Concerto: Dynamic Processor Scaling for Distributed Data Systems with Replication
نویسندگان
چکیده
A surge of interest in data-intensive computing has led to a drastic increase the demand for data centers. Given this growing popularity, centers are becoming primary contributor increased consumption energy worldwide. To mitigate problem, paper revisits DVFS (Dynamic Voltage Frequency Scaling), well-known technique reduce usage processors, from viewpoint distributed systems. Distributed systems typically adopt replication facility provide high availability and short latency. In type architecture, replicas maintained an asynchronous manner, while master synchronously operates via user requests. Based on relaxation constraint replica, we present novel called Concerto, which intentionally scales down frequency processors operating replicas. This mechanism can achieve considerable savings without user-perceived We implemented Concerto Redis 6.0.1, commercial-level key-value store, demonstrating that all associated performance issues were resolved. prevent delay read queries assigned replicas, offload independent part operation fast-running thread. also empirically demonstrate decreased replica does not cause lag because inherent load unbalance between hides latency replica. Performance evaluations with micro real-world benchmarks show saves 32% average up 51% under various workloads, minor losses Despite numerous studies saving centers, best our knowledge, is first approach considers clock-speed scaling at aggregate level, exploiting heterogeneous constraints across nodes.
منابع مشابه
Dynamic Replication based on Firefly Algorithm in Data Grid
In data grid, using reservation is accepted to provide scheduling and service quality. Users need to have an access to the stored data in geographical environment, which can be solved by using replication, and an action taken to reach certainty. As a result, users are directed toward the nearest version to access information. The most important point is to know in which sites and distributed sy...
متن کاملDynamic Voltage Scaling for Priority-Driven Distributed Real-Time Systems
Energy consumption is increasingly affecting battery life and cooling for computer systems. Dynamic Voltage and frequency Scaling (DVS) has been shown to substantially reduce the amount of power required for uniprocessor and multiprocessor real-time systems that have independent tasks or a statically computed schedule. However, no DVS algorithm has been demonstrated for tasks with precedence co...
متن کاملA Survey of Dynamic Replication Strategies for Improving Response Time in Data Grid Environment
Large-scale data management is a critical problem in a distributed system such as cloud,P2P system, World Wide Web (WWW), and Data Grid. One of the effective solutions is data replicationtechnique, which efficiently reduces the cost of communication and improves the data reliability andresponse time. Various replication methods can be proposed depending on when, where, and howreplicas are gener...
متن کاملData Replication for the Distributed Database using Decision Support Systems
Replication is a topic of interest in the distributed computing, distributed systems, and database communities. Decision support systems became practical with the development of minicomputer, timeshare operating systems and distributed computing. Replicated data may get insufficient due to system failure, fault tolerance, and reliability. A partial Replication is quantized in the replication sy...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Applied sciences
سال: 2021
ISSN: ['2076-3417']
DOI: https://doi.org/10.3390/app11125731