Follow Datanami:
December 3, 2020

Lenovo Looks to Unify Data Management


A set of data management building blocks unveiled this week by Lenovo’s Data Center Group seeks to unify data from edge to cloud via new data management software, an all-flash storage array that supports object storage as well as AI-driven storage management and a data fabric geared toward data analytics.

The data management suite is among the latest offered by infrastructure vendors who are targeting the growing volumes of data at the network edge with the promise of applying data analytics to achieve business goals.

“The biggest thing customers are trying to do is a new way of…accelerating [and] modernizing their IT,” said Kamran Amini, general manager of Lenovo’s server, storage and software-defined infrastructure.

“What we’re seeing happening is customers are either moving data to the edge or going ahead and modernizing their basic, core data centers” to support private and hybrid clouds, Amini added. The next step is utilizing the data generated by this infrastructure to “get some kind of business outcome.”

Lenovo’s Kamran Amini

Meanwhile, Amini said, enterprise consumption models are evolving as more customers run databases and other workloads in the cloud.

Hence, the server and storage specialist (HKSE: 992) (ADR: LNVGY) is focusing its “unified” data management architecture on cloud and edge computing, analytics and AI along with an “as-a-service” consumption model for smart infrastructure.

To that end, Lenovo on Thursday (Dec. 3) released new data management software that will run across its storage portfolio along with new cloud management capabilities and a Fibre Channel switch for storage networking. The data management package includes object storage support for edge applications that tend to be customized for the emerging data storage framework.

The accompanying monitoring software is a cloud-based management platform that uses AI to automate optimization of Lenovo’s ThinkSystem storage systems and servers.

The unified approach responds to the reality that data is moving from the core to the edge and the cloud, noted Stuart McRea, who oversees Lenovo’s storage platforms. “That requires a new approach for managing that data on the edge.” Hence, the IT vendor touts its data management approach as a way to deliver real-time analytics on the growing amounts of data generated at the network edge.

The data management stack also includes integrated Amazon S3 object storage support. Adding object storage “provides data management for all the different protocols that we see customers needing,” McRea said.

The package also includes the DM5100 all-flash array that provides all-NVMe storage. McRea said the upgrade improves storage performance by 45 percent over the previous Lenovo system. The array supports up to 1 petabyte of flash memory and delivers 25 petabytes in scale-out mode.

The flash array “provides lots of density for edge applications as well as core applications, depending on the user environment,” according to McRea.

The cloud- and AI-based monitoring software introduced this week provides predictive analytics for all Lenovo storage platforms. Among the uses cases is predicting storage infrastructure issues before they occur. It also handles storage capacity management, performance metrics and can recommend “preemptive” system software upgrades.

“As a cloud-based management [architecture], this is where we see customers going [so] they can manage all their storage arrays from one location,” said McRea.

Meanwhile, a new Fibre Channel switch is touted as reducing latency for analytics workloads by up to 50 percent.

Lenovo also introduced an AI training reference architecture developed in collaboration with NetApp (NASDAQ: NTAP) and Nvidia (NASDAQ: NVDA) to help novice users apply more analytics at the edge. It also extended its collaboration with SAP (NYSE: SAP) with the rollout of a new edition of its SAP HANA enterprise cloud that can be run on-premise.

Lenovo executives noted that customers are adopting new operational models that account for edge deployments and greater data volumes.

Ultimately, added Amini, “It’s not the data that’s important, it’s what you do with the data and where it resides and the speed of data,” hence the growing enterprise requirement for data management platforms.

Recent items:

Why Databases are Needed on the Edge

The Journey to Effective Data Management in HPC