The resulting manuscript strikes a balance between our two objectives, namely to address new and emerging issues, and maintain the main characteristics of the book in addressing the principles of distributed data management. The third edition is coming out at a time when there is renewed interest in distributed data management. The last ten years have seen an accelerated investigation of distributed data management technologies spurred by advent of high-speed networks, fast commodity hardware, very heavy parallelization of hardware, interest in cloud computing, and, of course, the increasing pervasiveness of the web. We hope the book contributes the the renewed discussion on these topics.
One of the concerns for the cloud-based solutions is the fact that the components responsible for service discovery, monitoring and load-balancing still employ centralized approaches. This, in turn, comes from traditional architectural model designed for Web Services which defines three types of participants  : 1) service providers, which create and own services, and advertise them to potential users; 2) service brokers, which maintain a registry of advertised services and might introduce service providers to service requestors; 3) service requestors, which search the registries of service brokers for suitable service providers and then contact a service provider to use its services. Due to its nature, the presence of central authority entities like service brokers is often inappropriate, since such solutions lack satisfactory scalability, present a single point of failure and lead to performance bottlenecks and network congestion. On the other hand, considering distributed nature of cloud-based architecture, it is reasonable to use distributed approach to cloud service management and discovery as well, which in turn leads to the idea of using inherently decentralized, fault tolerant and scalable peer-to-peer paradigm.
In this paper, we propose a Web Service discovery architecture based on structured P2P overlay network, which utilizes platform-independent attribute-value based method for describing services and an algorithm for service discovery that allows execution of range and multidimensional queries. Among ideas that inspired us to propose this system, we can name PWSD (Peer-to-Peer based Web service discovery) architecture, presented in  and Intentional Name System, described in detail in  . We propose a lightweight approach that could be based on any DHT overlay, uses attribute-value based service description without resorting to complex data description frameworks (e.g. Semantic Web) and is designed with modular approach in mind. While being inherently a framework for building reliable and scalable network for providing, discovering and using services, the approach presented in this paper is designed to be applied to the problem of efficient service provisioning in cloud-based solutions, that is, for building an overlay network that makes cloud-based services resilient, scalable and managed in a distributed manner.
2) Service storage: an efficient data storage which supports all common types of queries, allows storing typed data and can be easily deployed on each node in the overlay. In current implementation we decided to use document-oriented database MongoDB ( ), which satisfies all the requirements stated above and is more flexible than many traditional relational databases, since it is a lightweight solution which has little deployment overhead, do not require predefined schema and do not include unnecessary at this point functionality, such as transactional processing.
To further demonstrate the system operation, we chose a domain of dedicated gaming solution based on Google Cloud Platform ( -server-gaming-solution), since in our opinion it represents one of the typical scenarios found in distributed systems, requiring elaborate strategies for load-balancing and efficient service provisioning. 153554b96e