Overview of HPE Alletra
HPE announced in May 2021 that it would be transforming HPE Storage into an SDDS business that implements its Unified DataOps cloud operations model. This change will bring a new experience to cloud users that unifies data operations and addresses the explosive growth in edge-to-cloud computing for many data environments. It will collapse data silos and reduce the complexity of storage platforms, giving them greater agility to meet market changes. This transition of HPE Storage is a critical milestone in HPE’s goal to become a Platform-as-a-Service (PaaS) provider for edge-to-cloud workloads.
Unified DataOps Puts an End to Fragmented Infrastructure
The need for Unified DataOps arises from the complexity of infrastructure found in many organizations, especially enterprises with a long history. Their infrastructure is often a collection of fragmented hardware and software managed by manual processes that have developed over many years. This practice makes it difficult to remain competitive in today’s rapidly changing marketplace.
HPE GreenLake recognizes the need to deliver a consistent experience to cloud and edge users by allowing them to eliminate data silos and leverage their data to provide meaningful insights, regardless of where the data resides. Its new platform will be ideal for organizations developing artificial intelligence (AI) initiatives that are fueled by distributed data. The ability to drive digital transformation and improve customer experiences can benefit many organizations.
HPE recently engaged marketing research firm ESG to conduct a survey on data management to explore how organizations are dealing with the rapidly increasing complexity of their infrastructure. This survey showed that 93% of IT decision-makers said the complexity of storage and data management was impeding the organization’s digital transformation. Furthermore, two-thirds of the respondents believed that data fragmentation across hybrid clouds created business risks.
HPE’s vision of Unified DataOps integrates data-centric policies, cloud-native management, and AI to generate insights into data while eliminating silos. It will also simplify the management of data and infrastructure, which is essential for driving operational agility, promoting innovation, and minimizing business risks.
HPE’s plan to develop an operational model for data storage that’s based entirely in the cloud will change the way organizations look at storage. Bringing data operations to the place where it allows an organization’s digital transformation to proceed much more smoothly. This paradigm change is essential for organizations to deal with the rapidly growing amount of data they must manage. HPE’s original approach to this problem will continue to provide innovative solutions that increase agility and allow users to deliver outstanding customer service, especially for organizations with customers and partners throughout the world.
Cloud Data Services Simplifies Global Infrastructure Management
The cloud data services component of HPE Storage is a suite of software services provided to the user on a subscription basis. It greatly simplifies and automates the management of large-scale global infrastructure. HPE will deliver its data cloud services through a Data Ops Manager that can reach any location and any device through an intent-based approach to provisioning infrastructure. This method uses AI to focus on what the application is trying to do with the data, which facilitates provisioning on-demand and self-service. It also optimizes service level objectives (SLOs) by eliminating the guesswork that normally accompanies the provisioning of infrastructure.
Data Services Cloud Console Provides a Single Interface
Data Services Cloud Console allows an administrator to manage this entire infrastructure from a single interface, providing operational agility. It delivers unified data management and cloud data services based on the same proven technology utilized by Aruba Central. This cloud networking solution serves tens of thousands of customers around the world with millions of devices. Data Services Cloud Console uses a single API that allows customers to drive the automation of applications and custom data services. It also allows developers to access infrastructure as code.
HPE Alletra Introduces Cloud-Optimized Workloads
HPE Alletra is the newest component of HPE’s cloud-native data infrastructure. Data Services Cloud Console securely manages Alletra, so it can deliver an operational experience based in the cloud. Alletra includes a portfolio of systems optimized for specific workloads that consist of the All-NVMe Alletra 6000 and 9000 series of servers. They both offer the architectural flexibility needed to run a variety of applications without compromising their performance.
The Alletra servers are still very new, but the HPE Alletra data sheets are already available. They provide details about the contents of these boxes, including drive capacity and network ports. The most recent information released about HPE Alletra concerns the 6000 PCIe, which is currently in its fourth generation.
