Distributed query; Join processing
The distributed join is a query operator that combines two relations stored at different sites in the following way: each tuple from the first relation is concatenated with each tuple from the second relation that satisfies a given join condition, e.g., the match in two attributes. The main characteristics of a distributed join are that at least one of the operand relations has to be transferred to another site.
Techniques for evaluating joins on distributed relations have already been discussed in the context of the first prototypes of distributed database systems such as SDD-1, Distributed INGRES and R*. In Ref.  the basic strategies ship whole vs. fetch matches were discussed and results of experimental evaluations were reported. Another report on an experimental comparison of distributed join strategies was given in Ref. .
Special strategies for distributed join evaluation that aim at reducing the...
- 5.Lu H, Carey M. Some experimental results on distributed join algorithms in a local network. In: Proceedings of the 11th International Conference on Very Large Data Bases; 1985. p. 229–304.Google Scholar
- 7.Özsu MT, Valduriez P. Principles of distributed database systems. 2nd ed. London: Prentice Hall; 1999.Google Scholar
- 8.Roth MT, Schwarz P. Don’t scrap it, wrap it! A wrapper architecture for legacy data sources. In: Proceedings of the 23rd International Conference on Very Large Data Bases; 1997. p. 266–75.Google Scholar
- 10.Urhan T, Franklin MJ. XJoin: a reactively-scheduled pipelined join operator. Bull Tech Comm Data Eng. 2000;23(2):27–33.Google Scholar
- 11.Valduriez P. Semi-join algorithms for distributed database machines. In: Schneider H-J, editor. Distributed data bases. Amsterdam: North-Holland; 1982. p. 23–37.Google Scholar
- 12.Williams R, Daniels D, Hass L, Lapis G, Lindsay B, Ng P, Obermarck R, Selinger P, Walker A, Wilms P, Yost RR. An overview of the architecture. San Jose: IBM Research Lab; 1981.Google Scholar