作者: Ishai Menache , Peter Bodik , Gregory M. Foxman , N. M. Mosharaf Kabir Chowdhury , Peter Winkler
DOI:
关键词: Latency (engineering) 、 Response time 、 Fault tolerance 、 Fault management 、 Computer network 、 Network management station 、 Engineering 、 Bandwidth management
摘要: Upon receiving a request to improve one or more conditions of datacenter network, fault management system may analyze information the network including communication patterns among services provided in network. The determines logical machines associated with be moved from devices other select these for migration based on cost function factors tolerance, bandwidth usage, number moves and/or response time latency. tolerance without significantly affecting usage