Top IT/Storage Investment Priorities for 2010

 Here is the opinion of Hubert Yoshida, Vice President and Chief Technology Officer of Hitachi Data Systems on top IT/storage investment priorities for 2010.
 
While global economic recovery is still uncertain looking ahead to 2010, the growth rate of storage and the demand on storage infrastructures continue to intensify. Therefore, IT departments are tasked with staying ahead of the exponential growth of data in the enterprise. IT professionals are operating their businesses with very limited existing capital, making it difficult for them to acquire new technologies. Instead of continuing to add more storage systems, capital or operational investments must be maximized to the fullest extent and be more cost-effective and sustainable.
 
Below are some of the most critical investment priorities for IT professionals in 2010 according to Hitachi Data Systems:
 
Data Center Virtualization: Virtualization is a critical enabler of the dynamic data center of tomorrow. Data center virtualization presents many benefits, but also presents new challenges. Overall power consumption will be lower, but highly variable. There will be fewer servers, but each one will be more critical and will place a higher demand on storage resources than ever before. Applications can be dynamically reallocated at will, and the support infrastructure must be able to do the same.  The data center footprint will be smaller, but overall efficiency might still be suboptimal. The good news is that there are practical and affordable ways to address these challenges and improve data center efficiency in the process.
 
As more workloads are virtualized onto multi-core processors and network bandwidths increase to 8 Gbps on FC and 10 Gbps on Ethernet, the strain on the underlying storage infrastructure increases from both a capacity and a performance standpoint. It’s important that the storage infrastructure supporting the virtualized environment can scale up and out depending on performance and capacity requirements. A scale up storage system that can add more processing power, access ports, cache, and disk spindles for wide striping performance to meet the peak requirements of large server virtualizations. Scale out is required to lower the cost and disruption of adding new capacity or upgrading to new technologies.   In 2009, we have seen the trend towards lower cost modular storage that scales out, through loosely loose coupling around a switch technology like Ethernet or RapidIO. The shortcoming of scale-out modular storage is the inability to scale up. In 2010, we expect to see demand for storage systems that both have the ability to scale-up as well as scale-out out in order to meet the increasing demands of faster networks, processors, and virtual operating systems such as, VMware and Hyper V. 
 
Cloud Storage: Cloud computing is often used as a metaphor for the Internet. Cloud storage serves to mask the complexity of the IT infrastructure and enables access to storage capacity as ‘a pay as you grow’ service. The cloud will continue to grow in awareness into 2010, with a continued focus by private cloud builders and public cloud service providers on elasticity, reliability, multi-tenancy and security. Key to the success of cloud providers, as with the dot.com services providers of earlier days, will be the ability to leverage their resources and be more efficient in managing the growth of storage compared to their end users. Storage virtualization and the ability to dynamically scale up and scale out will be key drivers to adoption. We expect an increasing adoption of cloud storage as advancements are made on key capabilities such as, security, multi-tenancy, and payment models.
 
Automated Tiered Storage: As the difference in price and performance widens, with the introduction of Flash drives and TB SATA disks, the adoption of tiering becomes more attractive. The addition of automated tiered storage management eliminates the enormous staff overhead associated with moving and copying data minimizes the cost of expensive disk drives, leverages lower cost storage for inactive data and copies, and optimizes the use of high performance Tier 1 storage. Many IT organizations are moving towards tiered storage management automation that is policy-based to achieve economic efficiency while maintaining service level objectives for the business. Combining automated tiered storage management with storage virtualization and dynamic, thin provisioning will provide the greatest reduction in capital and operational costs. 
 
Managed Services: Managing data storage infrastructure requires experienced and costly full time staff.  At one time, a pool of in-house resources was available to manage these routine functions, but not anymore. In today’s IT organizations, provisioning for new apps or SAN performance reporting are now being managed through remote managed services. As this demand grows, expect new service offerings to be adopted as an efficient and cost effective alternative for fully managed, 24/7 real time storage expertise where resource skill sets gap or head count constraints exist. Utilizing services also eliminates the need to pull experienced staff away from revenue generating projects for critical, yet administrative tasks.
 
