# Complex transitive closure queries on a fragmented graph

## Abstract

In this paper we study the reformulation of transitive closure queries on a fragmented graph. We split a query into several subqueries, each requiring only a fragment of the graph. We prove this reformulation to be correct for shortest path and bill of material queries. Here we describe the reformulation for an abstract graph, elsewhere we have described an actual implementation of our approach and some promising simulation results.

We view the study of distributed computation of transitive closure queries as a result of the trend towards distributed computation. First selections were distributed to fragments of a relation, then fragmentation was used to compute joins in a distributed way, and now we are studying distributed computation of transitive closure queries. This should result in a deeper insight into the use and possible benefit of parallelism. Our work may be used in ordinary distributed databases as well as advanced multiprocessor database machines, such as PRISMA.

Although this research was started to efficiently use distributed computation, it turns out to be beneficiary in a central environment as well. This is due to the introduction of extra selections, stemming from an appropriate fragmentation. This leads to extra focus on relevant data.

## Keywords

Short Path Logic Programming Complementary Information Transitive Closure Large Data Base## Preview

Unable to display preview. Download preview PDF.

## References

- [1]Agrawal, R., Borgida, A. and Jagadish, H.V. ‘Efficient management of transitive relationships in large data and knowledge bases,’ in
*Proc. ACM-Sigmod Conference*, Portland, Oregon, 1989, pp. 253–262.Google Scholar - [2]Agrawal, R. and Jagadish, H.V. ‘Efficient search in very large databases,’ in
*Proc. 14th Int. Conf. on Very Large Data Bases*, Los Angeles, 1988, pp. 407–418.Google Scholar - [3]Agrawal, R. and Jagadish, H.V. “Multiprocessor transitive closure algorithms,” in
*Proc. Int. Symp. on Databases in Parallel and Distributed Systems*, Austin, Texas, Dec. 5–7 1988, pp. 56–66.Google Scholar - [4]Apers, P.M.G., Houtsma, M.A.W. and Brandse, F. “Processing recursive queries in relational algebra,” in
*Data and Knowledge (DS-2), Proc. of the 2nd IFIP 2.6 Working Conference on Database Semantics, Albufeira, Portugal, Nov. 3–7, 1986*, R.A. Meersman and A.C. Sernadas (eds.), North Holland, 1988, pp. 17–39.Google Scholar - [5]Apers, P.M.G., Kersten, M.L., and Oerlemans, H. “PRISMA database machine: A distributed main-memory approach,” in
*Advances in Database Technology-EDBT'88*, J.W. Schmidt, S. Ceri and M. Missikoff (eds.), Lecture Notes in Computer Science #**303**, Springer-Verlag, 1988, pp. 590–593.Google Scholar - [6]Bancilhon, F., Maier, D., Sagiv, Y., and Ullman, J.D. “Magic sets and other strange ways to implement logic programs,”
*Proc. ACM SIGMOD-SIGACT Symp. on Principles of Database Systems*, 1986.Google Scholar - [7]Cacace F., Ceri S., Crespi-Reghizzi C., Tanca L., and Zicari R. “Integrating object oriented data modeling with a rule-based programming language”, in
*Proc. ACM-Sigmod Conference*, Atlantic City, USA, May 1990.Google Scholar - [8]Cacace, F., Ceri, S., and Houtsma, M.A.W. “A survey of parallel execution strategies for transitive closure and logic programs,” Technical Report, University of Twente-Politecnico Milano, Sep. 1990. Submitted for publication.Google Scholar
- [9]Ceri, S., Cacace, F., and Houtsma, M.A.W. “An overview of parallel strategies for transitive closure on algebraic machines,”
*Proc. Workshop on Parallel Database Systems*, Noorwijk, the Netherlands, Sept. 1990, Lecture Notes in Computer Science, Springer-Verlag.Google Scholar - [10]Ceri, S. and Tanca, L. “Optimization of systems of algebraic equations for evaluating Datalog queries,” in
