Introduction

Because financial services institutions (FSIs) typically run hundreds of different mission critical applications, they are subject to Federal regulations such as Sarbanes-Oxley (SOX), and have been prone to M&A activity, as in the recent acquisition of Bear Stearns by J.P. Morgan Chase. As such, they require high performance IT architectures with high availability. In addition, the success of many financial firms is directly related to the implementation and use of data warehousing, data mining, web portals for on-line banking/trading services and mainframe operations. Data storage and data storage management are critical components of these systems.

This article discusses three key areas associated with data storage and data storage management:

• Storage Tiering

• Data Management Policy

• ‑Storage Provisioning Process and Policy

Focusing on these three areas can benefit any FSI’s data storage environment. Savings may be realized through the efficient use of data storage infrastructure; by putting management controls around the use, creation and storage of data; and through streamlined, accurate storage provisioning process and policy. Standardization, automation and reduced complexity will ultimately contribute to a lower total cost of ownership (TCO) for FSIs.

Storage Tiering

Storage tiering is a concept that stresses putting the most important applications and data on the “appropriate” storage infrastructure. Typically, a rationalization and risk mitigation process will be performed on core enterprise applications and associated data. A Business Impact Analysis (BIA) will assist in rationalizing the Recovery Time Objective (RTO) and Recovery Point Objective (RPO) of applications and associated data. Typically, the required RTO will be the key driver in determining on which storage tier the application’s data will reside. Other drivers will be the data type attributes such as importance level, availability, access and state characteristics (static versus dynamic).

Implementing a tiered storage environment is the process of assigning different categories of data to different types of storage media in order to reduce total storage cost. This is realized as various storage platforms have different footprints, power and cooling requirements. By invoking an enterprise-wide tiered data storage policy, categories may be based on levels of protection needed, performance requirements, frequency of use, and other considerations specific to an FSI.

Table 1 below illustrates a way to organize the parameters for an enterprise tiered storage environment and policy. It divides data storage by estimated cost per storage array and by storage type, storage environment and RTO.

By defining and communicating the tiered storage policy, data can be stored on the most cost-effective storage environment to meet the needs of business. Associating the policy with a valid chargeback model often aids in enforcing the tiered storage policy and communicates to business units the costs associated with storage. In this way cost savings can be realized by:

• ‑Efficient use of expensive storage platforms and resources

• ‑Reduced power and cooling requirements by having fewer high-end storage environments

• ‑Reduced data center footprint (space) by having fewer high-end storage environments

• ‑Improved Total Cost of Ownership (TCO) due to less complex storage environments to support

• ‑Standardized storage environment supports operational efficiency and reduces rework or remediation efforts

A step-by-step storage provisioning framework is represented in table 2 below:

Data Management Policy

Data management encompasses all areas related to managing data as a resource. According to the Data Management Association (DAMA), "Data Resource Management is the development and execution of architectures, policies, practices and procedures that properly manage the full data lifecycle needs of an enterprise." This is a broad definition that does not address a lower level definition of data management.

When speaking of data storage environments and data management policy, the simplest policy changes employed even at the lowest levels can produce significant savings. The following policies can promote the efficient utilization of storage resources while meeting the needs of a financial institution’s business.

• ‑Assign data to the pre-designated storage tier based on pre-determined data characteristics and business requirements, such as RTO and RPO.

• ‑Allocate storage to be 65% to 75% utilized. When data growth exceeds 85% to 90%, extend allocation based on estimated growth.

• ‑Minimize redundant copies of data, unused or accessed data stores, or multiple versions of seldom-used data sets.

• ‑Identify and purge unused, never accessed data unless specifically required to meet compliance or regulatory criteria, or archive onto removeable media.

• ‑Reclaim and reuse data storage space whenever feasible.

As with all data management policies, the goal is to get a minimal data footprint on the lowest tiered data storage (tiers 2 or 3) environment to meet the business objective and reduce associated costs.

Storage Provisioning Process and Policy

While discussing a storage provisioning process and policy, it is important to note that provisioning is a sub-process of Release Management. Storage Release Management involves the planning, designing, configuration, and testing of allocations prior to release into the storage environment, to create a set of release components for changes to the storage infrastructure.

The successful storage provisioning process relies primarily on:

• ‑An accurate workflow of storage request requirements among IT architecture groups, storage requestors (i.e. Crisis Management, Release Management, and Project Managers), storage provisioning resources (i.e. Provisioners and Zonies), database administrators, systems administrators, and related peer-level service providers

• ‑The establishment of clear responsibilities for each role in the process

• ‑Teamwork on behalf of all human resources involved in the provisioning process

• ‑An understanding by all involved that each team has individual objectives and deadlines to meet and will work to meet the common goal of having the appropriate amount and type of storage allocated in the agreed timeframe

The storage provisioning process provides policy guidelines on the requirements and criteria necessary to obtain completion verification of a storage allocation. The policies listed below are examples of storage provisioning policies that would support a storage provisioning process:

• ‑Approval of all changes into the production infrastructure resides with change management

• ‑No action should be performed on storage requests unless funding has been approved

• ‑Provisioning activities should be suspended until all requirements pertinent to the provisioning request are accurate and complete

By designing, communicating and adopting a storage provisioning process and supporting policy an FSI can realize noticeable benefits such as:

• ‑Reduced costs associated with unplanned system outages - Storage provisioning management controls allocations presented to the environment through workflow and resource management in order to properly ensure a stable production environment.

• ‑Reduced costs associated with inefficient resource usage - The storage provisioning process regulates environmental changes in a manner governed by business need, customer requirements and technical considerations.

• ‑Improved manageability - The storage provisioning process provides a structured, repeatable, traceable methodology for efficient and effective change management with continuous, closed-loop process improvement.

• ‑Improved customer satisfaction - The storage provisioning process establishes expected results as well as acceptance criteria while managing expectations throughout the workflow.

In summary, financial services institutions can realize significant savings by following the three simple steps described above starting with the implementation of a tiered storage environment and the adoption of data management policies that reduce the data footprint, but still service the needs of the business. Lastly, FSIs should implement a structured storage provisioning process that allocates storage in the correct tiered environment, adheres to data management policy, and meets the business objectives.

Mark Bortle is Sr. Manager, Data Center Practice at Acumen Solutions, Inc. He can be reached at
703-600-4038 or via email at mbortle@acumensolutions.com.