Data Stream Processing Management in the Cloud
Speaker: Evangelia Kalyvianaki
Date: 13 June 2016 Time: 12:00
Location: G-100
Host: Prof. E. Markatos


As users of “big data” applications expect fresh results, we witness a new breed of stream processing systems (SPS) that are designed to scale to large numbers of cloud-hosted machines. Such systems face new challenges: (i) to benefit from the “pay-as-you-go” model of cloud computing, they must scale out on demand, acquiring additional virtual machines (VMs) and parallelis-ing operators when the workload increases; (ii) failures are common with deployments on hundreds of VMs—systems must be fault-tolerant with fast recovery times, yet low per-machine overheads. An open question is how to achieve these two goals when stream queries include stateful operators, which must be scaled out and recovered without affecting query results.

In this talk I will describe a novel approach to externalise operator state explicitly to the SPS though a set of state management primitives. State externalisation enable us to handle both scale out and recovery from operators’ failures using the same primitives. Our system periodically checkpoints operator state and saves it to upstream VMs. Failed operators are recovered by restoring checkpointed state on a new VM.


Eva is a Lecturer (Assistant Professor) in the Department of Computer Science at City University London. Before this, she was a post-doctoral researcher in the Department of ComputingImperial College London. She holds a Ph.D. from the Computer Laboratory (SRG/netos group) in Cambridge University and an M.Sc. and B.Sc. from the Computer Science Department, University of Crete, Greece. Her interests span the areas of Cloud Computing, Data Stream Processing, Autonomic Computing, Distributed Systems and Systems Research in general.

© Copyright 2007 FOUNDATION FOR RESEARCH & TECHNOLOGY - HELLAS, All rights reserved.