*Proc. of the 13th Int. Conf. on Very Large Data Bases*, Brighton, Sept. 1987, pp. 31–42.Google Scholar - [11]Ceri, S., Gottlob, G., and Tanca, L.
*Logic Programming and Databases*, Springer-Verlag, 1990.Google Scholar - [12]Ceri, S. and Pelagatti, G.
*Distributed Databases, principles & systems*, McGraw-Hill, 1985.Google Scholar - [13]Cohen, A.R. and Wolfson, O. “Why a single parallelization strategy is not enough in knowledge bases,” Technical report no. 561, Technion, Israel, June 1989.Google Scholar
- [14]Ganguly S., Silberschatz A., and Tsur S. “A framework for the parallel processing of Datalog queries,”
*Proc. ACM-Sigmod Conference*, Atlantic City, USA, May 1990.Google Scholar - [15]Houtsma, M.A.W.
*Data and Knowledge Base Management Systems: Data Model and Query Processing*, Ph.D. Thesis, University of Twente, Enschede, the Netherlands, Nov. 1989.Google Scholar - [16]Houtsma, M.A.W., Apers, P.M.G., and Ceri, S. “Parallel computation of transitive closure queries on fragmented databases,” Technical Report INF-88-56, University of Twente, the Netherlands, Dec. 1988.Google Scholar
- [17]Houtsma, M.A.W., Apers, P.M.G., and Ceri, S. “Distributed transitive closure computations: The disconnection set approach,” Technical report INF-89-12, University of Twente, the Netherlands, Nov. 1989.Google Scholar
- [18]Houtsma, M.A.W., Apers, P.M.G., and Ceri, S. “Distributed transitive closure computations: The disconnection set approach,” in
*Proc. 16th Int. Conf. on Very Large Data Bases*, Brisbane, Australia, Aug. 1990.Google Scholar - [19]Houtsma, M.A.W., Cacace, F., and Ceri, S. “Parallel hierarchical evaluation of transitive closure queries,” Technical Report, University of Twente, Nov. 1990.Google Scholar
- [20]Hulin, G. “Parallel processing of recursive queries in distributed architectures” in
*Proc. 15th Int. Conf. on Very Large Data Bases*, Amsterdam, 1989, pp. 87–96.Google Scholar - [21]Ioannidis, Y.E. “On the computation of the transitive closure of relational operators,”
*Proc. of the 12th Int. Conf. on Very Large Data Bases*, Kyoto, Japan, Aug. 1986, pp. 403–411.Google Scholar - [22]Ioannidis, Y. and Ramakrishnan, R. ‘Efficient transitive closure algorithms,’ in
*Proc. 14th Int. Conf. on Very Large Data Bases*, Los Angeles, 1988, pp. 382–394.Google Scholar - [23]Kersten, M.L., Apers, P.M.G., Houtsma, M.A.W., van Kuijk, H.J.A., and van de Weg, R.L.W. “A distributed, main-memory database machine,” in
*Proc. of the 5th Int. Workshop on Database Machines*, Karuizawa, Japan, Oct. 5–8, 1987; and in*Database Machines and Knowledge Base Machines*, M. Kitsuregawa and H. Tanaka (eds.), Kluwer Academic Publishers, 1988, pp. 353–369.Google Scholar - [24]Kleinhuis, G. and Oskam, K.R. “Evaluation and simulation of parallel algorithms for the transitive closure operation,” M.Sc. Thesis, University of Twente, the Netherlands, May 1989.Google Scholar
- [25]Larson, P-A. and Deshpande, V. ‘A file structure supporting traversal recursion,’ in
*Proc. ACM-SIGMOD*, Portland, Oregon, 1989, pp. 243–252.Google Scholar - [26]
- [27]Nejdl W., Ceri S., and Wiederhold G. “Evaluating recursive queries in distributed databases,” Tech. Rep. 90-015, Politecnico di Milano, submitted for publication.Google Scholar
- [28]Raschid, L. and Su, S.Y.W. “A parallel strategy for evaluating recursive queries,” in
*Proc. of the 12th Int. Conf. on Very Large Data Bases*, Kyoto, Japan, Aug. 1986.Google Scholar - [29]Tandem Database Group “NonStop SQL, a distributed, high-performance, high-availability implementation of SQL,” Tandem report, April 1977.Google Scholar
- [30]Wolfson, O. “Sharing the Load of Logic-program Evaluation,” in
*Proc. Int. Symp. on Databases in Parallel and Distributed Systems*, Austin, Texas, Dec. 5–7 1988, pp. 46–55.Google Scholar