A New Era in Midrange Storage IDC Analyst paper
1 like394 views
The document discusses the evolution of midrange storage technology, focusing on IBM's Storwize V7000 as a scalable solution that addresses current market needs. It highlights the importance of scalability, efficiency, and simplicity in next-generation storage systems, emphasizing features like automated storage tiering and support for third-party systems. The document also outlines challenges for IBM as it transitions customers from its existing DS series to the new architecture while competing in a mature market.
A New Era in Midrange Storage IDC Analyst paper
- 1. WHITE P APER A New Era in Midrange Storage Sponsored by: IBM Benjamin Woo December 2010 IDC OPINION Storage has evolved from a peripheral to the foundational technology in IT. Data (and www.idc.com therefore, the protection of that data) underpins the internet, Cloud and internal datacenters worldwide. Having storage systems that allow users to scale up in performance, scale out in F.508.935.4015 capacity, and scale deep to leverage existing investments are the key tenets on which next generation datacenters are founded. IBM, in its announcement of the Storwize V7000 has taken all these into account. The V7000 meets all three scaling needs, is feature packed. It is also a clear example of P.508.872.8200 how leveraging industry standards can yield high return for both IBM and its customers. SITUATION OVERVIEW Global Headquarters: 5 Speen Street Framingham, MA 01701 USA An opportunistic market IDC research shows that external storage systems in price bands five through eight (representing storage systems that have an average selling price between US$25,001 and US$250,000) have a compounded annual growth rate (CAGR) of 7.3% between 2009 and 2014. These price bands are often considered by the industry as representing midrange storage systems. This represents a CAGR that is more than 20% more aggressive than the external storage market, and nearly a 45% improvement on the total storage market in the same period. Contrasting this positive news, in the highest two price bands that represent those external storage systems costing more than US$250,000, that market is expected to show a 2.6% decline over the same period. With these forecasts in hand, it is obvious that a focus on the midrange storage market will yield the most opportunities for any storage vendor.
- 2. FIGURE 1 2009 IBM Marketshare, Midrange Systems for External FC Storage Systems Source: IDC, 2007 In 2009, IBM held the leadership in terms of vendor share (with 19.5% of the market) for midrange storage systems as defined above for external fibre channel storage systems. For the last several years, IBM has served this market with a plethora of offerings – the DS3000, DS4000 and DS5000. In the most recent year IBM has been consolidating the number of offerings in this space. But what IBM needed was a new architecture that can parallel and scale with the data and information management messaging it has been pursuing throughout 2009 and 2010. What a next generation midrange storage system needs Next generation storage systems are built on three critical premises: Scalability Efficiency Simplicity Each of these elements help to define whether a storage system will meet the ever- growing and ever-changing demands of end-users. Scalability Scalability can be measured in three different ways: Scale Up 2 #225882 ©2010 IDC
- 3. Scale Out Scale Wide Scaling Up Traditionally, storage systems have ridden atop Moore's Law. As processors improve in performance and capabilities, storage systems vendors have simply leveraged these improvements in subsequent generations of storage systems or in larger versions of existing storage systems. The benefit of scaling up is that as systems scale, there are few interruptions to operations, and in many cases, components within a storage system can be repurposed in subsequent generations or upgraded versions of a given family of storage system. The challenge however, is that scale-up storage systems have typically been proprietary, particularly in monolithic storage systems. Historically, even in modular midrange systems, proprietary components and the software which drives the storage system is typically written to operate with a fixed number (usually two) controllers in mind, and as such limits the potential scalability of these storage systems Scaling Out In more recent years, there has been a tremendous focus on scaling out. The concept is simple, rather than leveraging proprietary hardware, rewrite the software code to take advantage of many industry standard server platforms and leverage the collective performance of these relatively inexpensive, yet (potentially) infinitely scalable hardware platforms. Scale out systems can also benefit from scaling up. As newer generations of industry standard servers are released, the benefits of these servers can be integrated with existing platforms to create a larger, more powerful storage server pool while protecting the investment made in existing storage server nodes. Scaling Wide While scale out architectures can assist in protecting investment in storage server and storage capacity elements of an existing storage system in the same family, it cannot provide the same investment protection to other third party storage systems (even if they are from the same vendor). This is where scaling wide has benefits. The ability to integrate third party storage systems behind a scale-out architecture (i.e. scaling wide), now provides all the benefits of the relatively simple, low-cost approach to scaling storage systems while being able to leverage external third party storage systems to extend the life of existing storage investments and/or to take advantage of specialized cost or features. Efficiency The most critical features and functions for storage systems today address storage efficiency. These technologies include: ©2010 IDC #225882 3
