The Hipster Approach for Improving Cloud System Efficiency

Abstract

In 2013, U.S. data centers accounted for 2.2% of the country's total electricity consumption, a gure that is projected to increase rapidly over the next decade. Many important data center workloads in cloud computing are interactive, and they demand strict levels of quality-of-service (QoS) to meet user expectations, making it challenging to optimize power consumption along with increasing performance demands. This paper introduces Hipster, a technique that combines heuristics and reinforcement learning to improve resource e ciency in cloud systems. Hipster explores heterogeneous multi-cores and dynamic voltage and frequency scaling (DVFS) for reducing energy consumption while managing the QoS of the latency-critical workloads. To improve data center utilization and make best usage of the available resources, Hipster can dynamically assign remaining cores to batch workloads without violating the QoS constraints for the latency-critical workloads. We perform experiments using a 64-bit ARM big.LITTLE platform, and show that, compared to prior work, Hipster improves the QoS guarantee for Web-Search from 80% to 96%, and for Memcached from 92% to 99%, while reducing the energy consumption by up to 18%. Hipster is also e ective in learning and adapting automatically to speci c requirements of new incoming workloads just enough to meet the QoS and optimize resource consumption. In this work, we extend our previous work in several ways. First, we present an analysis of the size of the reward lookup table and an optimization for the table to improve the scalability of our reinforcement learning mechanism. Second, we demonstrate Hipster's capability to adapt to changes in the latency-critical application at runtime and still satisfy QoS guarantees of the new incoming applications. Lastly, we present a deployment methodology for setting up new applications managed by Hipster's runtime system.

Twig: Multi-Agent Task Management for Colocated Latency-Critical Cloud Services

“…As systems increase in complexity (hardware and services), observability (more PMCs), and controllability (DVFS settings and core counts), it gets increasingly more expensive and errorprone to develop custom heuristics [12,13,71] with Hipster [15,72], Twig's use of a NN approximator for the state-space mapping means that: (1) Twig learns faster as it uses a NN instead of a Q-table, (2) Twig eliminates the need to explicitly traverse the state-action pairs to understand the quality of an action, (3) Twig reduces the memory usage by not storing the state-action space as a Q-table, (4) Twig understands the environment's state using a set of PMCs rather than a single metric, and (5) Twig can use transfer learning to quickly learn how to manage new services. Moreover, unlike other state-of-the-art approaches [12,13,40,58], Twig's use of PMCs avoids the need for service-specific instrumentation.…”

Leveraging application classes to save power in highly-utilized data centers

“…CPI 2 [47] uses long-term CPI patterns to identify stragglers and antagonists of latency-sensitive tasks. HipsterCo [28] uses reinforcement learning to place tasks on heterogeneous systems.…”