BGP Routemaps within VMware NSX allow administrators to manipulate BGP routing decisions based on specified conditions. They enable administrators to influence route selection, adjust attributes, and filter routes to meet the needs of their network architecture.
Key Components of BGP Route maps:
Before we delve into practical examples, let’s review the fundamental components of BGP Route maps in VMware NSX:
- Match Conditions: These define the criteria against which routes are evaluated. Match conditions include route prefixes, AS paths, route origins, etc.
- Actions: Actions specify what should happen to routes that match the specified conditions. Common actions include setting attributes like local preference, AS path prepend, and route filtering.
In each route map, multiple things can be set. In this section, we will look at two very common ways to control incoming and outgoing traffic.
AS Prepend
AS prepend is a technique that influences inbound traffic flow by manipulating the AS path attribute of outgoing BGP advertisements. Adding your AS number multiple times to the AS path can make certain routes less preferable or expensive to incoming traffic.
Local Preference
Local preference is an attribute used to influence outgoing traffic flow by indicating the preferred exit point for traffic leaving your network.
The setup
We already got our T0 setup with BGP Peers to our upstream routers.
We also have a VM behind a T1 that is connected to the T0. From here we can trace routes so we can see what path it will take.
A tracert from the VM shows us that we are getting out of the .185 peer.
How to
Create the prefix lists
If not already, then we need to define some prefix lists. It can be done with 0.0.0.0/0 or a more specific prefix if we want to specify what the route map will hit IPs. Head into the NSX manager, Networking, Tier-0 Gateways, and edit the T0. Under its Routing tab, you will see prefixes.
In this example, there is both a specific and an any. We will use the 0.0.0.0/0 here
Create the route maps
Next, we need to prepare the route maps. They can also be found under the tier-0 routing section. Here we will define two routemaps. one for prepend and one for local pref.
Apply the route maps
Now we have set together their route map with its criteria and prefixes. We now need to apply the route maps to the BGP Peers. Under the BGP section of the T0 press the “BGP Neighbors”.
Now we will add the local pref for the .177 peer. This will manipulate the use of this peer for the outgoing traffic for the specified prefixes.
Caution: i have seen that if you choose local pref on one bgp peer and the other peers dont have as prepend setup, then traffic will drop. Not acaltly sure why, but i assume that i could be due to asymetric route. eggress on on the peer with local pref and incoming on the other peer. But when setting as prepend on the other peers it works again.
Traffic from our test VM will now flow through the .177 peer instead of the .185 peer that was before. Looking at the incoming traffic it also goes to the peer with the best route, hence the peer that doesn’t have any prepend on its AS number.
Conclusion
With BGP Routemaps in VMware NSX, administrators can control traffic flow precisely, optimizing network performance, resource utilization, or migration. Experiment with different match conditions and actions to tailor route maps to the unique requirements of your NSX environment.