- 4. Automated storage tiering Space efficient backups (snapshots) On-site and Off-site Data Replication Transparent migration Automated storage tiering One of the major challenges as it relates to the efficient use of storage media is the need to put the right data (typically the most accessed) on the right price/performance data storage medium. Solid state storage technologies can help. Solid State Disks (SSDs) that leverages NAND Flash technologies to provide ultra-high performance, when used to target performance sensitive data should be deployed to address these needs. In most end- user environments, investment in ultra-high performance capacity does not need to exceed 10% of the storage capacity. The balance of the capacity should leverage serial-attached SCSI drives (SAS). SAS combines the cost effectiveness of serial-ATA (SATA) drives with the enterprise features, such as dual interface ports, of fibre channel (FC). With this infrastructure in place, the biggest challenge for end-users is ensuring that the right data is in the right tier of storage. The most effective approach is to let the storage system make that determination based on the actual I/O that it is reading and writing. As such, having automated storage tiering becomes critical in terms of meeting performance service levels while minimizing cost. Space efficient backups (snapshots) Managing the cost of storage capacity is not limited to primary, in-use data. Backups and snapshots also need to be efficient. Particularly where an end-user leverages its storage system in a mixed-use environment, being able to make space efficient backups for testing or for data recovery can end up being a big portion of a data storage investment. Additionally, these capacity efficient backups must also be fully integrated into mission critical applications – not only to eliminate backup windows but to ensure the integrity of the data during backups. On-site and off-site replication The ability to provide on-site and off-site replication is critical to the protection of digital assets. Storage systems need to incorporate the ability to create an almost instant copy of active data, which can be used for backup purposes or for parallel processing activities. To improve efficiencies, storage systems should also incorporate the ability to support incremental copies, copying only the portions of the source or target volume that have been updated. 4 #225882 ©2010 IDC
- 5. The combination of thin provisioning technologies along with efficient copy technologies results in end-users using only a fraction of the amount of storage needed for a full physical copy to improve overall storage utilization. These functions are only relevant and useful, if they can be managed within an overall management platform, preferably from a single pane of glass to maximize productivity. When creating data protection policies that go beyond a single storage system, it is necessary to leverage systems that can create remote mirroring capabilities between systems at different locations to help create copies of data for use in the event of a catastrophic event at a data center. Given the prevalence of VMware, these functions should ideally be designed to support VMware vCenter Site Recovery Manage to help speed disaster recovery. Transparent Migration Data migration, is perhaps the riskiest activity that goes on in a datacenter, and is one of the most common causes of planned downtime. The ability to create transparent migrations between existing platforms and future platforms will mitigate these risks. These functions should ideally occur while maintaining access to the data. The data migration function might be used, for example, when replacing older storage with newer storage, as part of load balancing work or when moving data in a tiered storage infrastructure. There are economic benefits to transparent migration functions that can can provide efficiency and business value. This can speed time-to-value from weeks or months to days, minimize downtime for migration, eliminate the cost of add-on migration tools, and may help avoid penalties and additional maintenance charges for lease extensions. Simplicity Given the number of data intensive functions with which storage systems are involved, from RAID calculations to automated data tiering, the management of storage systems can get quite involved. Nonetheless, with the focus of end-users being placed squarely on the delivery of service levels to support business applications, a simple management interface is critical. Ideally, the management interface can be integrated with an enterprise storage management platform, and ultimately with a data center wide management platform. Overall, storage systems need to provide as much reactive information as well as proactive recommended actions to the storage administrators. While some storage administrators desire to see intricate details regarding what a storage system is doing, others simply want to know that the prescribed tasks are completed. Finding this balance can sometimes be a difficult challenge for storage management software. But it is a necessary one. ©2010 IDC #225882 5