Sustainable IT: IT agendas are already full, especially given the added burdens imposed by the downturn. But executives and managers should be wary of using this as an excuse to delay engaging with top management on the sustainability initiatives. Sustainable IT practices offer significant opportunities for creating value and competitive advantage. 2010 will see greater rationalization of green IT projects. IT managers face internal competition for limited investment available, and will find that green IT projects are gaining more support from corporate programs. There will also be a new role within IT organizations – the Sustainable IT manager – responsible for identifying and managing Green IT programs. In addition, data center managers are likely to start putting greater focus on power consumption. With closer scrutiny on energy measurements, more IT managers will ensure the appropriate procedures are in place for documenting power consumption. At a technical level, this will result in increased focus on server, storage, and data center virtualization as well as cloud.
 
Security: IT managers must strike a balance between mitigating security risks and delivering the best infrastructures in terms of throughput, availability, scalability, cost and complexity. Each organization must make its own trade-off decisions based on its unique situation (e.g., deployed infrastructure, legal and regulatory requirements, and due care expectations) and the importance of its data. IT managers planning storage investments or use of third party services in 2010 will need to take into account key priorities including data confidentiality, privacy, sanitization/eradication and security.
 
Ethernet Convergence in the Data Center (FCOE/DCIB): The industry is moving to a dynamic infrastructure networking model that has highly utilized servers running many VMs per server and uses high bandwidth links to communicate with virtual storage and virtual networks. Over time, this technology has the potential to:

• Lower capital expenses through higher server, storage and network utilization and converged fabrics (which means less cabling and consolidation of HBAs and NICs.)
• Lower operational expenses through higher utilization of data center real estate
• Lower power consumption.

Adoption on the server side with FCoE to the switch will be first as the price for CNA’s and 10 Gbps infrastructure decreases. We expect to see adoption on the storage side of the switch pick up as additional industry standards are formed to enable multi-pathing and network decongestion. There is also a huge investment in Fibre Chanel (FC) infrastructure which will be transitioning to 8 Gbps FC. Since transition to 8 Gbps FC is less disruptive than the transition to 10 Gbps FCoE, the adoption of FCoE on the storage side, we expect may take longer before being widely adopted.
 
Improved Performance with Flash-based Drives or SSD: While flash-based drives are considered cost-prohibitive compared to traditional hard disk drives, they do offer advantages in performance – specifically very low latency and very fast I/Os — and have the added advantage of superior energy efficiency. Expect to see more flash-based technologies integrated into solutions portfolios as customer awareness and potential demand grows, but adopters should look for the added security of data encryption capabilities at the SSD drive level before moving forward.
 
Content Archive Platform Continued Growth: IDC estimates that content data will be the fastest growth area of data with a compounded annual growth rate of 121%. This is not hard to imagine since this is the data that accumulates year after year. In addition to content data, as much as 60 to 80% of structured and unstructured data in a data center may be stale data that could be moved to a content archive to reduce the working set of data that needs to be managed on a daily basis. While content data is the fastest growing data type, it may be the easiest to manage due to its static nature. Once it is ingested in a content archive platform there is no need to back it up as long as you have at least two copies and its life cycle can be managed automatically. When it is ingested it can be indexed so that it can be directly accessed, compressed, de-duped and stored on low cost large capacity disks, while also being encrypted to ensure privacy. Since this is the fastest growing type of data, a content platform must be able to scale to tens of Petabytes and be able to support multiple data formats, and we expect to see growing demand for Content Archive Platforms that meet these requirements.
 
Greater adoption of dynamic (thin) provisioning:
The single greatest tool for reducing operational costs is dynamic (thin) provisioning. Thin provisioning can:

• Eliminate the waste of allocated unused space
• Reduce the cost of moving and copying fat volumes by eliminating unused space
• Reclaim up to 40% or more capacity from existing fat volumes
• Reduce the provisioning of storage from hours to minutes
• Facilitates wide striping to increase performance by spreading the I/O across more disk spindles.

Given all of these advantages, we expect dynamic (thin) provisioning technology to be a top investment priority for customers with capabilities such as, zero page reclaim, dynamic rebalance of storage pools as capacity is added, multi-vendor storage support – while also boosting performance.