To 4,000 compute nodes and beyond: Network-aware vertex placement in large-scale graph processing systems

The explosive growth of "big data" is giving rise to a new breed of large scale graph systems, such as Pregel. This poster describes our ongoing work in characterizing and minimizing the communication cost of Bulk Synchronous Parallel (BSP) graph mining systems, like Pregel, when scaling to 4,096 compute nodes. Existing implementations generally assume a fixed communication cost. This is sufficient in small deployments as the BSP programming model (i.e., overlapping computation and communication) masks small variations in the underlying network. In large scale deployments, such variations can dominate the overall runtime characteristics. In this poster, we first quantify the impact of network communication on the total compute time of a Pregel system. We then propose an efficient vertex placement strategy that subsamples highly connected vertices and applies the Reverse Cuthill-McKee (RCM) algorithm to efficiently partition the input graph and place partitions closer to each other based on their expected communication patterns. We finally describe a vertex replication strategy to further reduce communication overhead.