This article originally appeared on SearchDataManagement.com.
Building and managing a successful database infrastructure requires a combination of the right people, processes and products.
On the people front, having the
There is no standard or ratio for database management staffing since organizations have different business and operational requirements. However, there are several common factors that affect the skill and number of staff required. These include:
- Breadth of services provided: Are you just monitoring your database operations, or are you conducting performance management, administration, and backup and recovery of your databases?
- Maintenance routines: Is this work automated and set on a regular schedule, or is it performed manually and one off?
- Change management windows: When and how is time set aside to perform the maintenance databases require?
- Geographical dispersion: Does staff have to be spread out across several locations or can you centralize?
- Hours of operation (e.g., 7x24, 5x24): Are there any times when an outage would not impact business operations?
- Complexity of the technological environment: Is your environment client/server or is there n-tier infrastructure complexity with application servers and web services?
- Number of managed databases, types of databases, and database growth: Do you manage the components of your database infrastructure individually or as an aggregate?
- Number and sophistication of customers using application: Do your customers rely on you for setup, query and reporting from the databases managed?
- Skill set, experience and training of database management staff: Is your staff updating its skills to handle growing complexity?
Running and managing a successful database infrastructure requires processes and procedures that measure and gauge the staff's ability to meet business and operational objectives. A Database Administrator's (DBA) ability to quickly and effectively manage change is critical to business application availability, making formal administration change management processes for database a necessity. For any database change, upgrade, or patch, there should be testing and validation prior to deployment, accompanied by a risk/benefit review. To prevent unforeseen complications, calculate the worst-case failure scenarios and establish viable fallback schemes.
As to staffing considerations, there are several common factors within organizations that affect the number and level of expertise for best practice build-out. Requirements that determine the level of staffing and effort include:
- Frequency and amount of database changes occurring within the environment
- Number of databases managed and their propensity for growth
- Level of customer service required
- Breadth of services provided – monitor? Or comprehensive database performance management, administration and backup and recovery?
- Complexity of the technological environment
- Frequency of staff turnover
- Type of automation and level of process efficiency (I.E SEI Level 3)
- Degree to which the organization values "standardizing" tasks
Making processes and procedures easy to follow and achieving "buy in" from the staff performing the management is vital to continued success. Well-defined procedures help facilitate determination of whether changes have had the desired effect. Measuring the "overall performance" of the operation through collected metrics enables the DBA to baseline and measure their own performance and effectiveness. This also helps DBA understand their business effectiveness and a measurable path for career growth.
When evaluating the products that manage your database infrastructure, review should start with what is installed, where it resides, and how it is managed in its current state. Assessments should be performed at least quarterly to help track the rate of change and to understand your infrastructure expansion. Periodic baselines give insight into database performance "norms" and lead to predictive and proactive management techniques for events outside of normal acceptance.
Regular reviews also prevent unauthorized augmentation of a database, which can negatively influence the performance or place it at risk. They uncover or prevent unapproved or non-standard practices or tools. DBAs should maintain both physical and logical models of their databases as part of their baseline. Documentation should reflect the current physical and logical state as well as weekly and monthly trends. Poor documentation contributes to delays in problem resolution and reduced service availability. Lastly, substantial changes to the database infrastructure may warrant additional or revised disaster recovery processes.
Whether you are initiating new database projects or simply maintaining your existing database environment, the role of managing databases can quickly become overwhelming. In most organizations today budgets are getting cut, business requirement are escalating, and complexity is increasing as data volumes mushroom. Routinely performing database management best-practice assessments increases the predictability of future project work and enhances return on investment by optimizing the performance of new and existing databases.
These reviews support the successful deployment of enterprise resource planning (ERP), supply chain, and web enabled applications, all of which rely on databases for the storage and exchange of information. Routinely verifying database management practices helps uncover rogue changes, security problems and even helps locate undocumented databases reducing the risks of outages caused from accidental or unauthorized alterations.
In short, best-practice reviews of your people, processes and products reduce the risk of outages or performance delays while increasing overall operational efficiency.
About the author
Steve Lemme is a director of database management solutions for CA, published author of Implementing and Managing Oracle Databases and columnist in Database Trends and Applications Magazine. Mr. Lemme is an Oracle Master DBA with over 15 years experience in mission-critical Oracle architecture and speaks on database management best practices to address regulatory compliance. Prior to CA, he managed critical computer and database systems for Allied Signal, Apple, GTE and Motorola, where downtime was $150,000+/hour. He holds the position of director in the Independent Oracle Users Group.