dev-install

dev-install installs TripleO standalone on a remote system for development use.

Host configuration

dev-install requires that:

• an appropriate OS has already been installed
• the machine running dev-install can SSH to the standalone host as either root or a user with passwordless sudo access

For OSP 16 and 17, the recommended OS is RHEL 8.2. There is no need to do any other configuration prior to running dev-install.

Running dev-install

dev-install is invoked using its Makefile. The simplest invocation is:

$ make config host=<standalone host>
$ make osp_full

make config initialises 2 local statefiles:

• inventory - this is an ansible inventory file, initialised such that standalone is an alias for your target host.
• local-overrides.yaml - this is an ansible vars file containing configuration which overrides the defaults in playbooks/vars/defaults.yaml.

Both of these files can be safely modified.

make osp_full performs the actual installation. On an example system with 12 cores and 192GB RAM and running in a Red Hat data centre this takes approximately 65 minutes to execute.

Configuring access to OpenStack from your workstation

To access OpenStack on your standalone host from your workstation you need 2 things:

• A route to the virtual network on the standalone host
• A correct clouds.yaml

dev-install configures both of these with:

make local_os_client

This will configure your local clouds.yaml with 2 entries:

• standalone - The admin user
• standalone_openshift - The appropriately configured non-admin openshift user

You can change the name of these entries by editing local-overrides.yaml and setting local_cloudname to something else.

Network configuration

dev-install will create a new OVS bridge called br-ex and move your external interface on to that bridge. It will also create a new interface called dummy0 which will be added to the br-ctlplane bridge when it is created by TripleO. From here there are 2 network configuration options.

Internal-only networking (default)

With no further configuration, dev-install will configure the public provider network on br-ctlplane with subnet 192.168.25.0/24. The OSP public endpoint will use 192.168.25.1.

For remote access to this network you can use sshuttle. make local_os_client writes a script to scripts/sshuttle-standalone.sh in the dev-install directory with appropriate arguments.

Externally routable public network

This is the best configuration when you can route multiple IP addresses to the external interface of your host, for example a lab system where you control the whole subnet, or a DSAL system where you requested additional FIPs.

dev-install will create the public provider network on br-ex, bridged to your external nic. You need to override several parameters in local-overrides.yaml to provide your networking details. For example, on my DSAL system I have:

public_cidr: 10.46.26.0/23
public_api: 10.46.27.66
public_gateway: 10.46.27.254
public_fip_pool_start: 10.46.27.67
public_fip_pool_end: 10.46.27.75
public_uses_external_nic: true

In this case, my external NIC is on the subnet 10.46.26.0/23. The host has a primary IP in that range, and the default gateway for the subnet is 10.46.27.254.

In addition, I have an allocation of 10 FIPs in the range 10.46.27.66-10.46.27.75 (NOTE: these are also in the same subnet). I have used the first of these, 10.46.27.66 as OSP's public API endpoint. The rest will be used as the allocation pool of OSP's public provider network. That is, floating IPs allocated by OSP are externally routable.

Finally, set public_uses_external_nic to tell dev-install to configure external networking.

Sizing

When idle, OSP17 standalone on RHEL8 uses approximately:

• 16GB RAM
• 15G on /
• 3.5G on /home
• 3.6G on /var/lib/cinder
• 3.6G on /var/lib/nova

There is no need to mount /var/lib/cinder and /var/lib/nova separately if / is large enough for your workload.