I spent an insightful hour this week at the “Clear the Path to VMware in AWS” webinar, and I wanted to share my key takeaways for fellow IT professionals trying to understand the intricacies of this implementation. The session, featuring Erin Stevens and Kyle Grossmiller from Pure Storage, alongside Ben Lipman from Amazon Web Services, cut through the marketing fluff and got right to the architectural heart of the matter. The speakers dove deep into how businesses can seamlessly integrate their on-premises VMware environments with the power and flexibility of Amazon Web Services. It’s clear that the journey to the cloud, while promising, often comes with its own set of hurdles, especially when dealing with complex, mission-critical applications.

Elastic VMware Service (EVS) is an AWS-managed service that enables customers to deploy a fully functional VMware Cloud Foundation environment directly within their own Amazon Virtual Private Cloud (VPC). This integration offers the flexibility to connect seamlessly with native AWS services while preserving the operational consistency and feature set familiar to VMware administrators. Furthermore, the offering can be connected to an on-premises VMware and Pure Storage FlashArray environments to facilitate data migration, disaster recovery and other use cases.

Illustration of AWS EVS

The Hybrid Cloud Problem Isn’t Going Away

The session kicked off by framing a reality we all live in: the hybrid cloud is not a temporary phase. As Ben Kipman highlighted, every enterprise today navigates a hybrid world – balancing public and private cloud, along with on-premises infrastructure. This isn’t just about choosing a single path; it’s about leveraging the unique benefits each environment offers for different applications, whether it’s performance, cost optimization, or security. The common thread? Data. You need your data available everywhere to truly excel.

However, Kyle elaborated on the common challenges: cloud migrations often go over budget and face delays. One of the main reasons? Not every application runs natively as a service in AWS. Imagine having a critical app on an older Windows Server version that isn’t directly supported. That’s where Amazon Elastic VMware Service (EVS) steps in as a vital bridge, allowing you to move these workloads to the cloud without immediate refactoring.

This is where the core of the discussion began.

Amazon EVS: The “Easy Button” for VMware in the Cloud

Ben Lipman from AWS provided a fantastic overview of Amazon Elastic VMware Service (EVS).

He then took us through the specifics of Amazon EVS, emphasizing its core strength: it’s essentially another deployment of VMware, just in an AWS region. This means you get the full VMware experience – vCenter, SDDC Manager, NSX Manager – exactly as you would on-premises. This familiarity is a game-changer for compatibility and ease of management.

EVS truly shines as a “bare metal as a service.” You don’t have to worry about installing vSphere or SDDC. AWS handles all that heavy lifting, delivering a fully configured four-host cluster in about four hours. Need more hosts? A few clicks and about 30 minutes, and you’re good to go. This drastically simplifies expanding your environment.

A key differentiator for EVS is its full administrative access. Unlike some managed services that restrict access, EVS gives you root-level control, allowing you to SSH into appliances, check logs, and install plugins just like you would on-premises. This level of control is crucial for organizations that want to maintain granular management over their VMware environments.

The key takeaway here is its simplicity and familiarity: EVS is a fully managed, native VMware SDDC running on dedicated bare-metal hosts within an AWS region.

What this means for IT pros:

  • It’s the Real Deal: You get the full VMware experience you already know—vCenter, SDDC Manager, NSX Manager—without any modifications. Your existing scripts and operational knowledge apply directly.
  • Bare Metal as a Service: AWS handles the heavy lifting of deploying and configuring the underlying SDDC. Ben noted a four-host cluster can be provisioned in about four hours, with new hosts added in around 30 minutes. This is a massive operational win.
  • Full Admin Control: This was a critical point. Unlike some managed services, EVS provides full root-level administrative access. You can SSH into appliances, check logs, and install third-party plugins just as you would on-premises, giving you the granular control necessary for enterprise environments.

Pure Storage Cloud Block Store: The Enterprise-Grade Storage Layer - The Missing Piece

This is where Kyle Grossmiller from Pure Storage explained how their solution complements EVS. Pure Storage Cloud Block Store (CBS) is essentially the Purity operating environment from their on-premises FlashArray, running natively as software in AWS.

Pure Storage is providing the external storage solution that complements EVS perfectly. As Kyle explained, Pure Storage’s Cloud Block Store runs the same Purity operating environment as their on-premises FlashArray. This means you get all the familiar enterprise features – data reduction, thin provisioning, SafeMode, and replication capabilities – extended seamlessly into the AWS cloud.

One of the most compelling benefits highlighted was the cost savings, especially for disaster recovery (DR). With Pure Storage’s replication, you only send reduced data across the wire, meaning you’re not paying for egress on duplicated bits. If you have 100TB on-premises with 40:1 data reduction, you’re only moving 25TB to the cloud for DR. And if you need to fail back after an outage, you only replicate the changed bits, further reducing egress costs and speeding up recovery.

The flexibility of Cloud Block Store also contributes to cost efficiency. It deploys at the lowest possible capacity and can be non-non-disruptively scaled up as your needs grow. This “pay-as-you-go” model for storage, combined with AWS’s compute savings plans for EVS, offers significant budget control.

For more information, see the official blog post:

Pure Storage Cloud Service for Amazon Elastic VMware Service | Pure Storage Blog

This creates a seamless storage experience with powerful benefits:

  • Consistent Enterprise Features: You get all the features you rely on in the data center—industry-leading data reduction, thin provisioning, SafeMode for ransomware protection, and asynchronous replication—extended directly into your AWS environment.
  • Massive DR Cost Savings: This was one of the most compelling points. Because Pure’s replication sends only unique, reduced data, egress costs for disaster recovery are slashed. If your 100TB on-prem array has a 4:1 data reduction, you only replicate and pay egress for 25TB of data to AWS. For failback, you only send the changed data back. This is a huge TCO advantage.
  • Pay-as-You-Grow Efficiency: CBS deploys with a minimal footprint and can be scaled non-disruptively. This aligns perfectly with the cloud consumption model and avoids overprovisioning.

NVMe/TCP: The Performance Linchpin

Both Kyle and Ben heavily emphasized using NVMe/TCP as the storage protocol between EVS and Cloud Block Store. While iSCSI is supported, NVMe/TCP is the modern, more efficient choice. It delivers lower latency, higher performance (up to 2.5x in some tests), and is more CPU-efficient, which means your expensive ESXi hosts spend their cycles running your VMs, not managing storage I/O.

Actionable Lessons Learned (The Pro-Tips)

The session concluded with practical advice from the preview phase, which is invaluable for anyone planning a deployment:

  • Co-locate everything: Deploy EVS and CBS in the same region, VPC, and ideally, the same Availability Zone to minimize latency and data traversal costs.
  • Match your MTU sizes: Ensure the MTU size is consistent between EVS and CBS (8500 is recommended).
  • Use the right VLANs: EVS has dedicated expansion VLANs for external storage. Use these to leverage the dual 75Gb network interfaces for high availability.
  • Default to NVMe/TCP: It’s the clear winner for performance and efficiency.

Detailed Implementation Guide

For detailed deployment instructions, please refer to the official guide: Elastic VMware Service and Pure Cloud Block Store iSCSI Implementation Guide

My Verdict

The Pure Storage + Amazon EVS partnership offers a robust, familiar, and surprisingly cost-effective solution for extending a VMware footprint into the AWS cloud. It removes major migration roadblocks for legacy apps, provides the enterprise-grade storage features that AWS’s native block storage lacks, and gives IT teams the full control they are used to. It’s a powerful combination that genuinely bridges the gap between on-prem and cloud.