We present an overview and initial performance assessment of a rule-based query optimizer written in VenusDB. VenusDB is an active-database rule language embedded in C++. Following the developments in extensible database query optimizers, first in rule-based form followed by optimizers written as object-oriented programs, the VenusDB optimizer avails the advantages of both. To date, development of rule-based query optimizers have included the definition and implementation of custom rule languages. Thus, extensibility required detailed understanding and often further development of the underlying search mechanism of the rule system. Objectoriented query optimizers appear to have achieved their goals with respect to a clear organization and encapsulation of an optimizer's elements. They do not, however, provide for the concise, declarative expression of domain specific heuristics. Our experience demonstrates that a rule-based query optimizer developed in VenusDB can be well structured, ...

VenusDB is a C++ embedded, forward-chaining rule language and compiler that includes linguistic elements and runtime support for accessing multiple databases across multiple platforms. Multidatabase access was a natural evolutionary step for Venus. Evaluation of Venus using an expert-database application revealed the need for explicit syntax for the expression of event conditions. Thus, VenusDB provides for both eventcondition -action (ECA) rules typical of active-database systems and condition action rules typical of expert systems and expert-database systems. The Venus compiler is readily extended by virtue of an abstract interface, the AMI, that encapsulates the details of data access. Although middleware elements can be amorphous, the AMI forms a well defined interface for the encapsulation of databases and their integration with a forward-chaining inference engine.

Declarative programming, in the form of forward-chaining rule languages, offers advantages complementary to procedurally based object-oriented programming languages. The Venus rule language addresses certain deficiencies with rule-based programming by adding object-oriented features to the rule paradigm. This differs from most other efforts at hybridizing the two paradigms. Most other efforts seek advantage by starting with an object-oriented language and adding rule facilities. Object-oriented features of the Venus language includes encapsulation of parameterized rule modules such that the behavior of a rule group may be defined using inheritance and improved embeddability via resolving polymorphism. An implementation of a large subset of Venus has been exploited for the implementation of benchmark programs and applications. Experience with the application of Venus as the basis of rule-based database query optimizers and a financial expert system demonstrates the reusability of Venus-...

Active-databases are a budding technology where rule-based expert systems can be developed in tight integration with database management systems. This paper presents VenusIDS: an active database component of the Network Exploitation Detection Analyst Assistant (NEDAA) developed as an enhancement to the analysis layer of a two-layer distributed network intrusion detection system using the VenusDB active database system. The layers consist of a network layer and an analysis layer. The network layer contains probes on each subnetwork that sniff network traffic and forward interesting packets in real time to a central Oracle database. The analysis layer comprises this central database and the mechanism to identify and report intrusions. For active-database technology to form an effective basis for intrusion detection, it must be capable of processing network events at least as fast as the network probes produce and log them. Our performance results show that VenusIDS is more than fast enough to handle this rate. Further, VenusIDS is scalable in the number of rules and size of the underlying database. As context for the VenusIDS component, we begin by describing the application architecture and the VenusDB system, with emphasis on the particular features that are important to distributed intrusion detection. We follow that with a description of the VenusIDS component and its performance profile that enables near real time intrusion detection. We conclude with a discussion of future topics for active-database analysis layers.

We present an overview and initial performance assessment of a rule-based query optimizer written in VenusDB. VenusDB is an active-database rule language embedded in C++. Following the developments in extensible database query optimizers, first in rule-based form followed by optimizers written as object-oriented programs, the VenusDB optimizer avails the advantages of both. To date, development of rule-based query optimizers have included the definition and implementation of custom rule languages. Thus, extensibility required detailed understanding and often further development of the underlying search mechanism of the rule system. Objectoriented query optimizers appear to have achieved their goals with respect to a clear organization and encapsulation of an optimizer’s elements. They do not, however, provide for the concise, declarative expression of domain specific heuristics. Our experience demonstrates that a rule-based query optimizer developed in VenusDB can be well structured, flexible, and demonstrate good performance. We compare a relational optimizer developed with Volcano and a functionally identical optimizer developed with VenusDB. The results demonstrate comparable performance on small queries with few joins, while the VenusDB optimizer scales better and outperforms Volcano on larger joinarity queries. Since we did not have to develop a specialized rule language or consider application specific programming constructs, the source code for the optimizer is small and straightforward, about one-third the size. Similar code comparisons with an object-based optimizer, the VenusDB optimizer reveals similar benefit.