|
Dwight Koop is cofounder and chief operating officer for CohesiveFT. His experience spans enterprise IT and entrepreneurial startups. Dwight was global head of data center operations and security for Swiss Banks capital markets, and O'Connor and Associates. He was one of the founders and an EVP of the Chicago Board Options Exchange during its early and rapid growth years. As COO of Bedouin, Inc, he was instrumental in its acquisition by Borland, and as a VP at Borland he played a significant role in its acquisition of Starbase. He was also COO of Signet Assurance, where he is proud to say his engineering team consisted of Eric Hughes, the noted cryptographer, and Bram Cohen, the founder of BitTorrent. Mr. Koop is also the Managing Member of Leporidae Holdings LLC, a private asset management company. Leporidae recently sold its interest in Rabbit Technologies Limited to VMWare.
|
|
"Managing Clouds Using Business Economic Models"
The MCUBEM Consortium Research Initiative
CohesiveFT and a number of leading cloud sector companies have joined forces with several Universities to advance the state-of-the-art of decision support for the cloud service spot and forward markets. The MCubeM Consortium's research is focusing on three types of Cloud economic modeling, all of which help both Cloud Users and Cloud Providers make informed decisions about economically efficient service pricing, purchasing decisions, and configuration of Cloud resource offerings.
The three types of models are:
- Local Cloud Optimization Models - Support Services to help Cloud Providers and Cloud Users organizing their Cloud infrastructure with respect to optimizing cost, utility, and profit by applying economic approaches.
- Unilateral Profit Maximization Models - Support Services to help Cloud providers and Cloud users maximizing their profits and utility, respectively, when interacting with each other.
- Public Cloud Trading Strategy Models - Support Services help Cloud providers understanding their business strategies in the context of the Cloud market (e.g. collaboration through Cloud usage information exchange, market movement through competition between providers).
There are two business-related issues for Clouds. First, the management of Clouds is challenging, since it not only requires the technical skills but also the understanding of the business side of these systems. Traditional distributed systems research did not require this understanding, simply assuming that all components involved are willing to cooperate according to the resource allocation algorithm prescribed. This cooperation assumption is not valid anymore if Clouds are commercial.
Secondly, since demand of Cloud customers can cause spikes anytime, a real-time view of the provider business with respect to revenue streams and costs becomes essential for Cloud providers in order to be able to respond in an economically efficient way. From the user perspective, the understanding of the different offerings of Clouds is also becoming none-trivial, requiring a support in selecting the best-fitting Cloud service.
A solution to these two issues is the use of business economics, which is a part of economic theory that focuses on enterprises. The objective of the MCUBEM Consortium is to design and implement decision support for managing Clouds, and for the use of Clouds. by using business economic models.
"Techno-Revolutionaries of 2010 - Their CLOUD Tactics and Techniques"
CLOUDS = Overwhelming Capacity to Engage + First Mover Advantage. Until now, the edge in Network Centric Warfare (NCW) has been on the side of those who build and maintain capital intensive data centers and the supporting infrastructure. Public CLOUD computing removes this competitive advantage - flattening the playing field. Today, a small group of independently-acting individuals can utilize massive commercial data center capabilities, high bandwidth networks, and on-demand hourly fees while hiding in plain sight in the fabric of the global Public CLOUD Infrastructure. All at minimal cost, and with low risk of detection alongside the business clients of the CLOUD vendors.
HOW:
1. Prepare task specific virtual server template configurations
2. Stage these templates into mission specific virtual clusters
3. Swarm the clusters across a fabric of CLOUD vendors
4. Dynamically redeploy the clusters to allude infiltration, capture, or forensics
This paper examines the tactics and techniques for conducting warfare using and in the CLOUD. Actual experience from building, deploying, and managing virtual stacks, and the role of the author's Elastic Server® Platform webservice are presented, including::
- CLOUD attack vectors
- Leveraging CLOUDS to build overwhelming capacity
- First move advantage - who has it now
- Diversification of CLOUD concentration risk
[If you do not have access to this paper, please contact Dwight directly to request a copy.]
"Methods for Leverage Cloud Computing with Assured Control & Security"
Cloud computing delivers rapid infrastructure scalability for on-demand deployments, cross data center integration, emergency backup, and survivability - all with usage-based pricing model. Cloud computing is rapidly being adopted in the commercial sector, HOWEVER, the gating factor preventing use in the special operation forces community is Cloud Control & Security.
This paper, co-authored with Peter Rung from IDRank Security, examines military applications of Cloud computing, and discusses use of CohesiveFT's VPN-Cubed® technology to create virtual Overlay Networks that are secure from each network end-point to each network end-point - not just - firewall to firewall security. Solutions, including:
- Encrypted communications between all virtual servers within a hyper-cloud.
- Intrusion detection and prevention.
- Extrusion detection and prevention.
- Service discovery and election, in non-multicast environments.
- Federated management of all security elements between clouds.
The paper explains techniques for deploying security as part of a cloud infrastructure to eliminate the risks of exposure to unauthorized personnel.
[If you do not have access to this paper, please contact Dwight directly to request a copy.]
|
|
|
