RAPID DATA STREAM APPLICATION DEVELOPMENT FRAMEWORK

• Main Author: ANDRIAN TANUJAYA (NIM : 13513071), WILHELMUS
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/24711
• Institution: Institut Teknologi Bandung

Data stream processing has become a trending phenomenon in recent times. There have been many data stream processing machines circulating around the community. However, application development on top of the data stream processing engine is still considered as slow and there are a lot of rewritings needed in developing different applications. In addition, general data stream processing engine can generally perform distributed computing. Distributed computing requires the program to be compiled before it is run on another machine. As a result, whenever an error occurs, the entire program must be recompiled before it is restarted on distributed machines. <br /> <br /> <br /> Based on these problems, an application framework has been developed to facilitate users in developing data processing applications stream over Apache Flink. This framework can be configured by reading an external file. Through the external file, applications built on the <br /> <br /> <br /> application framework can perform data processing in accordance with the options provided. This application framework can filter the data, aggregate data from different sources, and divide data into the processing window. In addition, the framework also provides a way for users to develop plugins that can be used specifically in the case of processing a specific data stream. <br /> <br /> <br /> By using this application framework, users do not need to write a lot of code to develop data processing applications. Users can also save the number of documents written to develop an application. In creating an application, users simply replace the configuration that is in the configuration file and create a data processing plugin necessary. <br />