Additional cell considerations and features


Multi cell support is only supported in Stein or later versions.

Availability Zones (AZ)

A nova AZ must be configured for each cell to make sure instances stay in the cell when performing migration and to be able to target a cell when an instance gets created. The central cell must also be configured as a specific AZs (or multiple AZs) rather than the default.

Configuring AZs for Nova (compute)

It’s also possible to configure the AZ for a compute node by adding it to a host aggregate after the deployment is completed. The following commands show creating a host aggregate, an associated AZ, and adding compute nodes to a cell-1 AZ:

source overcloudrc
openstack aggregate create cell1 --zone cell1
openstack aggregate add host cell1 hostA
openstack aggregate add host cell1 hostB


Right now we can not use OS::TripleO::Services::NovaAZConfig to auto create the AZ during the deployment as at this stage the initial cell creation is not complete. Further work is needed to fully automate the post cell creation steps before OS::TripleO::Services::NovaAZConfig can be used.

Routed networks

A routed spine and leaf networking layout can be used to deploy the additional cell nodes in a distributed nature. Not all nodes need to be co-located at the same physical location or datacenter. See Deploying Overcloud with L3 routed networking for more details.

Reusing networks from an already deployed stack

When deploying separate stacks it may be necessary to reuse networks, subnets, and VIP resources between stacks if desired. Only a single Heat stack can own a resource and be responsible for its creation and deletion, however the resources can be reused in other stacks.

Usually the internal api network in case of split cell controller and cell compute stacks are shared.

To reuse network related resources between stacks, the following parameters have been added to the network definitions in the network_data.yaml file format:

external_resource_network_id: Existing Network UUID
external_resource_subnet_id: Existing Subnet UUID
external_resource_segment_id: Existing Segment UUID
external_resource_vip_id: Existing VIP UUID

These parameters can be set on each network definition in the network_data.yaml file used for the deployment of the separate stack.

Not all networks need to be reused or shared across stacks. The external_resource_* parameters can be set for only the networks that are meant to be shared, while the other networks can be newly created and managed.

For example, to reuse the internal_api network from the cell controller stack in the compute stack, run the following commands to show the UUIDs for the related network resources:

openstack network show internal_api -c id -f value
openstack subnet show internal_api_subnet -c id -f value
openstack port show internal_api_virtual_ip -c id -f value

Save the values shown in the output of the above commands and add them to the network definition for the internal_api network in the network_data.yaml file for the separate stack.

In case the overcloud and the cell controller stack uses the same internal api network there are two ports with the name internal_api_virtual_ip. In this case it is required to identify the correct port and use the id instead of the name in the openstack port show command.

An example network definition would look like:

- name: InternalApi
  external_resource_network_id: 93861871-7814-4dbc-9e6c-7f51496b43af
  external_resource_subnet_id: c85c8670-51c1-4b17-a580-1cfb4344de27
  external_resource_vip_id: 8bb9d96f-72bf-4964-a05c-5d3fed203eb7
  name_lower: internal_api
  vip: true
  ip_subnet: ''
  allocation_pools: [{'start': '', 'end': ''}]
  ipv6_subnet: 'fd00:fd00:fd00:2000::/64'
  ipv6_allocation_pools: [{'start': 'fd00:fd00:fd00:2000::10', 'end': 'fd00:fd00:fd00:2000:ffff:ffff:ffff:fffe'}]
  mtu: 1400


When not sharing networks between stacks, each network defined in network_data.yaml must have a unique name across all deployed stacks. This requirement is necessary since regardless of the stack, all networks are created in the same tenant in Neutron on the undercloud.

For example, the network name internal_api can’t be reused between stacks, unless the intent is to share the network between the stacks. The network would need to be given a different name and name_lower property such as InternalApiCompute0 and internal_api_compute_0.

Configuring nova-metadata API per-cell


Deploying nova-metadata API per-cell is only supported in Train and later.


NovaLocalMetadataPerCell is only tested with ovn metadata agent to automatically forward requests to the nova metadata api.

It is possible to configure the nova-metadata API service local per-cell. In this situation the cell controllers also host the nova-metadata API service. The NovaLocalMetadataPerCell parameter, which defaults to false need to be set to true. Using nova-metadata API service per-cell can have better performance and data isolation in a multi-cell deployment. Users should consider the use of this configuration depending on how neutron is setup. If networks span cells, you might need to run nova-metadata API service centrally. If your networks are segmented along cell boundaries, then you can run nova-metadata API service per cell.

   NovaLocalMetadataPerCell: True

See also information on running nova-metadata API per cell as explained in the cells v2 layout section Local per cell