Microsoft Azure Stack HCI
HPE GreenLake for Microsoft Azure Stack HCI provides HPE and Microsoft validated integrated systems. This solution provides an as‑a‑service cloud experience for on-premises compute and storage resources and can include an Azure subscription if customers choose Hewlett Packard Enterprise as their Microsoft Cloud Solution Provider (CSP). Having HPE as the CSP also gives the confidence that the whole HPE GreenLake for Azure experience has HPE as the customer’s technical one-stop support specialist.
The HPE GreenLake for Microsoft Azure Stack HCI cloud experience provides on-demand capacity, combining the agility and economics of public cloud with the security and performance of on-premises deployment. When customers purchase this service, they can gain the ability to accelerate digital transformation with cloud benefits of fast deployment, scalability, and pay-per-use economics—all within their control and choice of on-premises or colocation environment. HPE provides an integrated Microsoft Azure hybrid cloud platform built on HPE GreenLake cloud-native infrastructure that incorporates compute, storage, and networking, to run applications and workloads on the Azure Stack HCI platform and Azure services on-premises.
HPE GreenLake for Microsoft Azure Stack HCI delivers three service options offering customers choice based on the level of management they prefer from HPE. The customer can choose from the following:
Managed Infrastructure
Provides a preconfigured, integrated system, with your choice of cloud module instances, services for startup, advanced configuration, and ongoing support for the integrated system. In addition, an account team is assigned and acts as the Customer’s advocate and operational and/or technical focal point for support. This includes the monitoring and alerting of issues, patches, and advisories that could impact the Customer environment. The customer is primarily responsible for proactive monitoring, managing, and operating the integrated system with HPE support.
Managed Platform
Includes Managed Infrastructure with additional management services for the Azure Stack HCI OS including proactive monitoring, incident resolution, and implementing patch updates. The customer is responsible for monitoring, managing, and operating the workloads that are deployed on top of the Azure Stack HCI OS.
Managed Workload
Includes Managed Platform plus day-to-day management and operation of specific workloads running on the Azure Stack HCI OS. It delivers a fully managed service for Azure workloads such as K8s containers, SQL databases, virtual desktops, and data services in a truly hybrid environment. This service provides an end-to-end managed solution experience for Microsoft Azure hybrid cloud customers.:
Service Benefits
The service is designed to provide the following benefits:
- A complete solution including scalable hyperconverged infrastructure, software, and services
- Faster time to value with hybrid-by-design, preconfigured HCI on industry-leading HPE systems
- Simplified IT with implementation, support, and monitoring; with operations and management options
- Convenient, pay-per-use billing based on metered usage; includes hardware, software, and services (can include Azure services in a single bill if HPE is the CSP for Azure subscription)
- Cloud-like flexibility to meet business needs
Additional benefits are available dependent on the service level selected:
- Free up customer operational resources from the day-to-day overhead of maintaining an optimized IT environment
- Help gain advanced insight and analytics into the day-to-day operational status of the Systems
- Help gain insight over time into the performance of the Systems to aid in making strategic decisions based on that information
- Tooling and automation to help maximize efficiency and effectiveness of IT operations
Service Feature
HPE GreenLake for Microsoft Azure Stack HCI provides the following:
| Feature | Managed Infrastructure | Managed Platform | Managed Workload |
|---|---|---|---|
| Relationship management | + | + | |
| HPE GreenLake integrated system | |||
| Service implementationm | + | + | |
| IT service management (ITSM) | + | + | |
| System operations and administration | + | + |
= Basic deliverables
+ = More expansive deliverables
HPE GreenLake Integrated System
HPE GreenLake integrated system for Microsoft Azure Stack HCI is purpose-built for deploying Microsoft’s hyperconverged solution. The solution is modular, flexible, and factory-built and can be used to run most common needs that correspond to Azure Stack HCI solution capabilities and technical use cases such as trusted enterprise virtualization, Azure Kubernetes Service, virtual desktop infrastructure, high‑performance Microsoft SQL Server.
HPE GreenLake leverages predefined, tested, optimized, and validated cloud modules for Azure Stack HCI, delivering fast time to value to our customers. The solution is designed to deliver reliable, solid performance and high availability. The solution offers various configurable options in the form of cloud module instances.
HPE GreenLake cloud module features
- Processing hardware
- Compute instances
- Network module
- Integrated system management plane
- Platform software: Azure Stack HCI OS, Windows Admin Center (WAC), HPE OneView, HPE Aruba Networking Fabric Composer
Cloud module instance options
| Compute instance code | S2i.N25 | S2i.N51d | S2i.N102d | S2i.N102q | S2i.Z153d | S2iv.Z61d |
|---|---|---|---|---|---|---|
| Storage type | NVMe | NVMe | NVMe | NVMe | SSD | SSD |
| Raw capacity | 25.6 TB | 51.2 TB | 102.4 TB | 102.4 TB | 153 TB | 61 TB |
| Memory | 384 GB | 768 GB | 768 GB | 1536 GB | 768 GB | 1024 GB |
| CPU | 2 x 24 Core Gold Ice Lak | 2 x 24 Core Gold Ice Lak | 2 x 24 Core Gold Ice Lak | 2 x 24 Core Gold Ice Lak | 2 x 24 Core Gold Ice Lak | 2 x 24 Core Gold Ice Lak |
| GPU | - | - | - | - | - | 2x NVIDIA A40 |
HPE metering
To measure infrastructure consumption, HPE GreenLake for Microsoft Azure Stack HCI solution offers one of two units of measure
- Allocated memory GiB: The amount of memory allocated by the hypervisor hosts for use by virtual machines (VMs) (allocated memory) as well as the memory consumed by the hypervisors themselves (hypervisor overhead).
- Metering tool reports in any 24-hour period the maximum amount of memory allocated on each hypervisor host summed with the maximum amount of memory overhead consumed by the HA environment. As the amount of memory overhead required increases with the number of VMs configured, it is necessary to include the overhead in charged usage to enable consistent capacity planning and forecasting.
- As the metering script includes the overhead as usage, the installed capacity is reported as the total amount of physical memory. As the metering script includes the overhead as usage, the installed capacity is reported as the total amount of physical memory.
- HPE collects memory metrics from every 15 minutes to once an hour over a 24-hour reporting period.
- The average of daily peak allocated, and associated memory overhead is invoiced each month.
- Storage raw GiB: The amount of raw storage needed to store customer written data and provide data redundancy. The level of overhead associated with data redundancy is a function of the redundancy level configured. Two-way mirroring will incur a storage efficiency of 50% (for example, for every 1 TB of data, at least 2 TB of storage capacity will be needed) while three-way mirroring will incur a storage efficiency of 33% (for example, for every 1 TB of data, at least 3 TB of storage capacity will be needed).
- Installed capacity is reported as the total amount of raw capacity.
- The maximum usable capacity is the total raw capacity minus the formatted capacity of either 2 or 4 times the disk size used. Dependent on the number of nodes present in the configuration.
- HPE collects storage metrics daily.
- Monthly billing will be based on the average of the daily values recorded in each sample.
Microsoft will also measure the use of Azure services and report this information to HPE when HPE is the CSP of record. HPE then bills the customer for this usage in addition to the HPE hardware measures of allocated memory or allocated storage
Summary
- Offered as a single rack solution for data center deployment
- Minimum two (three recommended for production deployment) instances ordered that will be shipped with rack / PDUs / network modules and a management plane
- Maximum 16 compute instances with homogeneous configurations are supported
- Add node option up to maximum 16 nodes (increments of one or more)
