1 Background Information: Big Data Systems Vs Relational Database:Big Data Systems for the Internet of Things (IoT) Karthikeyan Sugumar, Jay Buckler, Suprio Ray Background Information: Project Details: Goal: To Evaluate the write throughput and query performance of the below PostgreSQL Cassandra Hbase Logbase OpenTSDB Internet of Things (IoT) is a proposed development of the Internet in which everyday objects have network connectivity, allowing them to send and receive data. IoT Growth: 50 billion devices by 2020 $8 trillion revenue potential (source: CISCO and DHL) Dataset: Electricity Smart Metering Technology Trials from the Commission for Energy Regulation (CER), Ireland. 160 million records of meter readings from the trial. Data format: meterId, Timestamp, meterReading(kWh) Performance Evaluation factors: Batch load performance Live stream performance TPC-H benchmark queries Data management challenges Velocity + Volume: 100s of millions of updates per second Variety: interoperability Analytics: only 0.5% of the world’s data is actually analyzed Big Data Systems Scalability Availability High Performance Open Source Dynamic control of Data Fault Tolerance Big Data Systems Vs Relational Database: Business Application Considerations: Do you need to keep the application always online and serving customers? Do you need to serve customers with multiple interfaces ad in multiple locations? Do you need to consume and deliver lots of data very quickly? Do you need to easily add database capacity to handle increasing customer demand? Do you need to manage many different types of data (e.g. social media, etc.)? Do you need to easily run analysis on your line of business data? Do you need to easily search your line of business data? Do you need to receive strong payback for IT investments?