The data sheets use different units for drive capacity, with terabytes (TB) for the 6000 and tebibytes (TiB) for the 9000. Both units are 4U in size, meaning they’re about seven inches high.
Although the Alletra 6000 & 9000 seem alike, the two products have unique platforms, which are not interchangeable.
The Alletra 6000 is derived from Nimble, HPE’s predictive all-flash storage technology. It has 24 drive slots with 3.5-inch bays. The individual drives may have capacities of 1.92, 3.84, 7.68 or 15.36 TB. The maximum capacity of the 6000 is therefore about 368 TB. In comparison, the base enclosure for the Nimble has 48 drive bays, which includes the AF20, AF40, AF60 and AF80.
The 6000 uses AMD processors with PCIe gen 4 interfaces and an increased lane count. These improvements make the Alletra up to three times faster than Nimble, according to HPE. The DRAM for the controller increases with the five models in the 6000 series. The DRAM capacity for these models is as follows:
- 6010 64GB
- 6030 128GB
- 6050 256GB
- 6070 448GB
- 6090 896GB
The 6050, 6070 and 6090 models treat storage class memory as cache. The 6090 is a scale-out system with a capacity of four standard 6090 arrays, each of which has two all-flash shelves. The maximum raw storage capacity of a fully scaled system is therefore 3.26 PB. The 6000 series supports triple parity RAID, with 32 Gbit/s Fibre Channel (FC). It may also have a total of 1, 10, 25 and 100 GbitE ports, including active and standby controllers. The number of controllers is divided about evenly between these two controller types.
The Alletra 9000 is based on HPE Primera’s intelligent storage system that combines autonomous operation with high performance and resiliency. The 9000 has 48 drive slots that look thinner than those of the 6000. The 9000 has many more I/O ports in the back than the 6000, including up to 48 network interface ports. The Ethernet ports support speeds of 10 or 25 Gbit/s, while the FC ports supports speeds of 16 or 32 Gbit/s. The arrangement of the cooling fans in the back of the unit is also different from the Alletra 6000.
The 9000 series currently contains the 9060 and 9080, both of which use small form factor (SFF) solid-state drives (SSDs). The raw capacities of these drives may be 1.92, 3.84, 7.68 and 15.36 TB. The maximum storage capacities of both models in the 9000 series are therefore about 656 TiB, or 721.3 TB, assuming 48 15.36 TB drives. The 9000 series supports a variety of drive types, including NVM Express (NVMe) First nondestructive Interactive Partition Splitter (FiPS), NVMe FIPS encrypted and NVMe FIPS encrypted Trade Agreements Act (TAA) drives.
The 9060 has 256GiB (275GB) of cache per controller node, while the 9080 has 768GiB (824.6GB). Both models may have two or four controllers, but the 9080 will be significantly faster for memory-bound operations with either configuration. The RAID 6 data protection of the 9000 is also different from the 6000’s triple parity RAID. The 9000 doesn’t support storage-class memory, according to its latest datasheet.
Impact of Comport Implementing Storage as a Service
In April 2021, a manufacturer with custom ComportSecure managed services awarded Comport a four-year contract to implement GreenLake solutions.
Mike Vencel, president of Comport Consulting said,
“We were able to customize that offer to really reflect specifically what the customer was looking for.”
He added that Comport would probably triple its implementation of GreenLake at that site before the end of the year. Comport’s sales team and customers are gaining confidence in Storage as a Service according to Vencel. He says that Comport’s ComportSecure platform makes HPE’s as-a-Service offering more attractive because Comport uses Managed Services Provider (MSP) customization to deliver a complete platform to the customer.
Comport has been a storage solutions provider for more than 30 years. We specialize in cloud storage and have partnered with HPE, along with other major hardware providers like Cohesity and Nutanix. We use our own in-house team to implement cloud solutions, and team members have certifications in many of the most common storage technologies. These qualifications help ensure that we can select the best solution for your needs. Learn more about how Comport helps teams make informed IT decisions or contact us for a consultation on improving your company’s storage infrastructure.