The proliferation of cloud computing, Artificial Intelligence (AI), and the Internet of Things (IoT) has upended the traditional industries.
The digital wave continues to change the way we live and work and the new applications and economic changes it brings are driving a move away from traditional business models and industrial structures. This is especially true upon the advent of cloud computing, AI, and the IoT.
During digital transformation, enterprises want to streamline all workloads, improve collaboration efficiency upstream and downstream, implement automatic and intelligent workload management, and build an intelligent collaboration platform. Data has become one of their most important assets.
As data volumes increase, enterprises not only need to efficiently and intelligently store and manage data throughout its entire lifecycle, they also need to extract potential value from of massive amounts of data using interactions and big data analytics. Moreover, seamless system interconnections, rapid service Time-to-Market (TTM), and flexible scalability pose demanding requirements on traditional storage architectures in terms of resource utilization, elastic scaling, intelligence, and automation.
Intelligent Transformation of Traditional Storage
Enterprises usually purchase different storage devices for different applications and purchase different devices at different times, and this leads to numerous storage silos. Storage silos greatly complicate O&M and management and cause a costly waste of resources. All of this hinders enterprises’ efforts to digitalize all their workloads.
To effectively support ever-growing workloads, enterprises require unified, transparent monitoring and management of storage from different vendors and dynamic storage resource configuration based on the Service Level Agreement (SLA), in order to simplify heterogeneous storage management, implement intelligent, automatic management of storage resources, and achieve intelligent O&M by monitoring and warning the status of hardware devices. On top of intelligent heterogeneous storage management, enterprises also require global data resource sharing and a storage-based data analytics platform that allows them to unleash value of their data.
Emerging Internet and mobile applications can experience rapid growth and are updated frequently. To keep up with this pace, storage systems must be deployed within hours or days instead of weeks or months, and must be easy to deploy and capable of elastic scalability if they are to meet the needs of enterprises experiencing an explosion of data growth.
As data centers gradually evolve toward the cloud, hybrid cloud environments will become the new norm. Shifting from traditional storage device management to cloud storage service delivery becomes an important assessment factor by which enterprises evaluate storage technologies and solutions.
For all these reasons, enterprises need to formulate phased cloud storage planning, flexibly deploy cloud storage, and implement unified and intelligent management of cloud storage to reliably support enterprise business in a cost-effective manner. An ideal cloud storage system can build open source private clouds, private clouds based on different levels of virtualization, multiple hybrid clouds, as well as interconnection with different public clouds.
What Is Cloud-Ready, AI-Ready Next-Gen Storage?
Do we have next-generation storage that can rise to the coming challenges presented by cloud and AI? Software-Defined Storage (SDS) points the way. According to Gartner, by 2020, 70 to 80 percent of unstructured data will be stored and managed in low-cost SDS environments.
SDS is growing in popularity, but it must be noted that it still lacks a generally accepted or standardized definition. This has led to many different understandings in the industry. Using traditional disks as an example, decoupling of software systems from the hardware systems was the direction the industry endorsed. So then, does this software mark the beginning of push toward SDS? There is no real consensus on this point. The common view is that SDS is limited to scale-out distributed storage that primarily uses general-purpose servers with an independent storage software suite. Using the local storage available on the servers helps build resource pools that are easy to share.
Distributed storage comes in many forms and product varieties, including Hyper-Converged Infrastructure (HCI), Hypervisor-based SDS, Server SAN, distributed file-based storage, and distributed object storage.
Different forms of product modality mean different focuses. Although the methods used may vary, storage systems are designed to satisfy application requirements. Cloud computing is an important driver of sustained rapid development for enterprises. Newer forms of cloud applications emphasize resource sharing, flexibility, and on-demand scale-outs and scale-ins. These same concepts apply to SDS.
Many SDS products and solutions exist in the market to satisfy the various storage requirements of structured, semi-structured, and unstructured data. The uses cases are quite diverse. FusionStorage — Huawei’s distributed storage product — is perfectly geared toward the high scalability, performance, and compatibility requirements of cloud environments.
China Telecom (Zhejiang) constructed the first in-country petabyte-scale SDS resource pool built on Huawei FusionStorage. The solution has also been widely applied at many organizations deploying cloud applications, including China Merchants Bank (CMB), China Mobile (Liaoning), Deutsche Telekom, and Telefonica to name just a handful.
Back in 2013, Huawei started using FusionStorage to support the company’s three large cloud resource pools, which carry critical workloads for R&D, production, operations, and other activities. Deployment on such a large scale helped demonstrate the superior performance and reliability of the FusionStorage offering, and its ability to continuously satisfy mushrooming demands on capacity and performance.
The sheer complexity entailed in cloud applications in an intelligent world places much higher demands on the flexibility, efficiency, and ease of storage system use in the future. Equipping storage systems with the capability to deliver increased data agility across varying format types and improve data retrieval and management in cloud environments is precisely the role Huawei distributed storage was born to fulfill.
