Manage Multi-Tenant BIG-IP Devices with F5 LBaaSv2¶

Overview¶

BIG-IP® devices allow users to create and customize partitions for which specific features that meet a tenant’s needs can be enabled. This type of configuration, called multi-tenancy, allows a greater degree of flexibility in allocating network resources to multiple individual projects. [1]

Multi-tenant BIG-IP and F5 LBaaS

Prerequisites¶

• Licensed, operational BIG-IP device or device cluster.
• Operational OpenStack cloud (Mitaka release).
• Administrator access to both BIG-IP device(s) and OpenStack cloud.
• F5 agent and service provider driver installed on the Neutron controller and all other hosts from which you want to provision LBaaS services.
• Knowledge of OpenStack Networking concepts.
• Knowledge of BIG-IP system configuration, local traffic management, & device service clustering.

Caveats¶

When using BIG-IP Virtual Edition (VE) with the Linux bridge or Open vSwitch Neutron core plugins, F5 LBaaS can be deployed in two ways:

1. Global Routed Mode: BIG-IP VE is connected only to provider networks; all pool member L3 addresses must be routable using cloud infrastructure routers.
2. L2 Segmentation Modes using GRE or VxLAN tunnels: BIG-IP VE has a data connection to a provider network which can route IP packets to the local_ip VTEP addresses of the compute and network nodes through a cloud infrastructure router.

Configuration¶

1. Edit the Agent Configuration File:

$ sudo vi /etc/neutron/services/f5/f5-openstack-agent.ini

2. Configure the L2 Segmentation Mode settings as appropriate for your environment.

   • Device VLAN to interface and tag mapping
   • VLAN device and interface to port mappings
   • Device Tunneling (VTEP) selfips (if using VxLAN or GRE tunnels)
   • Tunnel Types
   • Static ARP population for members on tunnel networks

Footnotes