580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can multi-cloud brokerage architectures optimize resource allocation, service selection, and energy efficiency across heterogeneous cloud providers?
This research theme investigates the design and implementation of cloud broker systems that can arbitrage between multiple cloud providers (multi-cloud or inter-cloud environments) to optimize resource allocation, improve service quality (e.g., QoS, SLA), reduce energy consumption, and handle vendor lock-in issues. It matters because with increasing numbers of cloud providers and data centers, users face challenges in selecting suitable providers and avoiding lock-in, while providers seek to efficiently manage dynamically fluctuating workloads across geographically distributed data centers.
Key finding: This paper identifies key barriers in multi-cloud environments, including lack of standards causing vendor lock-in and customer lock-in, leading to difficulties in cross-provider resource sharing and energy inefficiency. The... Read more
Key finding: This review emphasizes three essential cloud broker tasks—service intermediation, aggregation, and arbitrage—and stresses that practical multi-cloud brokers should support integration of multiple providers to address vendor... Read more
Key finding: The authors propose a tuple space based coordination architecture for multi-cloud spot markets to enable seamless resource brokering across heterogeneous providers. This architecture allows flexible matching of consumer... Read more
Key finding: This work introduces a dynamic brokerage model for resource pricing and reservation that incorporates customers’ historical usage records to promote fairness and trust. By considering personal and social fairness, incentive... Read more
Key finding: This paper presents a two-level cloud broker architecture composed of a master broker and local brokers to facilitate resource discovery, control, and scheduling across multiple cloud providers. By using both standard... Read more
2. What are the system design and architectural approaches for implementing scalable and flexible cloud brokers?
This theme focuses on architectural design principles, distributed brokering service implementations, and middleware mechanisms for cloud service brokers. The objective is to support scalable, concurrent, and multi-user environments that facilitate service selection, negotiation, scheduling, and execution monitoring, enabling agility and efficiency in cloud resource management. This is critical as cloud environments grow complex and require brokers that can scale elastically and provide multi-cloud integration.
Key finding: This study proposes a distributed broker architecture decomposing a global broker into cooperating distributed brokers and agents, which dynamically scale with the cloud infrastructure. The approach addresses scalability... Read more
Key finding: The proposed two-tier broker architecture, with a master broker and local brokers at provider sites, enables efficient search and control over cloud resources while scaling across multiple providers. The design supports... Read more
Key finding: Mandi is introduced as a lightweight, platform-independent market exchange framework enabling flexible trading of computing services between consumers and multiple cloud providers using multiple coexistence negotiation... Read more
Key finding: This work analyzes the design criteria, system activities, and operational workflows of cloud brokering services. It emphasizes the broker’s role in gathering user requirements, analyzing data, and interfacing with cloud... Read more
3. How can cloud brokers improve service reliability and congestion management to meet SLA and QoS requirements?
This theme explores approaches for cloud brokers to enhance cloud service reliability, deal with outages, manage congestion, and enforce service-level agreements (SLAs) effectively. It addresses mechanisms to detect cloud data center failures, dynamically manage loads across multiple providers, prioritize client requests, and ensure trust and security in multi-cloud environments. These broker functions are crucial for sustaining quality and reliability in diverse and dynamic cloud service markets.
Key finding: This study proposes placing brokers at user ends to monitor multiple data centers, improving fault tolerance by selecting alternative data centers in case of outages. Feedback-based data center selection combined with load... Read more
Key finding: A Dynamic Congestion Management (DCM) system is developed that prioritizes high-priority cloud service requests via a forwarding policy inspired by Cisco congestion management techniques. The system enables load-aware request... Read more
Key finding: Beyond energy efficiency and multi-cloud resource selection, this paper stresses the importance of broker-mediated security and trust models to protect against vulnerabilities in multicloud environments. It proposes a... Read more
Key finding: This work investigates the governance role of cloud brokers within enterprise cloud adoption, focusing on SLA management, risk assessment, and operational monitoring. It highlights how brokers can enhance compliance, enforce... Read more
Key finding: The paper presents a brokerage system employing an authenticated index structure (Multi-Merkle B cloud-tree) and verification protocol to ensure clients can verify the completeness and authenticity of cloud service selections... Read more