Very frequently to the strong tendency to move workloads to the public cloud for cost benefits, efficiency and innovation, alongside the need to maintain specific on-premises application environments. The reasons can be different and range from compliance reasons, specific needs in terms of latency or for certain business reasons. Microsoft, aware of these needs, recently announced the release of a new version of Azure Stack HCI, the solution that allows you to build a hyper-converged infrastructure (HCI) to run virtual machines in an on-premises environment and that involves an easy and strategic connection to Azure services. This article lists the main features that will be introduced in the new version of Azure Stack HCI.
What is Azure Stack HCI?
With the arrival of Windows Server 2019, Microsoft introduced the solution Azure Stack HCI, which allows the execution of virtual machines and a wide access to different services offered by Azure.
This is a hyper-converged infrastructure (HCI), where different hardware components are removed, substitutes from the software, able to combine the layer of compute, storage and network in one solution. In this way there is a transition from a traditional "three tier" infrastructure, composed of network switches, appliance, physical systems with onboard hypervisors, storage fabric and SAN, toward hyper-converged infrastructure (HCI).
Figure 1 – "Three Tier" Infrastructure vs Hyper-Converged Infrastructure (HCI)
Azure Stack HCI belongs to the Azure Stack family, which includes a comprehensive and flexible range of solutions to meet the different needs for implementing infrastructure. The Azure Stack portfolio ranges from Azure Stack Hub, which is an Azure extension that can bring the agility and innovation of cloud computing to the on-premises environment, to Azure Stack Edge, a managed Azure appliance that can bring computational power, cloud storage and intelligence in a remote edge of the customer. For more information about the Azure Stack portfolio, see this article.
Figure 2 – Azure Stack portfolio
The new Azure Stack HCI solution, deployed as an Azure hybrid service is named Azure Stack HCI version 20H2 and includes important news.
Figure 3 - Overview of Azure Stack HCI version 20H2 components
Full stack for a Hyper-Converged infrastructure
The operating system of the new Azure Stack HCI solution is based on the core components of Windows Server and has been specially designed and optimized to provide a powerful Hyper-converged platform. The new version of Azure Stack HCI adopts well-established Windows Server technologies such as Hyper-V, software-defined networking and Storages Spaces Direct, and adds new specific features. Following, the innovation areas of this solution are reported.
Dedicated and solution-specific operating system
The operating system of the new solution Azure Stack HCI it is a specific operating system with a simplified composition and newer components than Windows Server 2019.
This operating system does not include roles that are not required for the solution, such as the print server, DNS role, DHCP server, Active Directory Domain Services, services relating to certificates and federated services.
Furthermore, there is the most recent hypervisor also used in the Azure environment, with software-defined networking and storage technologies optimized for virtualization.
The local user interface is minimal and is designed to be managed remotely.
Figure 4 - Azure Stack HCI OS interface
Disaster Recovery Features and virtual machine failover inherent in the solution
In the new version of Azure Stack HCI is included the ability to create stretched clusters to extend a cluster of Azure Stack HCI in two different locations (rooms, buildings or even two cities). This feature provides a replica of storage (synchronous or asynchronous) and contemplates encryption, on-premises site resiliency and automatic failover of virtual machines.
Figure 5 – Stretched cluster in a hyper-converged Azure Stack HCI architecture
In the build phase of creating a new cluster, you can select whether it is an implementation on a single site or stretched on two different sites.
Figure 6 – Options when creating an Azure Stack HCI cluster
If there is a stretched cluster, when creating a volume, you can configure storage replication between the two sites.
Figure 7 – Volume replication options when there is stretched cluster
Optimized the Storage Spaces resync process
In Azure Stack HCI version 20H2 has been completely re-engineered the Storage Spaces Resync, used for storage space repair, to the point where the length of the process is significantly reduced (up to 4-5 times). This improvement makes it possible to speed up the restart of the various systems after the updates are applied.
Figure 8 - Comparison of the times for the monthly application of operating system patches
Updates of the entire stack covered by the solution (full-stack updates)
To reduce the complexity and operational costs of the solution update process, in the new version of Azure Stack HCI a process is contemplated that involves full-stack updating (Firmware / driver along with the operating system) for certain selected partners.
