Medea: Scheduling of Long Running Applications in Shared Production Clusters
- Panagiotis Garefalakis ,
- Konstantinos Karanasos ,
- Peter Pietzuch ,
- Arun Suresh ,
- Sriram Rao
European Conference on Computer Systems (EuroSys) |
The rise in popularity of machine learning, streaming, and latency-sensitive online applications in shared production clusters has raised new challenges for cluster schedulers. To optimize their performance and resilience, these applications require precise control of their placements, by means of complex constraints, e.g., to collocate or separate their long-running containers across groups of nodes. In the presence of these applications, the cluster scheduler must attain global optimization objectives, such as maximizing the number of deployed applications or minimizing the violated constraints and the resource fragmentation, but without affecting the scheduling latency of short-running containers.
We present Medea, a new cluster scheduler designed for the placement of long- and short-running containers. Medea introduces powerful placement constraints with formal semantics to capture interactions among containers within and across applications. It follows a novel two-scheduler design: (i) for long-running containers, it applies an optimization-based approach that accounts for constraints and global objectives; (ii) for short-running containers, it uses a traditional task-based scheduler for low placement latency. Evaluated on a 400-node cluster, our implementation of Medea on Apache Hadoop YARN achieves placement of long-running applications with significant performance and resilience benefits compared to state-of-the-art schedulers.