What will be new or different in storage for 2016? We posed this question to EMC, Hitachi Data Systems, IBM, NetApp, and Oracle, and their responses largely echoed what we reported in September: Flash is in, disk is out, and tape’s coming back.
The most direct response came from Hitachi storage boss Hu Yoshida. “While there is a huge install base of disk, the demise of the hard disk is imminent,” he stated.
“Many people will refute this, given the continued existence of tape 60 years after the introduction of the disk drive. I believe that the staying power of tape is due to several factors. Tape is cheaper to acquire, can be stored offline, and the road map for tape densities is still following an aggressive growth curve. By contrast the HDD roadmap has flattened out and the only way to increase densities is to change the packaging by adding more disks and R/W heads, which adds cost and impacts performance,” he said.
Regarding flash memory, “The availability of multi-terabyte flash devices will enable flash to compete with high-performance 15K RPM disk drives on a capacity-cost basis. As a result, the majority of storage systems delivered in 2016 will contain a percentage of flash to boost response times and reduce the cost of managing storage performance,” he stated.
“At the last Flash Summit, Toshiba announced that they would be delivering a 16 TB (solid-state drive) and projected a 128 TB SSD by 2018 through the use of 3D flash technology. This goes way beyond what disk vendors will be able to deliver. Back in August when I blogged about these announcements at the Flash Summit I still thought it was premature to say that the disk drive is dead. But now, four months later, I am ready to join the believers.”
Yoshida expands on these ideas in his blog.
IBM’s Andy Walls, CTO and chief architect, projected his thoughts deeper into the future. “Flash will dominate disk storage by 2019,” he said. Meanwhile, “Cold storage will become a hot technology in 2016. Storage vendors will step up efforts in alternative technologies for cold data storage and how to make storage technology faster, cheaper, and denser. High capacity, low cost, and durability are important goals for cold storage. Presently, data retrieval and response time can be significantly slower for a cold storage system than for devices or systems designed for active data. Media choices for cold storage include tape or low-cost commodity hard disk drives. Object storage… is one of the most common disk-based storage system types for cold data.”
Evangelos Eleftheriou, of IBM Research – Zurich (where they’re working on even more exotic stuff than flash memory), cited International Data Corp.’s label of “Big Data Flash” as part of the disk-killing emerging trend. “This new category of storage will focus on reducing cost and increasing the effective storage capacity to serve cloud workloads that are dominated by reads and do not have stringent requirements on write performance and endurance,” he said. “We expect to see more companies move towards software-defined storage systems made of low-cost flash and commodity hardware components to store large data volumes in a cost-efficient way.
NetApp’s flash-focus is on price cuts. “The first wave of flash adoption was driven by its promise of performance, but used sparingly due to high costs. Until now, flash was reserved for cache and tier technologies or carefully applied to isolated performance applications such as databases… Flash sales are poised to double due to aggressive price cuts as large enterprise storage companies stake out claims on this bright spot in the market,” observed Lee Caswell, vice president of product, solutions, and services marketing.
Of course, not all new technologies pan out. EMC’s John Roese, senior vice-president and CTO, added what he categorized as a cautionary note.
“As any storage veteran can tell you, when new technologies achieve scale there are more often than not unforeseen issues that lead to industry wide endemic failure. I am not saying that this definitely will happen to flash, but it is at the very least a possibility that businesses should consider as the technology starts to achieve real scale in the year ahead,” Roese observed. “The adoption of innovative technology always involves a certain degree of risk management and success often comes down to how well businesses can cope when this technology experiences teething troubles.”
Oracle, which owns the former StorageTek tape business, replied with a statement from the tape side of the disk-squeezing trend.
“Disk has clearly supplanted tape as the medium of choice for rapid data backup and recovery, but this story isn’t over yet. Consider vast cost-reduced data archives, and you’ll appreciate the growing appeal of tape. Yes, we’re all creating zettabytes of information, but not all of it needs to be immediately accessible, does it? Tape is now enjoying a quiet renaissance in situations where long-term preservation and cost efficiency at scale really matter… In 2016 expect modern tape technology to start supplanting disk for cloud-scale archival data repositories.”
The role of storage admins in 2016
NetApp’s Caswell added his thoughts on the people who make the bits work.
“The role of storage admins will continue to evolve in response to transforming IT landscapes. As organizations move to a cloud delivery model to reduce costs and increase flexibility, they shift from being builders and operators of their own data centers to being brokers of services that span both private and public cloud resources,” he said. “In 2016, the classic storage administrator will either evolve into a data manager of the hybrid cloud with a seat at the executive table or hole up in a comfortable storage product minutia and become increasingly less relevant.”