Disk failures in the real world: what does an MTTF of 1,000,000 hours mean to you?
FAST '07 Proceedings of the 5th USENIX conference on File and Storage Technologies
A scalable, commodity data center network architecture
Proceedings of the ACM SIGCOMM 2008 conference on Data communication
The cost of a cloud: research problems in data center networks
ACM SIGCOMM Computer Communication Review
Computer
Algebraic Specifications of Computing as a Service with Applications to Cost Analysis
UCC '12 Proceedings of the 2012 IEEE/ACM Fifth International Conference on Utility and Cloud Computing
Iris: a scalable cloud file system with efficient integrity checks
Proceedings of the 28th Annual Computer Security Applications Conference
An approach for constructing private storage services as a unified fault-tolerant system
Journal of Systems and Software
Client-controlled cryptography-as-a-service in the cloud
ACNS'13 Proceedings of the 11th international conference on Applied Cryptography and Network Security
CorrectDB: SQL engine with practical query authentication
Proceedings of the VLDB Endowment
Hi-index | 0.00 |
In this paper we aim to understand the types of applications for which cloud computing is economically tenable, i.e., for which the cost savings associated with cloud placement outweigh any associated deployment costs. We discover two scenarios. (i) In an "unified client" scenario, where the cloud-hosted applications are only accessed by a single cloud customer (or small set of associates), it is important to ensure that the cost savings (mainly computation-related) can offset the often significant client-cloud distance (network costs etc). Today, from a technological, cost-centric point of view, this includes only compute-intensive applications with at least 1000 CPU cycles per each 32 bits of client-cloud transferred data. Naturally a number of other considerations may make clouds attractive even for less compute intensive tasks (services, security, pay-as-you-go nature etc). (ii) In a "multi-client" setting on the other hand, when outsourced applications serve numerous different third parties, we show that clouds begin to act similarly in nature to content-distribution networks - their better network integration is simply too good to pass on, when compared to locally hosting the applications (and incurring associated network costs). Thus, in multi-client scenarios, today's compute, energy and general technology costs suggest that outsourcing to clouds is profitable for almost any application. Ultimately, we hope this work will constitute a first step in an objective evaluation of the technological side of costs of outsourcing and computing in general.