Bucketing for sampling in Hadoop - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is bucketing in Hadoop?

Bucketing is a technique to divide data into fixed number of files or buckets based on a hash function on a column. It helps in organizing data for efficient sampling and querying.

beginner

How does bucketing help in sampling data?

Bucketing allows sampling by selecting specific buckets instead of scanning the whole dataset. This reduces data scanned and speeds up queries.

intermediate

What is the difference between bucketing and partitioning?

Partitioning divides data based on column values into folders, while bucketing divides data into fixed number of files using a hash function. Bucketing is better for sampling and joins.

intermediate

How do you create a bucketed table in Hive?

Use the CLUSTERED BY clause with a column and specify INTO n BUCKETS. For example: CREATE TABLE t (id INT) CLUSTERED BY (id) INTO 4 BUCKETS;

intermediate

Why is bucketing useful for joins in Hadoop?

Bucketing ensures rows with the same key go to the same bucket, enabling efficient map-side joins without shuffling all data.

What does bucketing in Hadoop do?

A Divides data into fixed number of files based on a hash function

B Splits data into folders based on column values

C Compresses data to save space

D Encrypts data for security

Bucketing divides data into fixed number of files (buckets) using a hash function on a column.

Which clause is used to create a bucketed table in Hive?

A CLUSTERED BY

B SORTED BY

C PARTITIONED BY

D DISTRIBUTED BY

The CLUSTERED BY clause specifies bucketing in Hive.

Bucketing helps sampling by:

A Compressing data files

B Selecting specific buckets instead of full data scan

C Encrypting data buckets

D Sorting data within partitions

Sampling can be done by reading only some buckets, reducing data scanned.

What is a key difference between partitioning and bucketing?

A Partitioning divides data into files, bucketing into folders

B Partitioning sorts data, bucketing samples data

C Partitioning divides data into folders, bucketing into files

D Partitioning compresses data, bucketing encrypts data

Partitioning creates folders by column values; bucketing creates fixed number of files using a hash.

Why is bucketing useful for joins?

A It compresses join data

B It sorts data for faster joins

C It encrypts join keys

D It ensures same keys go to same bucket enabling map-side joins

Bucketing places same keys in same bucket, allowing efficient map-side joins.

Explain what bucketing is and how it helps with sampling in Hadoop.

Think about dividing data into smaller parts to read less data.

Describe the difference between partitioning and bucketing and why bucketing is useful for joins.

Focus on how data is organized and used in joins.