A bottom‐up design for spatial search in large networks and clouds

Abstract

Information in networked systems often has spatial semantics: routers, sensors, or virtual machines have coordinates in a geographical or virtual space, for instance. In this paper, we propose a design for a spatial search system that processes queries against spatial information that is maintained in local databases inside a large networked system. In contrast to previous works in spatial databases and peer-to-peer designs, our design is bottom-up, which makes query routing network aware and thus efficient, and which facilitates system bootstrapping and adaptation. Key to our design is a protocol that creates and maintains a distributed index of object locations based on information from local databases and the underlying network topology. The index builds upon minimum bounding rectangles to efficiently encode locations. We present a generic search protocol that is based on an echo protocol and uses the index to prune the search space and perform query routing. The response times of search queries increase with the diameter of the network, which is asymptotically optimal. We study the performance of the protocol through simulation in static and dynamic network environments, for different network topologies, and for network sizes up to 100 000 nodes. In most experiments, the overhead incurred by our protocol lies well below 30% of a hypothetical optimal protocol. In addition, the protocol provides high accuracy under significant churn. Int J Network Mgmt. 2018;28:e2041. wileyonlinelibrary.com/journal/nemConsider a networking scenario where locations for routers, servers, and virtual machines are produced by a network coordinate system, such as Vivaldi. 4 In this case, the Euclidean distance between locations refers to the round-trip-time between the network entities at those locations. Spatial queries include finding a server that is closest to a client application or finding a server with a similar distance to a given set of clients. Second, consider a networking environment where locations refer to geographic coordinates and distances refer to geographical distances. A spatial query for this case is finding backup servers outside a given area to improve availability in case of failures. Third, consider an Internet of Things (IoT) scenario with garbage containers at different geographical locations. A spatial query is finding full containers within a certain distance from a given place, for example, to facilitate garbage collection. Fourth, consider an information and communication technology infrastructure, whereby locations are IP addresses mapped onto R 4 (in case of IPv4). A spatial query in this case is finding physical or virtual machines in a given address range for the purpose of security management. Lastly, consider the case where a router searches for the closest gateway in the context of performance-based routing in order to optimize service levels.We illustrate the use of the Euclidean spatial model with two spatial queries. We assume a search space O of objects, whereby each object o ∈ O has a locat...

A bottom-up approach to real-time search in large networks and clouds

“…The indexes are built by distributing a given R-tree onto peer nodes. More details about VBI-tree and DP-tree can be found in [99].…”

“…The main advantage of a k-d tree index is that the addition or removal of objects does not require a change in index. For a more in-depth discussion of the MIDAS design see [99].…”