HPE bulks up hybrid cloud features in its GreenLake as-a-service portfolio

Hewlett Packard Enterprise Co. is expanding its GreenLake portfolio of as-a-service offerings with new application, analytics and developer services that are aimed at helping organizations better manage multiple public and private clouds.

The announcements being made today at the company’s Discover Frankfurt event are led by expanded support for the Kubernetes orchestrator for software containers in the GreenLake for Private Cloud Enterprise pre-packaged private cloud offering that the company introduced early last summer. HPE said it will now support Amazon Web Services Inc.’s Elastic Kubernetes Service Anywhere, which enables customers to create and operate Kubernetes clusters on-premises with optional AWS support.

HPE also said it’s introducing a set of workload-optimized cloud instances that are available on pay-as-you-go pricing. It has expanded its partner ecosystem to include Red Hat Inc.’s OpenShift container platform and VMware Inc.’s virtualization software. The company is also rolling out a set of consumption analytics that are aimed at improving customer insight into usage and cost across the three major hyperscale cloud vendors as well as their local infrastructure.

AWS Kubernetes for the private cloud

GreenLake for Private Cloud Enterprise is a fully managed private cloud that can run both cloud-native and traditional applications. It provides a single global namespace for both public and private cloud services as well as self-service deployment of bare metal, virtual machines and container workloads from a common pool of infrastructure.

Support for Amazon EKS Anywhere lets customers run the same container runtimes on-premises that they use in the public cloud with a consistent experience across the hybrid cloud. Developers can deploy instances from a choice of a central console, via application program interfaces, command-line interfaces or as infrastructure-as-code. They can also select from a variety of operating systems, containerized applications and toolchain integration services. The platform leverages the same automation and continuous integration/continuous pipelines across the public cloud and private infrastructure.

It’s a way to “deploy optimized infrastructure but in a cloudy way,” said Bryan Thompson, vice president of product management for HPE Greenlake cloud services solutions. “Virtual machines are table stakes but we’ve also added container instances and bare metal. It’s delivered not as a bill of materials but with a rate card, so it’s different from what you might expect from a traditional HPE type of solution.”

The new optimized instances cover general-purpose, compute-optimized, memory-optimized and storage-optimized use cases. General purpose instances are aimed at such applications as web servers and data preparation for machine learning. Compute-optimized services target uses likes containers, virtual machines and NoSQL databases. The memory-optimized offering is tuned for in-memory databases and high-speed analytics. The storage-optimized version is meant to help customers manage data lakes and software-defined storage services.

“The infrastructure to support the private cloud is not homogenous,” said Vishal Lall, senior vice president and general manager of HPE Greenlake cloud services solutions. “We built infrastructure types that come in different flavors that enable performance for those particular workloads.”

Cross-cloud analytics

Enhancements to the consumption analytics service include simpler showback reporting and a dashboard customers can use to improve capacity planning and budgeting across a hybrid environment.

“We can now consume inbound data from AWS, [Microsoft Corp.] Azure and [Google LLC’s Google Cloud Platform] and bring that into the same reporting mechanism to report across the entire cloud estate what workloads are consuming whether in the private or public cloud,” Thompson said. “This is a private cloud solution was multi-region support and multi-data center support.”

Administrative features include role-based access controls, quota limits, resource provisioning guidelines, security policies and self-service policies.

HPE is also introducing an early access program for its GreenLake for Data Fabric and Ezmeral Unified Analytics products. The Data Fabric offering consolidates different data types across local, cloud and edge devices. Ezmeral supports open-source tools for data engineering, analytics and data science.

Image: HPE