Storage isn't just a cabinet full of disks blindly attached to a production network. Storage is a resource, and must be allocated and managed as a resource if it is to provide the greatest benefit to a corporation.
Successful data storage management leverages a suite of tools to configure, provision, archive and report storage activities, according to a defined set of management policies or processes. Policies ensure a uniform application of storage resources, and often support other business objectives, such as risk mitigation or legal compliance obligations. Furthermore, the tools and processes must work together -- the best tools are useless if they're applied inconsistently, and the most refined policies cannot be implemented without capable tools.
Data storage management tools
Storage administrators use software utilities and suites to monitor or manage storage resources in the data center. The many management tasks include configuration, migration, provisioning, archiving and storage monitoring/reporting, and so most administrators use a variety of tools. Many of these tools are classified as storage resource management (SRM) tools.
Configuration tools handle the set-up of storage resources. These tools might help organize and manage RAID groups by assigning groups, defining or changing levels, or assigning spare drives. Configuration can also define higher-level activities, such as array failover options.
Provisioning tools define and control access to storage resources -- preventing a network user from being able to see and use any storage. For example, some amount of data center storage may be provisioned for an Oracle database that might only be accessible to a purchasing department.
Migration and archiving tools are used to move data between disks or storage subsystems. Migration tools move data during upgrades or consolidations. For example, data might be migrated from an older server to a new EMC or HDS storage array. As part of a consolidation initiative, the data from several older arrays might be migrated to a new large array before taking the older arrays out of service. Archiving tools move data to secondary or tertiary storage platforms. For example, an archiving tool might move expired data from a Tier 1 storage array to a Tier 2 array or from a Tier 2 array to a Tier 3 storage platform such as a tape library.
Other data storage management tools handle storage monitoring, measurement and reporting tasks. Monitoring tools track storage traffic and subsystem behavior. Administrators can see who is accessing what data. Monitoring tools can also check for storage problems and alert technicians or administrators to problems via pager or e-mail -- shortening response times to storage trouble.
Measurement tools are performance analysis utilities that reveal behavioral information about a storage subsystem. An administrator can use that information for future capacity and upgrade planning. Most monitoring and measurement tools include a reporting feature, which presents detailed information that can be tailored to a range of users. For example, high-level reports for administrators might outline current storage utilization and provide a graph of total utilization versus capacity over time. A low-level report for technicians might list disk failures or suggest potential bottlenecks in the storage infrastructure. Tools are increasingly being used to chargeback users based on departmental storage utilization.
Storage management processes
Even the best data storage management tools rely on rules that govern your data storage processes and procedures. No software can decide which data to migrate or where to move that data -- those decisions are based on your own business needs. Storage management encompasses three areas: change management; performance and capacity planning; and tiering (tiered storage).
Change management is the process used to request, schedule, implement and evaluate adjustments to the storage infrastructure. For example, a new research project may require some amount of storage to be allocated in the data center. The change management process defines the way a request is made and approved, and documents the steps used to configure and provision the requested space on a storage array or server. Change management may also document processes like data migration, defining the formal process used to move data from one part of the storage infrastructure to another while maintaining the integrity and availability of that data for network users.
Storage infrastructures are always growing, but growth must be managed to ensure that storage is available when needed. Performance and capacity planning processes define the means used to measure storage system performance (often to verify storage service level agreements) and examine utilization trends. The result of performance and utilization analysis is then used to make prudent decisions about subsequent storage purchases.
Tiering has emerged as a prime data storage management consideration. The traditional storage paradigm placed all data on expensive, high-performance SAN storage, migrating old or unused data to archival tape. Today's falling disk costs now allow companies to deploy other disk storage systems using inexpensive high-capacity disks, such as SATA or SAS. So high-performance disk storage is reserved for current or performance-sensitive data, while older or infrequently used data is relegated to less expensive disks.
"Purpose-specific" storage platforms have also emerged to address specific business needs. For example, content-addressed storage offers disk-based storage for data that requires rapid recovery but is not frequently accessed. Virtual tape libraries (VTLs) use disk storage to emulate tape systems for greater backup reliability with shorter backup and recovery times [see the SearchStorage.com article on Virtual tape libraries].
Although the availability of various storage tiers has given rise to concepts like information lifecycle management and hierarchical storage management, no tool can decide what data should reside on a given tier. Management processes are required to define (or classify) corporate data based on its relative value or importance and then implement the necessary rules to move data between tiers as needed [see the SearchStorage.com article on data classification].
Corporate data storage is increasingly subject to government regulations that influence data retention and recovery requirements (aka compliance). As a rule, compliance regulations involve four key areas: data retrieval and recovery (discovery); data completeness (legibility); data integrity (authenticity); and third-party review potential (auditability). That is, data must be available for a prescribed length of time, yet remain retrievable for auditors on demand in its complete and unaltered state.
Compliance needs vary with individual regulations and regulations may vary by industry or state. For example, the Sarbanes-Oxley Act applies to any publicly held corporation, while the Health Insurance Protection and Portability Act (HIPPA) relates to companies in the healthcare industry. States such as California and New York implement regulations requiring client disclosure when personally identifiable data is lost or stolen -- federal regulations will likely soon follow. Companies found to be in violation of compliance regulations may be subject to significant fines and other penalties. The goal is to identify the regulations that relate to your business and take steps to ensure that the regulatory requirements are met through your data management policies or procedures.