Distributed frank-wolfe under pipelined stale synchronous parallelism

Iterative-convergent algorithms represent an important family of applications in big data analytics. These are typically run on distributed processing frameworks deployed on a cluster of machines. On the other hand, we are witnessing the move towards data center operating systems (OS), where resources are unified by a resource manager and processing frameworks coexist with each other. In this context, different processing framework job tasks can be scheduled on the same machine and slow down a worker (straggler problem). Existing work has shown that an iteration model with relaxed consistency such as the Stale Synchronous Parallel (SSP) model, while still guaranteeing convergence, is able to cope with stragglers. In this paper we propose a model for the integration of the SSP model on a pipelined distributed processing framework. We then apply SSP on a distributed version of the Frank-Wolfe algorithm. We theoretically show its sparsity bounds and convergence under SSP. Finally, we experimentally show that the Frank-Wolfe algorithm applied on LASSO regression under SSP is able to converge faster than its BSP counterpart, especially under load conditions similar to those encountered in a data center OS.