- 6. FUTURE OUTLOOK IBM is addressing the next generation midrange disk storage systems need with its Storwize V7000. The Storwize V7000 has taken appealing aspects of its XIV and DS8000 storage systems and SVC gateway and combined it into a single solution. With the introduction of the Storwize V7000, IBM will enable its customers to start from a minimal number of disk drives and ultimately scale up to nearly 240TB of storage. To fulfill its "scale wide" promise, IBM has also designed the Storwize V7000 to attach other external storage systems to extend the capacity even further. To maximize the density of the Storwize V7000, IBM is leveraging a consistent compact 2U form factor. It has also included a number of premium features that will be included with each Storwize V7000. These include: IBM's EasyTier automated data tiering FlashCopy Nondisruptive data movement Thin Provisioning Leveraging the success of the XIV storage system, IBM has adapted the XIV GUI to the Storwize V7000. The Storwize V7000 can also be managed by IBM's Tivoli Storage Productivity Center and IBM's Systems Director Storage Control. On the backend, IBM is one of the first leading storage vendors to leverage the Storage Bridge Bay (SBB) Working Group's SBB 2.0 architecture. The use of SBB by IBM makes a lot of sense. An SBB platform is essentially an x86 industry standard server blade plugged into storage shelf. This means that IBM can accelerate development and deployment of advanced storage services via software. It will also decrease the development and testing costs associated with bringing next generation features and functions to market. One such use may be to leverage its recently acquired compression technology to improve efficiency and reduce costs even further.. IBM could also develop (or migrate) almost any relevant x86-based application to the Storwize V7000. In essence, IBM could conceivably develop the system as a storage application platform. Overall the Storwize V7000 represents a highly scalable platform for IBM customers. Finally, IBM is able to deliver a storage platform that can help it address its small and medium business (SMB) customers through to its high end medium sized business. The consistency in interface between the Storwize V7000 and XIV enables IBM customers to seamlessly evolve to the larger platform. 6 #225882 ©2010 IDC
- 7. CHALLENGES/OPPORTUNITIES Despite the efficiency and consolidation efforts made by IBM in the last 18 months, the DS series of block based storage solutions has been too fragmented and confusing. The Storwize V7000 and XIV define truly a new generation and approach from IBM. It also demonstrates IBM's commitment to R&D, and reflects IBM's innovation in storage. With that said, the market into which the Storwize V7000 is being deployed is already very mature with many well regarded competitors. The Storwize V7000 is a departure from its DS family of storage solutions, and IBM will need to take care not to isolate those existing customers while evolving those customers to the new platform. Secondly, IBM may need to change the way it markets the N series to take advantage of the Storwize V7000. It may benefit IBM greatly to leverage the gateway solution from its partner NetApp in front of a Storwize V7000. With the exponential growth of fixed content and unstructured data, the Storwize V7000 should allows potential customers to leverage the Storwize V7000 and then virtualize existing storage investment as an archive. It behooves IBM to demonstrate to its customers the many use cases that the Storwize V7000 can be deployed. It needs to also advise its customers about the short to midterm roadmap for the Storwize V7000. Customers will want to understand where the Storwize V7000 will scale and at what time so they can understand when an XIV solution may be more suitable. The IBM midrange storage systems have been due for a significant refresh. The Storwize V7000 is an appropriate response, and builds the right platform for IBM to once again be a technology leader in the storage market. Sponsored by IBM Copyright Notice External Publication of IDC Information and Data — Any IDC information that is to be used in advertising, press releases, or promotional materials requires prior written approval from the appropriate IDC Vice President or Country Manager. A draft of the proposed document should accompany any such request. IDC reserves the right to deny approval of external usage for any reason. Copyright 2010 IDC. Reproduction without written permission is completely forbidden. TSL03019USEN-00 ©2010 IDC #225882 7