S3 to HDFS application


This application template demonstrates continuous big data sync from a source to destination while reading blocks of data from a configured S3 bucket. This could be easily utilized and extended by any developer to create a fast, fault tolerant and scalable Big Data Sync or Retention Application to serve business with continuous data.

Required Properties

End user must specify the values for these properties.

Property Type Example Notes
Access Key For S3Input String ACCESS_XXX_KEY_XX_ID AWS credentials access key id for S3
Bucket Name For S3Input String com.example.app.s3 Bucket name for AWS S3 input
Input Directory Or File Path On S3 String /path/to/input Directory path within S3 bucket
Output Directory Path On HDFS String
  • /user/dtuser /output/dir1
  • hdfs://node1.corp1.com /user/dtuser/output
HDFS path (absolute or relative)
Secret Key For S3Input String 8your+own0+AWS0 secret1230+8key8goes0here AWS Secret access key for accessing S3 input.

Advanced Properties (optional)

Property Default Type Notes
Maximum Readers For Dynamic Partitioning 4 int Maximum allowed partitions for Block Reader. Allocating more partitions is for scaling the application by allocating more resources to handle larger dataset
Number Of Blocks Per Window 16 int Limit to control number of blocks emitted to downstream operators.