## On the Complexity of Generating Optimal Plans with Cross Products (Extended Abstract) (1997)

Venue: | In PODS Conference |

Citations: | 21 - 1 self |

### BibTeX

@INPROCEEDINGS{Scheufele97onthe,

author = {Wolfgang Scheufele and Guido Moerkotte},

title = {On the Complexity of Generating Optimal Plans with Cross Products (Extended Abstract)},

booktitle = {In PODS Conference},

year = {1997},

pages = {238--248}

}

### Years of Citing Articles

### OpenURL

### Abstract

In modern advanced database systems the optimizer is often faced with the problem of finding optimal evaluation strategies for queries involving a large number of joins. Examples are queries generated by deductive database systems and path expressions in object-oriented database systems. The best plan can be found in the very large search space of bushy trees where plans are allowed to contain cross products. A general question arises: For which (sub-) problems can we expect to find polynomial algorithms generating the best plan? We attack this question from both ends of the spectrum. First, we show that we cannot expect to find any polynomial algorithm for any subproblem as long as optimal bushy trees are to be generated. More specifically, we show that the problem is NP-hard independent of the query graph.

### Citations

10926 |
Computers and Intractability: A Guide to the Theory of NPCompleteness. Freeman and
- Garey, Johnson
- 1979
(Show Context)
Citation Context ...a polynomial time Turing reduction to XR. We have chosen to take the exact cover with 3-sets (X3C ) as the problem of choice. The next definition recalls this problem which is known to be NP-complete =-=[2]-=-. Definition 2.2 (exact covering with 3-sets, X3C) Let S be a set with jSj = 3q elements. Further let C be a collection of subsets of S containing three elements each. The following decision problem i... |

8531 |
Introduction to Algorithms
- Cormen, Leiserson, et al.
- 2009
(Show Context)
Citation Context ...mposable subchains of the same extent. There is an obvious dynamic programming algorithm to compute optimal recursive decomposable subchains. It is not hard to see that Bellman's optimality principles=-=[7, 14]-=- holds and every optimal recursively decomposable subchain can be decomposed into smaller optimal recursively decomposable subchains. Example 5: In order to compute an optimal recursively decomposable... |

481 | Access Path Selection in a Relational Database Management System
- Selinger, Astrahan, et al.
- 1979
(Show Context)
Citation Context ...nly few exceptions [4, 6], no efficient, exact algorithms have been developed so far. Little is known about the structure of the problem and only general optimization methods like dynamic programming =-=[10]-=- and probabilistic algorithms (for an overview see [13]) seem to be applicable. In the last years the problem gained fresh impetus due to its relevance in modern advanced database systems like deducti... |

120 | Optimization of nonrecursive queries
- Krishnamurthy, Boral, et al.
- 1986
(Show Context)
Citation Context ...since the introduction of relational database systems, optimization of multiway joins has been a challenging problem. Despite the huge research efforts in query optimization, with only few exceptions =-=[4, 6]-=-, no efficient, exact algorithms have been developed so far. Little is known about the structure of the problem and only general optimization methods like dynamic programming [10] and probabilistic al... |

108 |
Mathematical Programming: Theory and Algorithms
- Minoux
- 1986
(Show Context)
Citation Context ...mposable subchains of the same extent. There is an obvious dynamic programming algorithm to compute optimal recursive decomposable subchains. It is not hard to see that Bellman's optimality principles=-=[7, 14]-=- holds and every optimal recursively decomposable subchain can be decomposed into smaller optimal recursively decomposable subchains. Example 5: In order to compute an optimal recursively decomposable... |

82 |
Measuring the complexity of join enumeration in query optimization
- Ono, Lohman
- 1990
(Show Context)
Citation Context ...an algorithm for job sequencing [8]. Lately, it was shown that including bushy trees and cross products into the search space yields non-neglectable performance gains over the restricted search space =-=[9]-=-. The question that immediately arises is whether we can expect polynomial algorithms for this more general problem. Let us discuss first the problem of generating optimal left-deep processing trees p... |

