Islami Bank PLC, Quality Assurance (QA) Engineer, Exam: 14.03.2025 (BUET)প্রশ্ন ও সমাধান

Limitations of traditional DBMS (RDBMS such as MySQL, PostgreSQL, Oracle):

• Scalability: Primarily scale vertically (bigger server); horizontal scaling (sharding across many servers) is hard and costly.
• Rigid schema: Requires a predefined schema; evolving it later is disruptive to running applications.
• Performance at scale: Complex joins/transactions slow down as data volume grows; not built for unstructured/semi-structured data.
• Poor fit for big data / real-time workloads: Struggles with high-velocity streams (IoT, logs, social feeds).
• Cost and complexity: Licensing and skilled DBA/administration overhead can be significant.

How NewSQL relates to SQL and NoSQL:
NewSQL is a class of database systems designed to give the horizontal scalability of NoSQL while keeping the ACID guarantees and relational model of SQL — it sits between the two rather than replacing either.

Feature	SQL (RDBMS)	NoSQL	NewSQL
Data model	Relational (tables)	Document, Key-Value, Graph, Column-family	Relational (tables)
Schema	Fixed	Schema-less / flexible	Fixed, but more flexible tooling
Scalability	Vertical	Horizontal	Horizontal
Consistency	Strong (ACID)	Often eventual (BASE)	Strong (ACID)
Typical use case	Banking, ERP, transactional apps	Big data, content stores, real-time feeds	High-throughput transactional apps needing scale (e.g. Google Spanner, CockroachDB)

Exam takeaway: SQL = consistency-first, NoSQL = scale-first (often at the cost of strict consistency), NewSQL = tries to deliver both simultaneously.