With the announcement of general availability (GA) for Azure Stack HCI, we would like to share how the CPPE (Customer Program and Partner Ecosystem) team worked to drive the product features and quality of the release and bring out the voice of the customer in our engineering effort. CPPE has established deep co-engineering relationships with more than 30 customers to deploy and test the early software releases that helped us get to this GA milestone. Along the way, we tracked hundreds of bugs, countless enhancement requests, and learned how customers intend to utilize Azure Stack HCI in their production environments.
Early adopter customers ranged in size from a few clusters in a handful of geographic locations to large, complex datacenters with thousands of branch locations across the world. From manufacturing to retail to financial services, core requirements are common to many customers including:
- the desire for a single control plane to manage cloud and on-premises resources with a global view of cluster from health and monitoring to resource provisioning
- the ability to provide disaster recovery and business continuity (BC/DR) across sites without manual intervention to fail over clustered resources
- the need for resilient branch offices that can run traditional virtual machine-based applications alongside modern containerized applications
- support for a variety of server hardware options with a range of form factors, mounting options, noise levels, ambient temperature ranges and connectivity options (WiFi/5G) in branch and remote offices where often there is no datacenter infrastructure or IT staff
Leveraging these customer insights, the Azure Stack HCI engineering team collaborated with our server hardware partners to create an aligned deployment and operations experience that addresses the primary customer requirements above. Also, in large datacenter deployments with high node count in the clusters it became obvious that an integrated systems offering from our partners would be beneficial due the complexity of rack, stack and integration. This improves time to value for these large deployments.
At the feature level, we’ve built the following into Azure Stack HCI based on the insights above:
- a new Azure-based control plane leveraging Azure Arc to manage clusters from the Azure portal
- an integrated Stretch Clustering engine with a simple setup from a web-based UI
- enhanced 2-node features to enable quick and easy deployment of resilient branch offices at a very affordable price including industry-leading cloud and file quorum witnesses
- support for leading edge server hardware innovation in a variety of form factors and price points from 1 processor / 4 core up to multi-processor / 32 core systems
Looking at vertical insights, we work with retail in consumer goods and grocery which share specific needs for POS and store security. With low numbers of virtual machines running per store (between 8 and 12), Azure Stack HCI with true 2 node deployment with local resiliency is an ideal fit in terms of size, capability, and pricing. With the pricing model of USD $10 per physical core, per month, these customers can choose the right core count server in the Azure Stack HCI catalog and get just the right amount of compute power to meet their needs.
Another important facet of Azure Stack HCI that is appreciated from all our customers is that their expertise in Hyper-V, the Azure hypervisor and other Microsoft tools, such as PowerShell and System Center are directly transferable making the learning curve for this new technology very simple. In many cases, customers can also reuse 3rd party ISVs that they were using for particular in-house processes. The ability of Azure Stack HCI to leverage the Windows Server partner ecosystem plays an important role to simplify deployment and integration.
Now that we’ve reached GA, we would like to expand customer insights to a broader audience. Please download Azure Stack HCI from http://azure.com/hci and give us feedback through the Azure Stack HCI UserVoice. If you are interested in engaging directly with Engineering on our EAP (Early Adopter Program) for future iterations of Azure Stack HCI, complete this survey.