Mon-Yen Luo

Constructing zero-loss Web services

Modern Web services must support large and rapidly growing user populations, and remain available 24 hours per day, 7 day per week. Server-cluster is the most promising approach to address this challenge. We further augment the server-cluster schemes with a novel mechanism that enables a Web request to be smoothly migrated and recovered on another working node in the presence of server failure or overload. The new mechanism provides a powerful solution to fault tolerance and dynamic load-distribution for Web services. The administrator can explicitly specify some services to be guaranteed for fault-tolerance or higher performance support. We present the details of our design, implementation, and performance data. The performance results show that the proposed mechanism is efficient and with low associated overhead.

A content placement and management system for distributed Web-server systems

Clusters of commodity computers are becoming an increasingly popular approach for building cost-effective high-performance Internet servers. However, how to place and manage content in such a distributed and complex system is becoming a challenging problem. In particular, such distributed servers tend to be more heterogeneous, and this heterogeneity will further increase the management burden. This paper describes the motivation, design, implementation and performance of a content placement and management system for a heterogeneous distributed Web server.

Realizing Fault Resilience in Web-Server Cluster

Today, a successful Internet service is absolutely critical to be up 100 percent of the time. Server clustering is the most promising approach to meet this requirement. However, the existing Web server-clustering solutions merely can provide high availability derived from its redundancy nature, but offer no guarantee about fault resilience for the service. In this paper, we address this problem by implementing an innovative mechanism that enables a Web request to be smoothly migrated and recovered on another working node in the presence of server failure. We will show that the request migration and recovery could be efficiently achieved in the manner of user transparency. The achieved capability of fault resilience is important and essential for a variety of critical services (e.g., E-commerce), which are increasingly widespread used. Our approach takes an important step toward providing a highly reliable Web service.

Software Defined Networking across Distributed Datacenters over Cloud

Cloud computing, in particular its system spread across several data centers, poses a unique set of networking challenges and management problems. This paper presents our design, implementation, and experience in tackling these challenges by using the concepts and technologies of software-defined network (SDN). We define a network abstraction to incorporate the physical and virtual data plane within a data center. We propose a new network primitive to address the interconnection issues among multiple data centers. A policy plane consolidating these mechanisms provides a programmable interface to enable a wide range of policies and new networking mechanisms. We present the performance result on a real deployment over the production networks. The experience and initial performance results demonstrate that the SDN is a promising solution to cloud networking.

Enabling fault resilience for web services

Today, a successful Internet service is absolutely critical to be up 100 percent of the time. Server clustering is the most promising approach to meet this requirement. However, the existing Web server-clustering solutions can merely provide high availability derived from its redundant nature, but offer no guarantee about fault resilience for the service. In this paper, we address this problem by implementing an innovative mechanism that enables a Web request to be smoothly migrated and recovered on another working node in the presence of server failure. We will show that the request migration and recovery could be efficiently achieved in the manner of user transparency. The achieved capability of fault resilience is important and essential for a variety of critical services (e.g. E-commerce), which are increasingly widespread in use. Our approach takes an important step towards providing a highly reliable Web service.

Building an adaptable, fault tolerant, and highly manageable web server on clusters of non-dedicated workstations

Clustered server architecture is increasingly being viewed as a successful and cost-effective approach to building a high-performance Web server. Existing server-clustering schemes have typically concentrated on the following issues: scalability, high availability, and user transparency. In this paper, we argue that the design goals of the Web server cluster should include adaptability, fault tolerance, and high manageability. In the presence of the Internets highly unpredictable workload, the server system should be self-adapting to changing circumstances. We address this problem by building a Web server on a cluster of non-dedicated workstations. Such a server can easily recruit non-dedicated nodes dynamically in response to load bursts. Based on such a scheme, we designed and implemented an innovative approach that enables an ongoing request to be smoothly migrated to another node either in response to a node failure or overload. We also designed and implemented a management system that enables the Web site manager to manage and maintain the distributed server as a single large system.

An effective mechanism for supporting content-based routing in scalable Web server clusters

Clustered server architecture has been employed for many years on the Internet as a way to increase performance, reliability and scalability in the presence of the Internets explosive growth. Routing mechanism for mapping requests to individual servers within cluster is at the heart of any server clustering techniques. In this paper, we first analyze the deficiencies of existing request routing approaches. Based on these observations, we argue that the request routing mechanism in the cluster-based server should factor in the content of a request in making decisions. Thus, we design and implement a new mechanism to efficiently support content-based routing in the Web server clusters. Performance evaluation on a prototype implementation demonstrates substantial performance improvements over contemporary routing scheme that use weighted round-robin to distribute requests. The proposed mechanism can enable many new capabilities in cluster-based server, such as sophisticated load balancing, differentiated service, special content deployment, session integrity, etc.

Towards Network Virtualization Management for Federated Cloud Systems

The federated cloud system provides the ability to integrate, manage, and use third-party resources from multiple institutions and combines the resources into one infrastructure. A virtual network management system is needed to supporting a seemingly infrastructure where services can be deployed on-demand and across different network domains. We present our work on designing and implementing an integrated system to enable virtual network services across federated clouds. We also propose a novel mechanism to enable network stitching via translating VLAN tag dynamically. The service can be partitioned on hosts across different network domain using different underlying layer-2 or layer-3 network technologies. This paper also provides the results of performance measurement from the implementation based on real production networks.

Network Virtualization Implementation over Global Research Production Networks

The concept of network virtualization has attracted increasing attention by researchers designing and modeling the next-generation Internet paradigm. Researchers need a new type of hybrid network to construct experimental environments for testing new services, architecture, and technologies, especially network protocols, without affecting existing traffic. This paper presents our work on implementing a large-scale testbed that has been created over production networks by integrating novel technologies such as cloud computing, software defined networking, and virtual switches. This approach provides researchers with an experimental networking testbed where traffic can be controlled via programmable virtual switches and dynamic adjustability for many types of network emulation experiments. This system also provides a feasible solution for the networking requirements of federated cloud system.

From Monolithic Systems to a Federated e-Learning Cloud System

Pervasive high speed networks and emerging cloud computing concepts offer a new capability and an opportunity to rethink approaches for delivering information services. However, little attention has been paid to how the legacy monolithic systems can be migrated to the cloud computing paradigm. The paper shares our experience and solutions on building a federated cloud infrastructure for E-learning services. The federated cloud system provides the ability to integrate, manage, and use third-party resources from multiple institutions and combines the resources into one infrastructure. We also provide experience and performance data from six month real trace of the system. While the discussion and experiments focus on cloud paradigm with education services, this research can serve as a reference design for other service providers who want to deploy services in the form of cloud computing.