AWS FSx Integration with VMware Cloud on AWS

This blog post will describe how to integrate AWS FSx for Windows File Server with Windows Virtual Machines running within VMware Cloud on AWS.

As our customers evaluate migrating to the Cloud, they would often review how to move their most critical asset (their data) securely to the Cloud.

While VMware Cloud on AWS makes cloud migration much easier than ever before, it is still worth for customers considering where they move that data across, especially for customers with a vast volume of data. It is worth breaking down their terabytes/petabytes of data into different buckets and transfer that data to where it’s more appropriate.

The storage on VMC on AWS is currently leveraging the high-performance All-Flash NVME storage on the i3 hosts and might be overkill for data that don’t require that level of performance.

For example, customers might want to:

• Migrate VMs and their virtual hard drive to VMware Cloud on AWS
• Migrate ‘object-based’ storage to S3 (and attach it to VMC over the ENI)
• Leverage S3 for back-up and snapshots of VMs running on VMC
• Migrate ‘database’ workloads to VMC on AWS or to RDS depending on performance and licensing requirements (if using RDS, you can connect it directly to VMC over the ENI)
• Migrate ‘file services’ to EFS (for your Linux VMs) or FSx (for your Windows VMs)

We will focus on the last point above – not the migration itself but leveraging FSx for Windows VMs running on VMware Cloud on AWS.

EFS is AWS Managed File Server for Linux VMs and FSx is the fairly recent AWS Managed File Server for Windows VMs. Brian Graf already documented attaching VMC VMs to EFS so this post will focus on attaching VMC VMs to FSx.

Let’s start by deploying FSx and then I will explain how to map a FSx network drive to our VMC VMs. Below is an illustration of what we’re going to build:

Note that this service has recently been introduced and is not only available across a limited set of AWS regions.

Start by going to the AWS console and get started.

Select the file system type – I picked Amazon FSx for Windows File Server.

I created the file system with the bare minimum (300 GB of storage). The File Server is going to be deployed within a VPC and within an AZ so select the VPC attached to your VMware Cloud on AWS SDDC and and select the same AZ (as traffic over the ENI is free both ways).

You also need to select a Security Group – remember, for traffic to transit to and from VMware Cloud on AWS to its attached AWS VPC, you need to make sure you allow the traffic in two places:

1. On your AWS security group you select below
2. On the VMC Compute Gateway

The ports you need to allow are described in this link.

Note that FSx requires the use of AWS Managed Active Directory… thankfully I have already blogged about it in a previous post (Integrating AWS Directory Services with VMware Cloud on AWS).

The AD I created in my previous post will be what we authenticate against to access the FSx file server.

Once you have deployed your File Server, you will see the details on the console and it’s fairly straight-forward to then go ahead and access it.

See that “fs-04fee22293450b8f4.ad.corp.local” name? That’s what you will need to map to on your network drive.

Let’s jump on a VMware Cloud on AWS Windows VM. Map the network drive and don’t forget to add the \share at the end (I had some issue accessing the folder until my colleague Rob Riemer pointed out to me I was missing the \share).

I can see access the network drive and see the “hello world” file I had created from an earlier Windows machine.

And that’s it! Hopefully you can start seeing how easy it is to build hybrid applications – and by ‘hybrid’, I mean applications that leverage native AWS services and VMs running on VMware Cloud on AWS.

Thanks for reading.