NSX-T 3.0 Lab: Federation RTEP Configuration

Intro

Welcome to Part 4 of the NSX-T 3.0 Lab Federation Series.
In my Last Post I Setup the Global Manager and configured our Locations DCA and DCB.
In this post we will be configuring the RTEP on our Edge clusters.

What is an RTEP?

Put simply an RTEP is a Remote Tunnel End Point.
Just like the host and Edge TEPs, the Geneve traffic needs to be encapsulated and de-encapsulated by a TEP. The RTEP is used for cross site traffic from Edge node to Edge node between sites.

The Build

RTEP Configuration

Before we get into the meat of the build there are some pre-requisites that we need to do first.

Prerequisites

  • Verify that each location participating in the stretched network has at least one Edge cluster. 
  • Determine which layer 3 networks and VLANs to use for RTEP networks.
  • Intra-location tunnel endpoints (TEP) and inter-location tunnel endpoints (RTEP) must use separate VLANs and layer 3 subnets.
  • Verify that all RTEP networks used in a given Federation environment have IP connectivity to each other.
  • Verify that external firewalls allow cross-location RTEP tunnels, and BGP sessions between Edges.
  • Configure the MTU for RTEP on each Local Manager. The default is 1500.

So I already have an Edge cluster in each site, my nested setup allows Layer 3 between the sites theres no firewall in between for me.
So that means I need to configure the RTEP MTU and I will need an RTEP IP Pool.

RTEP MTU

Before I setup the RTEP Pool or configure the RTEP we need to set the RTEP MTU value, this should be set to as high as your physical network supports and ideally 1600 minimum.
The setting can be changed by logging into each of your local managers or you can use the global manager and connect to the local ones using the drop-down at the top.

Once on the local manager navigate to System, Fabric, Settings then on ‘Remote Tunnel Endpoint’ click EDIT

Set the desired MTU value and click SAVE.
Repeat for the other local manager/s.

RTEP IP Pool

The process is the same as setting up the host TEP IP Pool.
Connect to the local manager and go to Networking, IP Address Pool then under ‘IP ADDRESS POOLS’ click ADD IP ADDRESS POOL

Enter a name then click Set under ‘Subnets’, click ADD SUBNET, then IP Ranges I’m going to use 10.152.1.0/24 subnet for DCA and 10.157.1.0/24 for DCB once you have the configuration set click ADD then APPLY

Finally click SAVE.
Now repeat for the other Local manager/s

Configure TEP on the Edge Cluster

We can now configure our Edge RTEPs, there are a couple of ways of getting to the configuration dependant on which system you are connected to.
If you are on the Global manager then click Location Manager then on the location click NETWORKING.

You will be prompted to select the relevant edge cluster, click CONFIGURE

If you are on the Local manager then go to System, Get Started and on ‘Configure Remote Tunnel Endpoint’ click GET STARTED

Either way you will end up on this screen on the Local manager.

Click SELECT ALL all hosts in the cluster must be configured, select the ‘Host Switch’ from the drop-down menu, the Teaming Policy can be left blank. Enter the ‘RTEP VLAN’ and select the ‘RTEP Pool’ from the drop-down.
The ‘Inter Location MTU’ is populated by default and uses the value we set earlier.
Click SAVE.
Repeat for the other Local manager/s

We will get the message “All Edge Nodes has been configured successfully. Edit can be done under “System >> Fabric >> Nodes >> Edge Transport Nodes >> Tunnels”.
so lets go there and take a look.
We can see the configuration details in the ‘Remote Tunnel Endpoint’ section the rest is blank as we haven’t setup any networking yet.

Lets jump over to our Global Manager and do a quick check.
It’s here that we run into a bug as the Global Manager still shows the ‘Remote Tunnel Endpoint Status’ as ‘Not Configured’.

To fix this we need to remove and re-add our Edge Nodes from the Edge cluster.
Go to Edge Clusters select the cluster and click EDIT.

Select the Edge nodes and move them to the left then click SAVE.

Edit the Edge cluster again this time moving the Edge nodes to the right.
Repeat for the other DC’s Edge clusters.
Once done go back to the Global manager.
Go to System, Overview again. You may need to refresh the screen.
We can see up RTEPs are now up 🙂

Thats it for RTEPs, in the next post I’ll be deploying our stretched Tier-0

NSX-T 3.0 Lab: Federation Tier-0

Leave a Reply

Your email address will not be published.