Title :
An Optimal Data Center Availability and Investment Trade-Offs
Author :
Wiboonrat, Montri
Author_Institution :
Grad. Sch. of Inf. Technol., Assumption Univ., Bangkok
Abstract :
Business operation downtimes adversely affect not only costs of recovery times and opportunity losses, but also costs of companypsilas reputation and customerspsila confidence. Reliability of power distribution system for data centers depend on numerous of electronic and mechanical components. Redundancy represents a possible approach to enhance system reliability. In a series-parallel design methodology, serial systems reduce reliability; while more systems in parallel help in increasing it. The redundancy scheme improves the overall system reliability even though it comes with double investments. The data center reliability and cost trade-offs have become controversial issues among all concerned, including investors, IT managers, and consultants. Different business activities are subjected to system failures or power outages which can be lead to vary degrees of risk. A question arises on ldquoWhat represents the most appropriate level of data center reliability at the most affordable cost? This research employs the reliability block diagrams (RBD) with reliability information obtained from IEEE 493, or the so-called Gold Book, and component vendorspsila field test data. The investigation attempts to simulate and compare reliabilities and costs between two data centers (DC); TIA 942psilas Tier III and Tier IV power distribution system standards applying to real cases. Research findings propose an optimal balance between data center system reliability and investment costs for each case.
Keywords :
information centres; power distribution; redundancy; reliability; business activities; data center reliability; data centers; investment trade-off; optimal data center availability; power distribution system; redundancy scheme; reliability block diagrams; series-parallel design methodology; system reliability; Availability; Books; Costs; Design methodology; Gold; Investments; Power distribution; Power system management; Power system reliability; Redundancy; Mean Time Between Failures; Mean Time To Repair; System Availability;
Conference_Titel :
Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing, 2008. SNPD '08. Ninth ACIS International Conference on
Conference_Location :
Phuket
Print_ISBN :
978-0-7695-3263-9
DOI :
10.1109/SNPD.2008.121