_qpid`
        4. `qpid_hostname = 192.168.0.3`
        5. `core_plugin = core_plugin = neutron.plugins.openvswitch.ovs_neutron_plugin.OVSNeutronPluginV2`
        6. `auth_strategy = keystone`
        7. `auth_host = 192.168.0.3`
        8. `admin_tenant_name = service`
        9. `admin_user = neutron`
        10. `admin_password = password`
    2. `$EDITOR /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini`
        1. `[database]
            sql_connection = mysql://neutron:password@192.1680.3/neutron`
        2. `[ovs]
            tenant_network_type = gre
            tunnel_id_ranges = 1:1000
            local_ip = 10.10.10.4
            integration_bridge = br-int
            tunnel_bridge = br-tun
            enable_tunneling = True

14. (Re)start Neutron
    `service neutron-openvswitch-agent restart`

Notes
=====

Swift's endpoint caused some issues.  It was not found in the
basic-install/yum version of the guide

should we use `openstack-db --init --service keystone` instead of the
`[service]-manage` executables?

If you have issues with the repos not being found, try `yum clean all` first

WebSocketProxy may complain about an unknown option 'no_parent'.  Simply
update python-websockify to solve this issue

Notes: the compute node needs this
`iptables -I INPUT 10 -p tcp -m multiport --dports 5900:5999 -j ACCEPT` for vnc
to work (check correct rule number -- `iptables -nvL` and look for the REJECT
rule that rejects everything) (assuming you are using the IPTables firewall --
packstack actually inserts this line itself, it would seem)

Note: iptables on the network node don't like gre tunnels, and thus block gre
packets.  As you may surmise, this is not particularly useful.  TODO: figure
out why Neutron doesn't automatically unblock GRE packets in iptables.  For
now simply disabling iptables seems to work fine, but this is not a good
long-term solution

neutron dhcp agent dies after launch, [-may need to chown -R neutron:neutron
/var/lib/neutron-]

need iptables for snat, but can disable firewall by setting firewall to Noop
in both neutron.

in order to get snat working, need to delete the first line of
neutron-l3-agent-postrouting?