HBase data model (column families) in Hadoop - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a column family in HBase?

A column family in HBase is a group of columns that are stored together physically. It helps organize data and improves read/write efficiency by grouping related columns.

beginner

How does HBase store data within column families?

HBase stores all columns of a column family together on disk. This means data in the same column family is stored in the same place, making access faster for those columns.

intermediate

Why should you keep the number of column families small in HBase?

Because each column family is stored separately, having many column families can slow down performance and increase storage overhead. It's best to keep column families few and group related columns.

beginner

Explain the relationship between rows, column families, and columns in HBase.

In HBase, data is organized as rows identified by a row key. Each row contains one or more column families. Each column family contains multiple columns. This hierarchy helps organize and access data efficiently.

intermediate

Can you add new columns to an existing column family in HBase without changing the schema?

Yes. HBase allows adding new columns dynamically within an existing column family without changing the table schema. This flexibility is a key feature of HBase's data model.

What does a column family in HBase group together?

A Indexes for faster search

B Rows with similar keys

C Tables with the same schema

D Related columns stored together physically

A column family groups related columns that are stored together physically to improve access speed.

Why is it recommended to keep the number of column families small in HBase?

A Because each column family requires separate storage and too many slow down performance

B Because column families cannot have more than 3 columns

C Because column families are deleted automatically if too many exist

D Because column families are only used for indexing

Each column family is stored separately, so many column families increase storage overhead and reduce performance.

In HBase, what is the smallest unit of data storage?

A Table

B Column family

C Cell (intersection of row and column)

D Region

A cell, which is the intersection of a row and a column, is the smallest unit of data storage in HBase.

Can you add new columns to an existing column family without changing the schema in HBase?

A Yes, columns can be added dynamically within a column family

B No, schema must be updated first

C Only if the table is recreated

D Only during off-peak hours

HBase allows adding new columns dynamically within existing column families without schema changes.

How are column families stored in HBase?

A All column families are stored together in one file

B Each column family is stored separately on disk

C Column families are stored in memory only

D Column families are stored as separate tables

Each column family is stored separately on disk to optimize access and storage.

Describe how data is organized in HBase using rows, column families, and columns.

Think about the hierarchy from row to column family to column.

Explain why it is important to limit the number of column families in an HBase table.

Consider how storage and performance relate to column families.