Figure 9 – Solution updates of a Dell EMC-branded Azure Stack HCI solution
Azure Hybrid Service
This new version of Azure Stack HCI is provided as an Azure service, applying a subscription-based licensing model and offering integrated hybrid capabilities.
To expand the capabilities of your solution, you can use Azure solutions to monitor, activate disaster recovery scenarios, manage backup protection, as well as a centralized view of the various implementations of Azure Stack HCI direct from the Azure Portal. Following, details about this Azure hybrid service are reported.
Native integration in Azure
The new Azure Stack HCI natively integrates with Azure services and Azure Resource Manager (ARM). No agent is required for this integration, but Azure Arc is integrated directly into the operating system. This allows you to view, direct from the Azure Portal, the cluster Azure Stack HCI on-premises exactly like an Azure resource.
Figure 10 – Azure Stack HCI integration scheme in Azure
By integrating with Azure Resource Manager, you can take advantage of the following benefits of Azure-based management:
- Adopting Standard Azure Resource Manager-Based Constructs (ARM)
- Classification of Clusters with Tags
- Organizing Clusters in Resource Groups
- Viewing all clusters Azure Stack HCI in one centralized view
- Managing access using Azure Identity Access Management (IAM)
Billing based on a subscription model
Despite being running on-premises, Azure Stack HCI provides invoicing based on Azure subscription, just like any other Azure cloud service. The model is simple and has a cost of 10$ / core / Month, which depends on the cores of the physical processor. In the new pricing model there is no minimum or maximum on the number of licensed cores, much less in the activation duration.
Figure 11 – New licensing model applied for Azure Stack HCI
Dedicated Azure Support Team
Azure Stack HCI becomes an Azure solution, therefore it will be covered by Azure support with the following features:
- You can easily request technical support directly from the Azure portal.
- Support will be provided by a new team of experts dedicated to supporting the new solution Azure Stack HCI.
- You can choose from different support plans, depending on your needs.
For more information, you can access this page.
Familiarity in management and operation
The Azure Stack HCI solution can be activated on different hardware models of your choice and does not require specific software tools to be administered.
Choosing and customizing your hardware
There are several hardware vendors that offer suitable solutions to run Azure Stack HCI and can be consulted by accessing this link. The choice is wide and falls on more than 200 solutions of more than 20 different partners. Azure Stack HCI requires hardware that is specifically tested and validated by various vendors.
The solutions Azure Stack HCI included in the catalog are composed of:
- A server system
- An host bus adapter
- A family of network adapters
Furthermore, you can customize your hardware solution to suit your needs, going to configure the processor, memory, storage and features of network adapters, always respecting the supplier's compatibility matrices.
Figure 12 – Hardware composition for Azure Stack HCI solutions
Management and integration tools
The administrative management of Azure Stack HCI does not require specific software, but you can use existing management tools such as Admin Center, PowerShell, System Center Virtual Machine Manager and even third-party tools.
Using the Windows Admin Center, you can install and configure new architectures Azure Stack HCI and activate virtual systems. Furthermore, With native Windows Admin Center integration with Azure, you can extend functionality with different Azure services, including:
- Azure Site Recovery to implement disaster recovery scenarios.
- Azure Monitor to monitor, in a centralized way, what happens at the application level, on the network and in its hyper-converged infrastructure, with advanced analysis using artificial intelligence.
- Azure Backup for offsite protection of your infrastructure.
- Azure Security Center for monitoring and detecting security threats in virtual machines
- Azure Update Management to make an assessment of the missing updates and proceed with its distribution, for both Windows and Linux systems, regardless of their location, Azure or on-premises.
- Cloud Witness to use Azure storage account as cluster quorum.
Conclusions
The innovations introduced in Microsoft's new hyper-converged solution are very interesting and concern various areas. Azure Stack HCI integrates seamlessly with the existing on-premises environment and offers an important added value: the ability to connect Azure Stack HCI with Azure services to achieve a hybrid hyper-converged solution. This aspect in particular strongly differentiates it from other competitors who offer solutions in this area. Thanks to the changes introduced by this new version it is possible to obtain a complete and more integrated and performing proposition for hyper-converged scenarios.
