- LPU MCA Semester 1
- CAP570
- Unit 6
CAP570 · Unit 6: Advanced databases
Current trends in database technology and big data
10 practice questions · in the ETE syllabus
Key points
- Big data: volume, velocity, variety, plus veracity and value
- Structured, semi-structured and unstructured data
- Hadoop uses HDFS storage and MapReduce processing
- Trends: NoSQL, NewSQL, cloud and in-memory databases
Big data is data too large, fast or varied for traditional tools. Its classic characteristics are the three Vs: volume (amount), velocity (speed of generation and processing) and variety (many formats). Veracity (trustworthiness of data) and value are often added.
| Kind of data | Examples |
|---|---|
| Structured | Rows and columns in relational tables |
| Semi-structured | JSON, XML (tags or keys, flexible schema) |
| Unstructured | Photos, audio, video, free text |
A single server can only scale up so far, so big data is spread over clusters of ordinary machines. Hadoop stores data in HDFS and processes it with MapReduce: the map step emits (key, value) pairs, such as (word, 1), and the reduce step combines values per key, for example summing counts for a word count.
Current database trends include NoSQL and NewSQL systems, cloud databases (database as a service), in-memory databases that keep data in RAM for speed, graph databases for relationships, and data lakes that store raw data of every kind.
Exam trap: velocity is about speed, volume about size, variety about formats and veracity about trust. Mixing formats shows variety, not volume.
Try two questions
A company stores customer tables, social media posts and CCTV video together. Which big data characteristic does this mixture of formats illustrate?
- Volume
- Velocity
- Veracity
- Variety
Show the answer
D. Variety means data arrives in many formats at once.
Why not A: Volume is about the amount of data.
Why not B: Velocity is about the speed of arrival.
Why not C: Veracity is about the trustworthiness of the data.
Photos, audio recordings and videos are usually classified as
- Structured data
- Semi-structured data
- Unstructured data
- Relational data
Show the answer
C. Unstructured data has no predefined data model.
Why not A: Structured data fits rows and columns.
Why not B: Semi-structured data has tags or keys.
Why not D: Relational data is stored in tables.