Data Administration vs. Database Administration

Two different jobs that sound alike

The exam reliably tests the contrast between data administration and database administration — they share a word but are different functions.

- Data administration is a *strategic, organizational* function responsible for the overall management of data *resources*: setting enterprise-wide data definitions, policies, and standards; planning the information architecture; resolving data-ownership conflicts between departments; and managing the information repository (metadata). The data administrator is best filled by a *respected senior business manager*, not a deep technical specialist — the role is about politics and cross-departmental cooperation. Think of it as the *controller* of the data, analogous to a financial controller. - Database administration is a *technical, operational* function responsible for physical database design and the day-to-day technical reality: security enforcement, performance tuning, backup and recovery, and database availability. The DBA *implements* the standards that data administration sets and spends most of their time tuning performance and ensuring availability. The modern DBA role is expanding into Hadoop/Spark clusters, cloud vendor management, and data-lake support.

Worked example: who owns which task?

Task                                              Owner
--------------------------------------------------------------
Define the enterprise-wide "customer" standard    Data administration  (strategic / policy)
Mediate a dept dispute over data ownership         Data administration  (organizational)
Maintain the metadata repository                   Data administration  (metadata)
Tune slow queries / choose indexes                 Database administration (technical)
Configure backup, recovery, and journaling         Database administration (technical)
Install/upgrade the DBMS, manage user accounts     Database administration (technical)
Enforce security and ensure 24/7 availability      Database administration (technical)

Rule of thumb: policy, planning, conflicts, and metadata → data administration; everything physical and operational → database administration. When an organization has no separate data-administration function, its responsibilities fall to the DBA and other IT roles.

Open-source DBMSs and availability

An open-source DBMS is free (or nearly free) database software whose source code is public — MySQL and PostgreSQL are the leaders. Advantages: no per-deployment licensing fees, a large community of developers/testers, no vendor lock-in, and source you can inspect. Considerations when choosing one: feature completeness (subqueries, stored procedures, transactions), support and documentation, stability, speed, and the specific open-source *license* terms. The total-cost-of-ownership gap is large (open source can be an order of magnitude cheaper than commercial licenses), which is why startups, education, and non-profits favor it.

A core DBA responsibility is availability. The gold standard is "five nines" (99.999%) uptime — only about *five minutes* of downtime per year. The four threats to availability and their counters: hardware failure → redundant/clustered servers; data loss/corruption → mirroring + backup/recovery; human error (the leading cause of outages) → mature, repeatable procedures and training (COBIT, ITIL); maintenance downtime → nondisruptive utilities that work while the system stays online.