Advice for CIOs: Building a Trustworthy Storage Foundation for Diverse Applications

To protect valuable data assets, professional storage devices must be fast and reliable.

Over the past few years, Huawei OceanStor Dorado all-flash storage has become popular across multiple industries thanks to its high performance and reliability. More than 40 of the world’s top 100 banks have chosen Huawei OceanStor Dorado to go digital.

We are entering the yottabyte data era, where more data applications are being released and more data is being produced than ever before. Against this backdrop, Huawei data storage has built a trustworthy storage foundation to drive digital transformation.

Huawei’s OceanStor Dorado all-flash storage enhances NAS capabilities and provides industry-leading SAN features. Both NAS and SAN are built on the FlashLink algorithms and SmartMatrix architecture. FlashLink algorithms and end-to-end NVMe deliver a latency as low as 0.05 milliseconds, while SmartMatrix architecture enables high tolerance to hardware failures and only 3 seconds of service interruption every year. OceanStor Dorado provides the industry’s only integrated SAN and NAS active-active solution.

OceanStor Pacific Scale-out storage houses 120 disks per 5 U chassis. Such a high-density design combines with the 22+2 high-ratio erasure coding (EC) algorithm to deliver a 20% higher disk density per unit space than similar products. Huge breakthroughs have been made in technical architecture, including data flows adaptive to large and small I/O, converged indexing for unstructured data, ultra-high-density hardware, and EC algorithms. These mean that a single storage unit can make data analytics 30% more efficient by supporting hybrid workloads across high-performance computing (HPC), big data analytics, and AI computing.

OceanProtect backup storage provides up to 155 TB/h backup bandwidth and 172 TB/h recovery bandwidth, beating the next-best player by 3x and 5x, respectively. OceanProtect’s advanced data reduction capability allows you to work with more data for less. Innovative technologies achieve a premium data reduction ratio of up to 72:1 — 20% better than the industry’s next best product.

To protect against ransomware, Huawei has developed a ransomware protection storage solution that combines primary storage and backup storage. We use machine learning models to detect ransomware, with a detection rate of 99%.

FusionCube pre-integrates a distributed storage engine, virtualization software, and cloud management software to support on-demand resource allocation and linear expansion. The Data Management Engine (DME) frees O&M personnel from complicated tasks. On average, DME increases resource efficiency by over 20% and shortens service TTM tenfold, giving personnel more time and resources to focus on what matters most. In addition, DME provides unified management GUIs and standard APIs to centrally manage Huawei and third-party storage devices. Its open APIs, scripts, and plug-ins allow the DME to connect with mainstream cloud management platforms and IT Service Management (ITSM).

The Data Center Virtualization Solution (DCS) combines virtualization software and ICT hardware to build lightweight, flexible, diversified, and open data center infrastructure for enterprises. DCS supports a minimum of two nodes with flexible capacity expansion, supporting large-scale applications. DCS can also add container, DR, backup, and unified management software on demand, reducing costs.

In the future, storage architecture innovations will mainly focus on reshaping hardware and software architectures through a datacentric approach that use various data application acceleration engines. Together, these can improve storage performance by 10 times.

• Currently, data flows across two buses: the Double Data Rate (DDR) memory bus and the PCIe system bus. A global data bus with high throughput allows us to integrate these two buses, greatly improving data flow efficiency.

Figure 1: Reshaping storage architecture

• In the past, data was indexed and cached at each layer, from application and computing to network and storage. In the future, control and data plane separation will greatly reduce data access latency between applications and the storage pool.
• For all applications, including service logic and data processing, we can place processing power near the storage system based on application features. This will enable neardata processing and accelerate applications.

What we suggest

As data storage develops, enterprises aim to provide a consistent experience both on and off the cloud and enable real-time data processing and automatic management. To meet these requirements, we propose the Intelligent Storage solution, which embeds AI into data storage.

Aiming to enable autonomous and intelligent data storage, Intelligent Storage enables us to:

• Support autonomous decision making in scenarios like the layout, scheduling, and reduction of different types of data within storage devices.
• Intelligently allocate CPU and memory resources to overcome local resource bottlenecks.
• Enable automatic root cause analysis for unknown faults across all scenarios.
• Enable data flow and application deployment in a multi-cloud environment.

Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.