gary0416 / hdfs-browser

Fast desktop client for Hadoop Distributed File System

Changelog

Fixed

• directly close MainWindow will not close fileSystem.

Added

• copy path to clipboard
• hotkeys

hotkey	function
ctrl+shift+c	Copy path to clipboard
ctrl+u	Copy from local(upload)
ctrl+d	Copy to local(download)
ctrl+x	Cut
ctrl+c	Copy
ctrl+v	Past
ctrl+n	Create directory
delete	Move to trash
shift+delete	Delete permanently
F5	Refresh current view
ctrl+g	Go to directory
F2	rename

• add rename button
• duplicate connection
• count folder and file numbers
• support chinese

Changed

• UI
• allow edit connection without rename
• auto select user home dir in DirectoryChooser

FSBrowser

FSBrowser is a desktop GUI client for Hadoop Distributed File System. The application can be used as a more efficient alternative to Hue's file browser, especially on IO heavy operations, since it is based on native hadoop libraries instead of HttpFS.

Features

• Sending files and directories:
  • Between local file system and HDFS
  • Between different HDFS clusters (DistCp)
• Browsing, searching, sorting directories with milions of files
• Support for creating and browsing Hadoop Archives (HARs)
• GUI interface for hadoop tools:
  • Distributed copy (DistCp)
  • File System Check (fsck)
• Utils for cleaning temporary and staging directories
• Multiple active connections (tabs)

Installation

1. Maven and JDK8 are the only compile-time prerequisites. Build application with mvn package.
2. Run java -jar target/fsbrowser.jar.

Note that some actions require additional native hadoop libraries (for example DistCp or Hadoop Archive tools). While FSBrowser will run without them, it's recommended to have them installed. For Windows, you can download precompiled binary package here. After extracting, set HADOOP_HOME environment variable to the extracted directory and add bin subdirectory to your PATH variable.

Acknowledgments

• This project uses Fugue Icons by Yusuke Kamiyamane