A Cluster Based System for Analyzing Ethereum Blockchain Transaction Data

Abstract

Ability to perform fast analysis on massive public blockchain transaction data is needed in various finance applications such as tracing of fraudulent activities. The blockchain data that is synced as a node is accessible as a sequence of blocks containing transactions. This way of accessing transaction data, however, is too slow for applications that require a transaction graph to be constructed. We develop a cluster based system that constructs a distributed transaction graph in parallel. Since blockchain data is continuously growing, our parallel system also offers the advantage of being able to scale by simply increasing the number of nodes in the cluster. Our system has been developed using the MPI message passing interface. We report performance results from our system operating on the whole 9.5 million block (roughly 4 year) Ethereum mainnet blockchain data. We report timings obtained from tests involving distributed transaction graph construction, partitioning, page ranking of addresses, degree distribution and token transaction counting on a 16 node economical cluster set up on the Amazon cloud. In particular, our system is able to construct distributed graph of 658 million ether and 31 major token transactions in 188 seconds.