Taking a holistic approach to evaluating systems and policies is the key to recognizing where storage expenditures should be directed. It is not uncommon for an IT manager today to point to one storage array or identify one policy and characterize it as the single biggest problem within the organization. This mistake is often made when taking an event focus rather than a system wide focus and leads to ad-hoc spending.
In today's changing landscape where the value of IT expenditures is constantly being called into question, managers must deliver measurable improvements and real world returns when focusing on data storage. Recent opinion has also downplayed the importance of total cost of ownership (TCO) and return on investment (ROI) analysis in making these decisions. To calibrate and justify expenditures, today's IT manager can call on proven methodologies found outside traditional technology management.
Prioritizing storage expenditures has become a challenging activity, but it can be achieved with some elements common to a Total Quality Management (TQM) approach. Whereas TQM is more commonly associated with the manufacturing world, a focus on tangible benefits within the intangible world of data storage can make a positive impact.
The following three steps are required for prioritizing storage expenditures:
1. Develop measurable and prioritized goals, whether financial or operational.
2. Evaluate current policies, procedures and systems for "best of breed" and "worst of breed" solutions.
3. Establish where and why capital expenditures will have the most impact.
Spending to extinguish "fires" without real analysis may be acceptable in the short term, but will miss some key "root cause" issues. Target goals can be developed through methodologies such as process maturity evaluation or even a tool such as a Pareto chart. Both approaches attempt to classify the relative importance of a specific goal within the context of a defined environment, in this case a data storage infrastructure. Financial goals can establish cost savings or ROI metrics while operational goals establish availability or service level agreements.
Within any given storage infrastructure there are "best of breed" and "worst of breed" systems and policies. Taken individually these systems and policies may meet an organization's requirements. For example, the combination of a "best of breed" storage subsystem with "worst of breed" replication software could meet technical requirements. However, "worst of breed" replication software and "worst of breed" disaster recovery policy could create a significant negative impact larger than the sum of the individual parts would indicate.
For example, let's analyze the following scenario: Upon identifying inadequate restoration times within a disaster recovery system, a common reaction would be to direct storage expenditures to new replication software, networking infrastructure or even new storage subsystems. Is this additional capital addressing the single biggest flaw in the system, or masking it behind more complexity? In this case, legacy storage policies could be the real issue, and targeting these policies would allow an organization to commit resources to deliver the most impact.
- Simon Lloyd is a storage consultant for GlassHouse Technologies, Inc. Simon is responsible for delivering storage strategy engagements focusing on process methodology, best practices, disaster recovery and architecture.