What is Western Digital’s Zoned Storage Initiative?


Data volumes created by enterprises, machines, and consumer-generated content continue to drive demand for data center storage capacities at multi-petabyte to multi-exabyte levels. To date, current SSD and HDD technologies have been keeping up with these large data storage demands.

Zoned Storage is an open-source, standards-based initiative to enable data centers to scale efficiently for the zettabyte storage capacity era. It involves the ability to store and retrieve information using shingled magnetic recording (SMR) in hard disk drives (HDDs) to increase the storage density and its companion technology called Zoned Name Spaces in solid state drives (SSDs).

Zoned Name Spaces is an extension of the NVMe standard. This technology allows us to offer SSDs that are zoned similarly to HDDs, thereby allowing all storage, whether it be HDD or SSD, to be seen as though it were one technology. The Zoned Storage Initiative offers the ability to leverage both SMR and ZNS technologies, open source standards, and purpose-built computing in a disaggregated environment.

The Zoned Storage Initiative offers the ability to leverage both SMR and ZNS technologies, open source standards, and purpose-built computing in a disaggregated environment.

zoned_storage_zns_western_digital
Can the next hundreds of exabytes of data be stored on existing storage technology at the best possible TCO? Dare we think about a zettabyte of storage capacity and beyond?

It has become clear that traditional data center architectures will be increasingly unable to meet the requirements of storing enormous amounts of data at a reasonable TCO. To manage data centers of the zettabyte scale, existing SSD and HDD technologies need an innovative approach for more efficient storage capacity without sacrificing performance.

Innovating for a Data-Centric World

Featured Video

Architecting Zoned Storage for the Zettabyte Age


Western Digital's CTO, Martin Fink, talks with Sr. Vice President Chris Bergey about architecting data centers with purpose-built data storage technologies for economies of scale: shingled magnetic recording (SMR) for hard disk drives and its companion technology for solid state drives, Zoned Namespaces (ZNS).

Press Releases

Western Digital Champions Zoned Storage, an Open Initiative to Redefine Data Centers for Zettabyte Scale

June 11, 2019

With a focus on enabling purpose-built, open and scalable data center architectures, Western Digital today launched Zoned Storage, an initiative that brings together new innovations and industry standards for cloud...

zoned-storage-video-2-western-digital

Video

Ceph as Part of the Data Infrastructure for Zoned Storage

May 23, 2019

Zoned Block Devices is a recently introduced new category of storage devices introduced to address the needs of large-scale data infrastructures. In this presentation we will describe Zone Block Devices and how Ceph...

zoned-storage-video-1

Video

Characteristics of Zoned Namespaces

May 15, 2019

Matias Bjoerling from Western Digital talks about the characteristics of Zoned Namespaces (ZNS) and open-channel solid state drive (SSD) architectures in this speaking session at OCP in San Jose, CA.

What is SMR?


SMR is an acronym for “Shingled Magnetic Recording,” an important technology utilized to increase capacity and enable lower cost per TB in hard disk drives.

It is primarily used in HDDs that populate the world’s largest cloud data centers. As the amount of data created continues to grow at an ever-increasing pace, cloud service providers are seeking ways to drive down costs to enable new applications and provide the benefits of cloud storage to simplify our lives and increase the value of the services they provide. Connected devices provide the opportunity to capture and share more rich data in the form of photos, video, and audio.

Hard disk drives represent the most cost-effective technology to store data, but the challenge is to continually increase capacity and the amount of data that can be stored on each disk, or “platter” in the drive. Increasing drive capacity and storage density provides multi-dimensional benefits. The cost per TB in the device is reduced along with the cost of the corresponding infrastructure in the data center, including real estate, racks, servers and networking equipment.

So how does SMR contribute to this cost reduction? Every platter in a disk drive stores data in a track format. To increase capacity, those tracks have to be written narrower and narrower, while maintaining some space or buffer between discreet tracks. This buffer has to be managed in the device to maintain data integrity, even though the physical environment may be challenging due to vibration created by the cooling fans in the system. SMR circumvents this challenge by writing the tracks in an “overlapping” fashion, much like the shingles on a roof; hence, the name “shingled magnetic recording.”

SMR significantly increases the capacity that can be stored on every disk, improving what we call “areal density” and lowering costs as described above. In a given generation of HDD technology, this could result in approximately 15% additional capacity with cost reduction at both the device and system level.

zoned_storage_smr_western_digital

HDDs have historically used CMR, or “conventional magnetic recording,” without shingled tracks nor the requirement to manage sequential zones. CMR drives support completely random read and write operation.

Increasingly though, cloud service providers are investing in host software and file systems that can manage the complexities of SMR to deliver cost savings to their customers at scale. This is a trend that is fundamentally re-shaping the cloud data center and will continue to deliver lower cost and higher value data far into the future.

To learn more about SMR technology, you can find a detailed White Paper here
Developers can find development libraries and tools here: https://zonedstorage.io

Featured articles from our blog

{{getMergeArray(feed.title, '')}}

{{getEllipsis(getUnEscapedHTMLContent(getMergeArray(feed.description)), 180)}}

Read More