Orna Agmon Ben-Yehuda

Deconstructing Amazon EC2 Spot Instance Pricing

Cloud providers possessing large quantities of spare capacity must either incentivize clients to purchase it or suffer losses. Amazon is the first cloud provider to address this challenge, by allowing clients to bid on spare capacity and by granting resources to bidders while their bids exceed a periodically changing spot price. Amazon publicizes the spot price but does not disclose how it is determined. By analyzing the spot price histories of Amazons EC2 cloud, we reverse engineer how prices are set and construct a model that generates prices consistent with existing price traces. We find that prices are usually not market-driven as sometimes previously assumed. Rather, they are typically generated at random from within a tight price interval via a dynamic hidden reserve price. Our model could help clients make informed bids, cloud providers design profitable systems, and researchers design pricing algorithms.

The rise of RaaS: the resource-as-a-service cloud

In the RaaS cloud, virtual machines trade in fine-grain resources on the fly.

The nonkernel: a kernel designed for the cloud

Infrastructure-as-a-Service (IaaS) cloud computing is causing a fundamental shift in the way computing resources are bought, sold, and used. We foresee a future whereby every CPU cycle, every memory word, and every byte of network bandwidth in the cloud would have a constantly changing market-driven price. We argue that, in such an environment, the underlying resources should be exposed directly to applications without kernel or hypervisor involvement. We propose the nonkernel, an architecture for operating system kernel construction designed for such cloud computing platforms. A nonkernel uses modern architectural support for machine virtualization to securely provide unprivileged user programs with pervasive access to the underlying resources. We motivate the need for the nonkernel, we contrast it against its predecessor the exokernel, and we outline how one could go about building a nonkernel operating system.

Attacks in the Resource-as-a-Service (RaaS) Cloud Context

The Infrastructure-as-a-Service (IaaS) cloud is evolving towards the Resource-as-a-Service (RaaS) cloud: a cloud which requires economic decisions to be taken in real time by automatic agents. Does the economic angle introduce new vulnerabilities? Can old vulnerabilities be exploited on RaaS clouds from different angles? How should RaaS clouds be designed to protect them from attacks? In this survey we analyze relevant literature in view of RaaS cloud mechanisms and propose directions for the design of RaaS clouds.

Collusion in Cloud Computing Auctions

Cloud computing provides flexibility to clients by allowing them to pay per use for the rental of services and VMs. Renting reduces the waste of prepurchased but unutilized hardware. Recently, cloud computing has been moving towards the more economical Resource-as-a-Service model (RaaS) [2]: instead of horizontal scaling (purchasing more VMs), RaaS clouds enable vertical scaling—purchasing more resources (e.g., CPU, RAM, and I/O resources) for a few seconds at a time, at sub-second granularity. For example, CloudSigma charges separately for resources and adjusts prices every few minutes. Amazon Web Services (AWS), Azure and Google Cloud Platform all offer pay-as-you-go pricing. AWS Lambda and Azure Functions allow uploading code and paying for computing time only when the code is triggered to run. RaaS systems use economic mechanisms, such as auctions, to allocate resources [3, 4]. AWS EC2 [1], Alibaba Cloud and Packet spot instances are examples of auctions in horizontal elasticity. Vickrey-Clarke-Groves (VCG) auctions, as used by Facebook to allocate ad spaces, are likely to be used in the near future for vertical elasticity. Ginseng systems are examples of VCG auctions used for RAM [3] or cache [4] allocation. In Ginseng [3], guests run economic agents who bid for resources auctioned by the host. Thus, the mechanism incentivizes selfish, rational agents, who only care about their own profit, to bid with their true valuation of the RAM.