anshuiisc/Aadhar-uidaiBenchmark

Benchmarking Fast Data Systems for the Aadhaar Biometric Database [WBDB-2015]

Benchmarking Fast Data Systems for the Aadhaar Biometric Database

Machine Configuration for Benchmarking

• Apache Storm v0.9.4
• CentOS 7
• OpenJDK v1.7
• 1 node of commodity cluster for running Zookeeper, Nimbus and UI
• 9 nodes of commodity cluster for executing the topologies(supervisors)

About Code

Storm Topology

• bolts directory contains different bolts used in aadhaar enrollment topology
  
• spout directory has AadhaarSpout which feed packets to topology
  
• config directory contains global configuration files used by topology
  
• topology directory has Enrollment topology
  

Event generation and logging framework

• Used to create events from given input file according to timestamps values and logs the data generated during execution of topology for future analysis.

How to Run code

• Go to bin directory and run following command:
  "./runUidaiTopology.sh experimentID" where experimentID is a number of experiment e.g. 100, 101 etc.
• You can change different parameters used in runUidaiTopology.sh file.
• By default 15 minutes events are given, if you want to create 24 hours events then either make changes in runUidaiTopology.sh file or run "python /path/to/.../data/enroll/scripts/eventGen.py" on terminal.

Please refer the paper for detailed info - http://arxiv.org/pdf/1510.04160v2.pdf at [WBDB-2015] [WBDB-2015]: http://clds.sdsc.edu/wbdb2015.in