A Study Of Continuous Query Processing Techniques

Increasingly pervasive networks are leading towards a world where data is constantly in motion. In such a world, conventional techniques for query processing, which were developed under the assumption of a far more static and predictable computational environment, will not be sufficient. In many recent applications, data may take the form of continuous data streams, which are dynamic changing. This requires modern database systems to handle continuous queries for such data streams.There are several critical techniques, like scalability, incremental maintenance, adaptive capability, etc., involved in continuous query processing for dynamic environments. We firstly use a general and flexible framework architecture to analysis the previous technologies, such as triggers, materialized views and producing partial results to determine their value for the study of continuous queries. After that, we illustrate the three critical techniques in detail. The scalability is to resist the huge number of concurrent queries, using a scalable group policy to share queries. We present two types of scalable technology in this paper, which are NiagaraCQ, applying in the web system, and the Flux, solving clustered scalability in dynamic environments. Since the continuous queries are continuous running, and the data sources for queries are continuous changing, the technology of the incremental view maintenance apply various rules and transitions to the view relations to implement the incremental computing, and finally transfer the update results to clients. To accommodate the dynamic environment which continuous queries often involve, we need adaptive capability, adaptively routing tuples through operators in the query plan, to realize efficient query computing. This paper explains a typical adaptive mechanism, named eddy as well as its using utility.Last but not least, we synthetically apply all the critical technologies mentioned above into the mobile computing environment, based on the spatial-temporal database, and implement a highly adaptive and scalable continuous query processing paradigm across wide-area network with some experiment results about the system performance.