Reducing Storage Complexities
As the entrance of the new decade eases in, new technologies being developed by the day are creating voluminous amounts of data that accrue in stale storage systems, gathering dust in the bin.
Research firm IDC predicts in its annual Digital Universe report that data stored in various storage systems are expected to balloon 44-fold in the next ten years, considering the current rate of data growth pegged at 0.8 Zettabytes (800 billion gigabytes) in 2009.
The amount of data growth in recent years--fueled by cascading amounts of corporate data, bolstered by renewed preference for multimedia data, and necessitated by the need to keep and make use of digitized information--is prompting CIOs and IT managers to acquire and deploy a host of technology solutions that answer the need to contain and manage data, all while trying to curb expenditures.
"As customers become more interactive, companies need to have access to information they might have gathered from previous transactions," says Mohammad Saif, deputy director for consulting, ICT Practice, Frost & Sullivan South Asia and Middle East.
This information is compounded by firms' online interactions, as well as the need to keep and have access to records as mandated by law, as in the case of the telecoms sector. "At many places, digitization of records is also creating a big space for storage consumption," Saif adds.
"Data growth is a reality that few businesses can afford to ignore," posits Ravi Rajendran, vice president and general manager, Hitachi Data Systems (HDS) ASEAN. Embattled IT administrators are, therefore, forced to tackle a multi-pronged problem that is inevitable for most organizations.
Faced with the reality of data explosion occurring every year, companies are forced to espouse a keen foresight of data requirements in their data centers. "Because of the way storage is packaged, businesses end up buying up to 75% more capacity than they actually need," Rajendran explains.
For Rajendran, the problem sets in when organizations throw more capacities in an attempt to solve their storage crisis. "The result is a confusing tangle of heterogeneous storage systems and management software, which requires more IT staff expertise, and massively underutilized assets and 'stranded storage,' which are a waste of already shrinking budgets," he points out.
In this complex scenario, the age-old adage of beginning with the end in mind comes to fore. "[An] important area is Information Lifecycle Management (ILM), which can help companies predict storage needs and related costs," suggests Erwin Chuaunsu, country manager for systems and technology group, IBM Philippines.
ILM, according to an IBM whitepaper entitled "The Toxic Terabyte," is a process technology for managing information throughout its lifecycle, based on its intrinsic value to the company and in a way that makes the most efficient use of storage while minimizing the cost of retrieval.
This technology ensures that low-value information are eliminated and transferred to managed permanent storage as soon as possible, in order to make room for mission-critical data to take its place.
ILM offers reprieve from having to manage huge pools of data, but that is assuming most acquired data are structured. "Surprisingly, statistics show a rise in unstructured data (such as documents and media files) rather than structured data (databases)," Chuaunsu shares.
This come-from-behind rise in unstructured data--bolstered by high-bandwidth connectivity which allows for movement of huge amounts of data over the Internet--can be controlled by effectively managing the use of data. "The use of data must be viewed separately," proposes Reco Li, research manager, storage of enterprise hardware systems, domain research group, IDC Asia Pacific.
Li suggests that core business applications with mostly structured data be stored in highperforming disk systems, such as Fiber Channel SANs (Storage Area Networks), in order to be retrieved very quickly. "The unstructured data, and some nonmission-critical data are secured with IP SANs, NAS systems, and DAS (direct-attached storage), which are seeing increased market demands today," he adds.
Gaining access to data conveniently may have more wisdom to it than just mere technical efficiency, according to HDS's Rajendran. "Searching for the right information may be difficult and time-consuming and, in addition, may carry a heavy refinancial impact," he says, adding that having a system in place increases productivity, enables better business decisions, and helps firms achieve greater successes.
THE PROVERBIAL SPRAWL
Yet managing storage systems doesn't only entail zooming in to the granularity of data, but zooming out to handle the hardware as well. As data inevitably grows, storage systems and disks proportionally pile up, usually leading to what is called the "IT Sprawl" where heterogeneous storage systems operate in separate silos, making it difficult to scale when the need arises (and, in the case of data, that need is ever-present).
This is where virtualization comes in. "Technologies such as [virtualization] allow for more efficient use of raw capacity, and minimize the need for frequent capex, as in expansion and upgrades," explains IBM's Chuaunsu.
