HDFS command line interface in Hadoop

Introduction

HDFS command line interface lets you manage files in Hadoop's storage system easily using simple commands.

You want to upload files from your computer to Hadoop storage.

You need to check what files are stored in Hadoop.

You want to delete or move files inside Hadoop storage.

You want to see the content of a file stored in Hadoop.

You want to check the space used by files in Hadoop.

Syntax

Hadoop

hdfs dfs -command [options] [path]

# Example commands:
hdfs dfs -ls /path/to/directory
hdfs dfs -put localfile.txt /hdfs/path/
hdfs dfs -cat /hdfs/path/file.txt
hdfs dfs -rm /hdfs/path/file.txt

All commands start with hdfs dfs followed by the operation.

Paths starting with / refer to locations inside HDFS.

Examples

Lists all files and folders inside the /user/hadoop directory in HDFS.

Hadoop

hdfs dfs -ls /user/hadoop

Uploads a file named localfile.txt from your computer to the HDFS directory /user/hadoop/.

Hadoop

hdfs dfs -put localfile.txt /user/hadoop/

Shows the content of file.txt stored in HDFS on the screen.

Hadoop

hdfs dfs -cat /user/hadoop/file.txt

Deletes file.txt from the HDFS directory /user/hadoop/.

Hadoop

hdfs dfs -rm /user/hadoop/file.txt

Sample Program

This script shows how to list files, upload a file, check its content, and delete it using HDFS CLI.

Hadoop

# This is a shell script to demonstrate HDFS CLI commands

# Step 1: List files in root directory (likely empty or default files)
hdfs dfs -ls /

# Step 2: Upload a local file to HDFS
# Assume localfile.txt exists in current directory
hdfs dfs -put localfile.txt /user/hadoop/

# Step 3: List files in /user/hadoop to confirm upload
hdfs dfs -ls /user/hadoop

# Step 4: Show content of uploaded file
hdfs dfs -cat /user/hadoop/localfile.txt

# Step 5: Remove the file from HDFS
hdfs dfs -rm /user/hadoop/localfile.txt

# Step 6: List files again to confirm removal
hdfs dfs -ls /user/hadoop

Important Notes

Time complexity: Commands run quickly but depend on network and cluster size.

Space complexity: Commands themselves use little memory; storage depends on file size.

Common mistake: Forgetting to specify full HDFS path or mixing local and HDFS paths.

Use HDFS CLI for quick file management; use Hadoop APIs for complex data processing.

Summary

HDFS CLI helps manage files in Hadoop storage with simple commands.

Common commands include -ls, -put, -cat, and -rm.

Always use full HDFS paths and check file presence after operations.

Practice

1. What does the HDFS command hdfs dfs -ls /user/hadoop do?

easy

A. Uploads local files to /user/hadoop in HDFS

B. Deletes the directory /user/hadoop from HDFS

C. Lists all files and directories in the HDFS path /user/hadoop

D. Displays the content of files in /user/hadoop

Solution

1. Step 1: Understand the command structure

   The command hdfs dfs -ls is used to list files and directories in HDFS.

2. Step 2: Analyze the given path

   The path /user/hadoop specifies the directory to list.

3. Final Answer:

   Lists all files and directories in the HDFS path /user/hadoop -> Option C

4. Quick Check:

   hdfs dfs -ls lists files [OK]

Hint: Remember: '-ls' always lists files in given HDFS path [OK]

Common Mistakes:

• Confusing '-ls' with '-rm' which deletes files
• Thinking '-ls' uploads files
• Assuming it shows file content

2. Which of the following is the correct syntax to upload a local file data.txt to HDFS directory /user/hadoop?

easy

A. hdfs dfs -put /user/hadoop data.txt

B. hdfs dfs -copyFromLocal /user/hadoop data.txt

C. hdfs dfs -get data.txt /user/hadoop

D. hdfs dfs -put data.txt /user/hadoop

Solution

1. Step 1: Identify the correct command for uploading

   -put uploads local files to HDFS, and the syntax is hdfs dfs -put <local_path> <hdfs_path>.

