M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. Int. Conf. on Very Large Data Bases (VLDB), pages 228--238, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....etc. These papers are orthogonal to this one in that they treat optimization issues, and assume that caching strategies are either efficient or non existent. Kemper, Moerkotte, Peithner and Steinbrunn have proposed bypass techniques for optimizing and executing disjunctive queries [KMPS94, SPMK95] This work alleviates some of the problems of redundant computation in disjunctive queries, but does not present a solution to the problem of redundant method invocation in general for example, it does not avoid duplicate method invocation for the simple thumbnail example above. This paper ....

Michael Steinbrunn, Klaus Peithner, Guido Moerkotte, and Alfons Kemper. Bypassing Joins in Disjunctive Queries. In Proc. 21st International Conference on Very Large Data Bases, Zurich, September 1995.

....of the evaluation of the query. Further, the complementary output allows ecient evaluation of disjunctive queries: we need to evaluate the second disjunct only on the reject (complementary output) of the rst. This technique is termed bypassing and has recently been pointed out in the literature [SPMK95] Moreover, the operator set has a spread operator that allows replication of its input to a number of outputs, allowing multiple sub queries to share the same intermediate result. These features of the operators result in query graphs that are directed acyclic graphs. The advantage of using the ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing Joins in Disjunctive Queries. In Proceedings of the Conference on Very Large Databases, Zurich, Switzerland, pages 228-238, September 1995.

....is the last resort then. This paper shows that it is possible to find efficient execution strategies for the monoid comprehension calculus. Moreover, we present a transformation strategy that yields efficient execution plans. We will especially integrate the recently proposed bypass plans [SPMK95] Such approaches have not been undertaken for the monoid calculus yet and go beyond the optimization through normalization ideas of [FM95a] Also, algebra based optimizers that are aware of such execution strategies are still to be developed. Its apparent potential made the monoid comprehension ....

....materialisation or join strategies like semijoins or nestjoins [Ste95] of the underlying query engine. Another way to improve query performance is to analyze the ways data flows through potentially expensive predicates. This gives rise to techniques like the recently proposed bypass joins [SPMK95] which can be smoothly incorporated into the monoid calculus framework. We will discuss both ways in Section 4. 3 Monoid Comprehensions as a Target for OQL Translation This section shows that the ODMG OQL constructs can be mapped to the comprehension calculus rather straightforward. As the query ....

[Article contains additional citation context not shown here]

Michael Steinbrunn, Klaus Peithner, Guido Moerkotte, and Alfons Kemper. Bypassing Joins in Disjunctive Queries. In Proceedings of the 21st Int'l Conference on Very Large Databases (VLDB), September 1995.

....rise to E 1 x;y:p(x;y) E 2 . Finding the optimal order of generators and predicates is a task that is already known as the join enumeration or join ordering problem. To detect situations where sophisticated join types, like semijoins, nestjoins [Ste95] or the recently proposed bypass join [SPMK95] would allow an efficient evaluation of comprehensions, seems to be very hard if not impossible after normalization. One is stuck with a naive nested loop strategy of comprehension evaluation, having its algebraic analogy in computing cartesian products. We will see later on that we can do ....

Michael Steinbrunn, Klaus Peithner, Guido Moerkotte, and Alfons Kemper. Bypassing Joins in Disjunctive Queries. In Proceedings of the 21st Int'l Conference on Very Large Databases (VLDB), September 1995.

....database setting for caching expensive methods. However, as we show in this paper, memoization is an inadequate solution for the large volumes of data present in database systems. Kemper, Steinbrunn, et al. have proposed bypass techniques for optimizing and executing disjunctive queries [KMPS94, SPMK95] This work alleviates some of the problems of redundant computation in disjunctive queries, but does not present a solution to the problem of redundant method invocation in general for example, it does not avoid duplicate method invocation for the simple thumbnail example above. This paper ....

Michael Steinbrunn, Klaus Peithner, Guido Moerkotte, and Alfons Kemper. Bypassing Joins in Disjunctive Queries. In Proc. 21st International Conference on Very Large Data Bases, Zurich, September 1995.

....plan. It is important to note that these issues are indeed orthogonal to our problems of predicate placement once queries have been rewritten into cheaper forms, they still need to have their predicates optimally placed into a query plan. Kemper, Steinbrunn, et al. Kemper et al. 1994; Steinbrunn et al. 1995] address the problem of planning disjunctive queries with expensive predicates; their work is not easily integrated with a traditional (conjunct based) optimizer. Like most System R based optimizers, Illustra focuses on Boolean factors (conjuncts) Within Boolean factors, the operands of OR are ....

Steinbrunn, M., Peithner, K., Moerkotte, G., and Kemper, A. 1995. Bypassing Joins in Disjunctive Queries. In Proc. 21st International Conference on Very Large Data Bases (Zurich, Sept. 1995).

.... in [KM90a, KM90b] and [Ber91] Optimization of Object Oriented nested queries has been considered in [CM93] and [CZ96] Methods in queries make it necessary to reconsider ways to index [MS86, Hwa94] and to consider the order in which predicates should be evaluated on data [HS93, Hel94, KMPS94, SPMK95] Many areas of querying impacted by objects have been well studied, but as far as we know, mutability s impact on querying has not been considered before. Transformation rules over complex object bases have been proposed in [BK90] which provides rules for multiple collection types) Osb88] ....

