PACMan: Coordinated Memory Caching for Parallel Jobs
نویسندگان
چکیده
▪ Replayed Facebook and Bing workloads ▪ LIFE reduces average completion time by 53% and 51% in Facebook and Bing workloads ▪ Small jobs see 77% improvement ▪ LFU-F improves cluster utilization by 47% and 53% in the Facebook and Bing workloads ▪ LIFE and LFU-F beat Belady’s MIN despite lower cache hit-ratio ▪ Pre-fetch & Pre-replace → Ideal (87%) speedup ▪ Pre-replacement ~ Oracle cache eviction EVALUATION
منابع مشابه
Replication Algorithms in a Remote Caching Architecture
We study the cache performance in a remote caching architecture. The high performance networks in many distributed systems enable a site to access the main memory of other sites in less time than required by local disk access. Remote memory is thus introduced as an additional layer in the memory hierarchy between local memory and disks. Eficient use of remote memory implies that the system cach...
متن کاملArchitectural Support and Mechanisms for Object Caching in Dynamic Multithreaded Computations
High-level parallel programming models supporting dynamic fine-grained threads in a global object space are becoming increasingly popular for expressing irregular applications based on sophisticated adaptive algorithms and pointer-based data structures. However, implementing these multithreaded computations on scalable parallel machines poses significant challenges, particularly with respect to...
متن کاملCoordinating Parallel Processes on Networks of Workstations
The network of workstations (NOW) we consider for scheduling is heterogeneous and nondedicated, where computing power varies among the workstations and local and parallel jobs may interact with each other in execution. An effective NOW scheduling scheme needs sufficient information about system heterogeneity and job interactions. We use the measured power weight of each workstation to quantify ...
متن کاملCoordinating Parallel Processes on Networks of Workstations Running Head: Coordinating Parallel Processes on Nows
The Network of Workstations (NOW) we consider for scheduling is heterogeneous and non-dedicated, where computing power varies among the workstations, and local and parallel jobs may interact with each other in execution. An eeective NOW scheduling scheme needs suucient information about system heterogeneity and job interactions. We use measured power weight of each workstation to quantify the d...
متن کاملSolving the Problem of Scheduling Unrelated Parallel Machines with Limited Access to Jobs
Nowadays, by successful application of on time production concept in other concepts like production management and storage, the need to complete the processing of jobs in their delivery time is considered a key issue in industrial environments. Unrelated parallel machines scheduling is a general mood of classic problems of parallel machines. In some of the applications of unrelated parallel mac...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2012