2. Step 2: Match the syntax with options

   hdfs dfs -put data.txt /user/hadoop correctly places data.txt as local file and /user/hadoop as HDFS destination.

3. Final Answer:

   hdfs dfs -put data.txt /user/hadoop -> Option D

4. Quick Check:

   -put local_file hdfs_path is correct [OK]

Hint: Remember: local file first, then HDFS path in '-put' [OK]

Common Mistakes:

• Swapping local and HDFS paths
• Using '-get' which downloads files
• Using '-copyFromLocal' with wrong argument order

3. What will be the output of the command hdfs dfs -cat /user/hadoop/sample.txt if the file contains:

Hello Hadoop
Welcome to HDFS

medium

A. Lists the file details of sample.txt

B. Prints the content of sample.txt to the console

C. Deletes sample.txt from HDFS

D. Uploads sample.txt to HDFS

Solution

1. Step 1: Understand the '-cat' command

   hdfs dfs -cat prints the content of a file in HDFS to the console.

2. Step 2: Analyze the file content

   The file sample.txt contains two lines: 'Hello Hadoop' and 'Welcome to HDFS'.

3. Final Answer:

   Prints the content of sample.txt to the console -> Option B

4. Quick Check:

   -cat shows file content [OK]

Hint: Use '-cat' to display file content in HDFS [OK]

Common Mistakes:

• Confusing '-cat' with '-ls' which lists files
• Thinking it deletes files
• Assuming it uploads files

4. You run the command hdfs dfs -rm /user/hadoop/data.csv but get an error saying file not found. What is the most likely cause?

medium

A. The file /user/hadoop/data.csv does not exist in HDFS

B. You used the wrong command; '-rm' cannot delete files

C. You need to use '-put' to delete files

D. The HDFS service is down

Solution

1. Step 1: Understand the error message

   The error 'file not found' means the specified file does not exist at the given path.

2. Step 2: Check command correctness

   -rm is the correct command to delete files in HDFS, so the error is not due to wrong command.

3. Final Answer:

   The file /user/hadoop/data.csv does not exist in HDFS -> Option A

4. Quick Check:

   File missing causes 'file not found' error [OK]

Hint: Check file existence before deleting with '-rm' [OK]

Common Mistakes:

• Assuming '-rm' is wrong command
• Trying to delete with '-put'
• Ignoring file path correctness

5. You want to copy all files from local directory /local/data/ to HDFS directory /user/hadoop/data/ and then verify the files are uploaded. Which sequence of commands is correct?

hard

A. hdfs dfs -put /local/data/* /user/hadoop/data/ then hdfs dfs -ls /user/hadoop/data/

B. hdfs dfs -get /local/data/* /user/hadoop/data/ then hdfs dfs -cat /user/hadoop/data/

C. hdfs dfs -rm /local/data/* then hdfs dfs -ls /user/hadoop/data/

D. hdfs dfs -copyToLocal /local/data/* /user/hadoop/data/ then hdfs dfs -rm /user/hadoop/data/

Solution

1. Step 1: Choose the correct upload command

   -put uploads local files to HDFS. Using /local/data/* uploads all files from local directory.

2. Step 2: Verify files in HDFS

   hdfs dfs -ls /user/hadoop/data/ lists files in the HDFS directory to confirm upload.

3. Final Answer:

   hdfs dfs -put /local/data/* /user/hadoop/data/ then hdfs dfs -ls /user/hadoop/data/ -> Option A

4. Quick Check:

   Upload with '-put' then list with '-ls' [OK]

Hint: Upload with '-put' then check with '-ls' [OK]

Common Mistakes:

• Using '-get' which downloads files
• Deleting local files before upload
• Using '-copyToLocal' which copies from HDFS to local