Michael Steinbrunn, Klaus Peithner, Guido Moerkotte, and Alfons Kemper. Bypassing joins in disjunctive queries. In 21st International Conference on Very Large Data Bases, Zurich, Switzerland, September 1995.

No context found.

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. Int. Conf. on Very Large Data Bases (VLDB), pages 228--238, 1995.

....plans for the three main query classes. Beforehand, we have to introduce the used algebra operators. 3.1 Algebra Operators For the subsequent evaluation plans, we enhance OQL by an if : then : expression. It is useful for rewriting outer restrictions as proposed in [Mur88, SPMK95, CM95a]. At the algebraic level, this is reflected by an algebra operator if p (E true ; E false ) where E true is the result if p evaluates to true and otherwise E false is the result. Actually, the if constructs are much more often used in the simplification rules to optimize the 13 less important ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 228--238, Zrich, Switzerland, September 1995.

....for the three main query classes. Beforehand, we have to introduce the used algebra operators. 3.1 Algebra Operators if Expression For the subsequent evaluation plans, we enhance OQL by an iif : then : else : j expression. It is useful for rewriting outer restrictions as proposed in [Mur88, SPMK95, CM95a]. At the algebraic level, this is reAEected by an algebra operator if p (E true ; E false ) where E true is the result if p evaluates to true and otherwise E false is the result. Actually, the if constructs are much more often used in the simpli cation rules of Section 4 to optimize the 13 ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 228238, Z#rich, Switzerland, September 1995.

....plans for the three main query classes. Beforehand, we have to introduce the used algebra operators. 3.1 Algebra Operators For the subsequent evaluation plans, we enhance OQL by an if : then : else : expression. It is useful for rewriting outer restrictions as proposed in [Mur88, SPMK95, CM95a]. At the algebraic level, this is reflected by an algebra operator if p (E true ; E false ) where E true is the result if p evaluates to true and otherwise E false is the result. Actually, the if constructs are much more often used in the simplification rules to optimize the 13 less important ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 228--238, Zürich, Switzerland, September 1995.

....and Stonebraker present an algorithm for the optimal placement of expensive selections within a join graph. By applying their predicate migration algorithm to every possible join ordering, they gave an O(m (m s) 4 log(m s) algorithm for ordering m joins and s selections. The authors of [19] introduced a new type of evaluation plans called bypass evaluation plans which are superior to traditional plans in the case of disjunctive queries. They gave an O(2 n Gamma1 n ) algorithm for determining near optimal bypass plans for disjunctive queries with expensive predicates. By ....

....table below compares the different approaches of ordering joins and selections. The parameter m denotes the number of relations and n denotes the total number of operators in the query. dynamic IK KBZ dynamic predicate bypassing this progr. 6, 10] progr. migration joins paper [17] 16] 4, 5] [19] query graph arbitrary acyclic arbitrary arbitrary arbitrary acyclic query type conjunctive conjunctive arbitrary conj. selection arbitrary conjunctive predicates processing bushy left deep bushy bushy bypass left deep trees trees trees trees trees plans trees cross prod. yes no yes yes no no ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. Int. Conf. on Very Large Data Bases (VLDB), pages 228--238, Zurich, 1995.

....evaluate the bypass evaluation plans against the traditional evaluation techniques utilizing a CNF or DNF based query predicate. This work was supported by the German Research Council under contracts DFG Ke 401 6 2 and Ke 401 7 1. Excerpts of this work appeared in [KMPS94] and [SPMK95] Contents 1 Introduction 4 2 Bypass Technique 6 2.1 Running Example . 6 2.2 Drawbacks of Common Techniques . 8 2.3 Commercial Database Systems . ....

....possible to bypass certain costly or less selective predicates if the fate of the corresponding stream can be determined without considering the particular predicate. The ideas of bypass evaluation have been presented in two prior conference papers: KMPS94] introduced bypass selection and [SPMK95] contained the bypass join processing. In this paper we have covered the necessary issues to incorporate bypass evaluation in a real system: ffl we defined the underlying algebra, ffl the equations for generating bypass evaluation plans by the query optimizer were given and the OPT and the ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 228--238, Zurich, Switzerland, September 1995.

.... Gamma [ Figure 1: Optimal (Bypass) plan for example query Average cost: 16; 000 units note that in this section, we shall give just the results of the cost calculations according to our cost model. The cost model itself and a sample application (namely, for the bypass plan) is provided in [SPMK94] The bypass evaluation plan starts on the left hand side with the semijoin operation Delta C address of Publisher and Work and the selection oe C quantity of Order on the right hand side. Tuples satisfying C address are certain to be elements of the result set, hence they may bypass the other ....

M. Steinbrunn, K. Peithner, G. Moerkotte, and A. Kemper. Bypassing joins in disjunctive queries. Technical Report MIP-9412, Universit at Passau, 94030 Passau, Germany, 1994. WWW: ftp://dodgers.fmi.uni-passau.de/pub/ papers/techreports/MIP9412.ps.Z

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC