Oddly enough, many data governance initiatives leave out the first line of defense: the database administrator, according to Henry Olson, Embarcadero Technology’s director of product management. Last month, the company unveiled DB PowerStudio® XE3.5, the latest release of its database management and development platform. This version includes a metadata governance platform to involve database administrators in data governance. Olson explains to IT Business Edge’s Loraine Lawson why Embarcadero sees the DBA as essential to any governance effort.
Lawson: How do you know DBA's are usually overlooked in data governance?
Olson: We call on multiple departments. It’s not uncommon to uncover a governance initiative underway in one part of the organization and find out that the DBA team in IT doesn’t know about it. Successful governance efforts don’t ordinarily start in IT, and when IT gets involved, it usually begins with the architecture group.
Lawson: If IT isn't involving DBAs, who do they typically involve in data governance?
Olson: Governance is business-driven, focused on policy and process, and is usually led by those with a stake in the intended outcomes – data quality, security and usability. Many governance initiatives get hung up on the organizational, definitional and budgetary aspects of governance to the detriment of effective execution. These program initiation issues are non-trivial and are similar in nature to the issues faced by the enterprise data warehousing teams of yore.
Success in governance requires a cross-disciplinary effort, cross-organizational concurrence on definitions, shared authority, and staffing and funding sourced from stakeholders across the organization. Although there is some evidence for success with “guerilla governance” efforts, cross-organizational commitments, staffing and funding for data governance usually require a mandate from the senior leadership team. Initiatives that get past the organizational, funding and alignment hurdles must next take inventory.
Developing an accurate picture of the enterprise information landscape requires diverse knowledge and skills. Broadly speaking, the architects understand the structure of the data, the DBAs understand its deployment and implementation, and the business team understands the meaning and importance of the data. Successful governance efforts learn the lay of the land, and to do so must enable effective collaboration amongst a diverse group of technologists, subject matter experts and business stakeholders.
Lawson: Analysts talk about how data should belong to the business, not IT, and therefore should own governance. In light of that, why should DBAs be involved? What's their role?
Olson: Here, we must make a distinction between decision-making authority and execution responsibility. While the business team may “own” the data from an authority and policy standpoint, their interest is in using the data, not managing it.
IT professionals are still required to implement, manage and improve the underlying systems, and DBAs are on the front line with responsibility for database availability, performance, integrity and security. Actually effecting change and implementing the policies agreed by the governance team requires IT alignment and a shared understanding of the data and the policies affecting it.
DBAs, along with other members of the governance implementation team and the business, need to be aware of the meaning, value and sensitivity of the data they’re working with. For example, if a DBA is creating a development snapshot of a production database, they will need to know which data elements are sensitive — for example, PII — and must by policy be masked for non-production use.
If a data analyst needs to dig into an operational system to create an ad-hoc report not supported by the data warehouse, they’ll need to understand the meaning and use of the data in that system in order to quickly create a quality, correct report. Because systems and metadata are subject to change, tools for collaboration and coordination are critical to governance success in complex environments.
Our customers have asked us to deliver capabilities for collaborative authoring and publication of an authoritative catalog of information assets – what we term the “Enterprise Information Map.” With that map at hand, governance teams can stay better aligned and on track from policy definition to system management and optimization. Data governance is better with metadata.