Talia Chen participated in the Student Research Competition of SOSP’24, undergraduate category, with a poster of her work, “Scaling GNN Sampling on Large-Scale Graphs with io_uring“. Talia made it to the final round, where she gave a short presentation in front of the judges and SOSP attendees. Congratulations, Talia!
Our paper “CAPSys: Contention-aware task placement for data stream processing” was accepted at EuroSys’25! We performed an empirical evaluation study to show that task placement not only significantly affects streaming query performance but also the convergence and accuracy of auto-scaling controllers. To address this issue, we propose CAPSys, an adaptive resource controller for dataflow stream […]
The latest Apache Flink Kubernetes Operator release includes an autoscaler component based on the OSDI’18 paper “Three steps is all you need: fast, accurate, automatic scaling decisions for distributed streaming dataflows” co-authored by Vasiliki Kalavri and John Liagouris. We are glad to see our research impact and very grateful to the Flink community for seeing […]
Our paper “Learning on streaming graphs with experience replay” has been accepted to appear at the 2022 ACM/SIGAPP Symposium on Applied Computing (SAC’22). This is a collaboration with Massimo Perini (University of Edinburgh), Giorgia Ramponi (ETH Zurich), and Paris Carbone (KTH Royal Institute of Technology). See the preprint pdf here.
Showan’s submission “Toward Workload-Aware State Management in Streaming Systems” was accepted for presentation at the 15th EuroSys Doctoral Workshop (EuroDW 2021). Here’s the abstract: Modern streaming systems rely on persistent KV stores to perform stateful processing on data streams. Although the choice of the state store is crucial for the system’s performance, there has been […]
We virtually visited VMware Research today where John Liagouris gave a talk on Secrecy. See the talk slides here.
In out latest work, we study the problem of composing and optimizing relational query plans under secure multi-party computation (MPC). MPC enables mutually distrusting parties to jointly compute arbitrary functions over private data, while preserving data privacy from each other and from external entities. We have released a new preprint on the arxiv, where we […]