Tech giant HP, through its Converged Infrastructure thrust, enables firms to scale up to as much as 16 petabytes, by utilizing the largest single namespace product called the X9000. With the solution, firms can leverage non-disruptive and automatic rebalancer of storage loads across all systems.
According to Ong Swee Lye, StorageWorks regional sales manager for Asia Pacific and Japan at HP, common industries seeing increased need for scalable storage systems include financial services, telecom, media and entertainment, oil and gas, and life sciences.
"By pooling together all storage assets and providing a single interface for management, virtualization can help improve storage utilization by reclaiming stranded [capacity]," Rajendran adds.
Cloud storage, on the other hand, proves useful in managing the complex amount of data, at the same time liberalizing the storage portfolio to make it available to the entire spectrum of business. "Cloud storage would make the latest technologies available for all sets of businesses, starting from SOHO to
large enterprises," relates Frost & Sullivan's Saif.
IDC's Li, meanwhile, concedes that the cloud offers the agility needed by modern-day storage systems. "As a business solution-focused approach, many cloud end-users [can enjoy] IT agility and lower cost," he says.
Not all clouds come with a silver lining, however. Rajendran cautions users before jumping on to cloud storage because of eventual management issues. "The problem with this 'one size fits all' approach is that creating another silo for the cloud somewhat defeats the purpose," he argues.
HDS's strategy for cloud storage, he says, allows for customers to define the system based on their needs. "[Other cloud offerings do] not address the more holistic practice of running traditional IT from a single, integrated infrastructure," he clarifies.
MISSION : UPTIME
Virtualization and the cloud are both attractive options, but they answer only part of a multitude of storage issues enterprises face this day. In most large corporations, the remaining portions of the storage debacle are relegated to backup, disaster recovery, and business continuity, a segment of the enterprise controlled largely by compliance to government regulations.
"Businesses are now challenged to interpret a slew of new mandates and protect their key data assets," Rajendran relates. The need to answer to government rules in protection of precious customer data is likewise putting a pressure on CIOs and IT managers to manage their data store efficiently.
At one end, technologies such as tape libraries provide reliable methods of backup storage. "At the core of IBM's backup and base tier disaster recovery offerings is tape storage. IBM continues to push the limits of tape in terms of speed and capacity with products such as the new LTO 5 series of tape drives and libraries, and the TS1130 enterprise tape drive," Chuaunsu notes.
On the other end, an emerging middleman technology is rising in the form of data deduplication--or "dedupe," for short--a data backup mechanism which identifies if certain files being saved have duplicates, and will only save it once in the simplest form of storage, according to Ronnie Latinazo, country manager, EMC Philippines.
"In the past, companies do 1:1 backups," Latinazo stresses. "With deduplication, they can achieve 20:1 to as much as 500:1 backup and compression ratio." By identifying duplicate files--which, Latinazo notes, are prevalent most especially in corporate emails where end-users pass the same file around over and over again, multiplying the need to store multiple files of the same kind--firms eliminate the need to have more storage disks for backup.
Additionally, deduplication offers businesses the capacity to reduce risks, which are commonly attributed to doing backups using tape libraries. "There is a lot of risk about tape. Just transporting the tape to the backup site poses a lot of risks," the EMC executive explains, adding that with dedupe technology, ease of transport is not a problem as the transfer is done entirely online.
Just recently, EMC released what they claim is the largest and fastest inline dedupe system yet, the Data Domain Global Deduplication Array. It offers speed of up to 12.8TB/hour, up to 280TB usable capacity, and up to 270 concurrent wire streams.
IF THE PRICE IS RIGHT
On top of all these concerns, the need to keep costs in check is likewise weighing down on businesses, especially since storage and disk systems don't come cheap at all. "First of all, reducing total cost of IT ownership is becoming crucial," remarks IDC's Li. Acquiring technologies that reduce data complexity is important, but ensuring sustainability of the businesses remains a top concern for most executives.
"Keeping the cost of managing storage infrastructure from escalating is a major problem," adds HDS's Rajendran. The key, according to him, is to have a single window of management, and to promote consolidation, because "no single product can magically solve the challenges."
"A holistic approach--that seamlessly integrates virtualization, thin provisioning, and deduplication--can significantly augment an organization's storage management strategy and accelerate savings," he specifies.