TIPS & TRICKS

Hunk is a Big Data Platform for Building Applications on Hadoop

Hunk is not only a revolutionary new software product for exploring, analyzing and visualizing data in Hadoop, it’s also a powerful platform for rapidly building applications powered by data stored in Hadoop Distributed File System (HDFS). If you’re a developer, you can build on the Hunk platform using your choice of popular languages, frameworks and tools without having to manually program MapReduce jobs. Hunk enables you to work with data in Hadoop using your existing skills and a variety of standards-based technologies. If you’re familiar with the developer platform for Splunk Enterprise, you know everything you need to know to develop with Hunk.  Taking into account some of the fundamental differences between Hunk and Splunk Enterprise – the data in Hadoop is at rest, so no “real-time” searches – Hunk doesn’t manage data collection, so any functionality related to data ingestion (REST API endpoints, modular inputs) don’t apply – the developer experience is consistent. So what exactly can you do with Hunk?

Build Big Data apps

With Hunk, you can build apps powered by data stored in HDFS  that deliver insights like clickstream analysis, deep customer behavioral modeling and security analysis at enterprise-grade scale. By connecting your data in HDFS to a virtual index in Hunk, you gain access to the capabilities of the Splunk Web Framework (which is integrated into Hunk).

The Splunk Web Framework makes building an app on top of Hadoop look and feel like building any modern web application. You can quickly style and customize your Splunk app using Splunk’s visual dashboard editor or convert your dashboard to HTML with one click for more powerful customization and integration with JavaScript and HTML5. If you’re a web developer familiar with modern web development technologies and model-view patterns, you can easily build apps on Hunk with advanced functionality and capabilities using JavaScript and the popular Django framework.

Integrate and Extend Hunk

Hunk also lets you integrate data from HDFS into other applications and systems across the enterprise, from custom-built mobile reporting apps to Web Parts in Microsoft SharePoint. You can do it easily using the REST API and Software Development Kits (SDKs) for Java, JavaScript, Python, C#, Ruby and PHP. Hunk provides a fully-documented and supported REST API with over 200 endpoints that lets developers programmatically search and visualize data in Hunk from any application. The Splunk SDKs include documentation, code samples, resources and tools to make it faster and more efficient to program against the Splunk REST API using constructs and syntax familiar to developers experienced with Java, Python, JavaScript, PHP, Ruby and C#.

Search and Correlate Data in HDFS

Hunk offers ad hoc exploration, analysis and visualization of historical data at rest in Hadoop. You can dynamically query data in HDFS using the familiar search bar or write a custom search command in a few lines of Python without having to cobble together a bunch of Apache projects and components or set up MapReduce. Hunk utilizes the Splunk Search Processing Language (SPL™), the industry-leading method that lets you interactively explore data across large, diverse data sets. With Hunk’s schema-on-the-fly, you can immediately query and interrogate raw data in Hadoop through visual interactions and SPL for deeper analysis.

The developer platform enables you to expand the search language and write your own custom search commands to perform specific processing or calculations. Let’s say you have semantic customer-related data in HDFS from Twitter, product reviews and other direct feedback. Hunk lets you write your own sentiment analysis command in Python to analyze the data sitting in HDFS.

Also, customers with both Splunk Enterprise and Hunk licenses can search across data stored both in Hadoop and in native indexes in Splunk Enterprise – all in the same search.

Get Started

----------------------------------------------------
Thanks!
Jon Rooney

Splunk
Posted by

Splunk

Join the Discussion