70 | Left-deep vs. bushy trees: An analysis of strategy spaces and its implications for query optimization
- Ioannidis, Kang
- 1991
(Show Context)
Citation Context ...e of a polynomial algorithm for constructing optimal bushy trees, the only chance is to develop heuristics or to apply probabilistic optimization procedures. Concerning the latter, Ioannidis and Kang =-=[5]-=- make some important observations. In fact, they conclude that for probabilistic algorithms producing good bushy trees seems to be easier than constructing good left-deep trees. If this observation al... |

50 | Rapid bushy join-order optimization with cartesian products
- Vance, Maier
- 1996
(Show Context)
Citation Context ...and fewer experiments for random queries with up to 300 relations to compare the results of our algorithms. For ns15 we additionally compared the results with a standard dynamic programming algorithm =-=[10, 15]-=-. The results of all our experiments can be summarized as follows. ffl All algorithms yielded identical results. ffl Backtracking always led to exactly one sequence of contradictory chains. ffl In the... |

38 |
Sequencing with Series-Parallel Precedence Constraints
- Monma, Sidney
- 1979
(Show Context)
Citation Context ...rithms exist for the hash-join and nested-loop join cost functions [6], and a cost function which sums up all intermediate result sizes [1]. All these algorithms apply an algorithm for job sequencing =-=[8]-=-. Lately, it was shown that including bushy trees and cross products into the search space yields non-neglectable performance gains over the restricted search space [9]. The question that immediately ... |

37 | Multi-Join Optimization for Symmetric Multiprocessors
- Shekita, Young, et al.
- 1993
(Show Context)
Citation Context ... heuristics to be developed, then our result may not be as bad as it seems to be at first sight. There is some evidence that this is the case. Even simple greedy algorithms seem to perform quite well =-=[12]-=-. What is needed in the future are careful studies concerning heuristics for the construction of good bushy trees. Only after that, the builders of query optimizers can trust them. Another way for res... |

36 |
Optimal nesting for computing Nrelational joins
- Ibaraki, Kameda
- 1984
(Show Context)
Citation Context ...since the introduction of relational database systems, optimization of multiway joins has been a challenging problem. Despite the huge research efforts in query optimization, with only few exceptions =-=[4, 6]-=-, no efficient, exact algorithms have been developed so far. Little is known about the structure of the problem and only general optimization methods like dynamic programming [10] and probabilistic al... |

25 | Optimizing join orders
- Steinbrunn, Moerkotte, et al.
- 1993
(Show Context)
Citation Context ...ms have been developed so far. Little is known about the structure of the problem and only general optimization methods like dynamic programming [10] and probabilistic algorithms (for an overview see =-=[13]-=-) seem to be applicable. In the last years the problem gained fresh impetus due to its relevance in modern advanced database systems like deductive and object-oriented database systems. In object-orie... |

21 | On the complexity of generating optimal left-deep processing trees with cross products
- Cluet, Moerkotte
- 1995
(Show Context)
Citation Context ...orithm deals with the blockwise nested-loop join [4]. Other algorithms exist for the hash-join and nested-loop join cost functions [6], and a cost function which sums up all intermediate result sizes =-=[1]-=-. All these algorithms apply an algorithm for job sequencing [8]. Lately, it was shown that including bushy trees and cross products into the search space yields non-neglectable performance gains over... |

18 |
Bootstrap percolation, the schrÃ¶der numbers, and the n-kings problem
- Shapiro, Stephens
- 1991
(Show Context)
Citation Context .... Each contradictory chain can be completely recursively decomposed into adjacent pairs of connected subchains. Subchains with this property are defined next (similar types of decompositions occur in =-=[3, 11]-=-). Definition 3.5 ((optimal) recursively decomposable subchain)sA recursively decomposable subchain [s] u is inductively defined as follows. 1. If s is a single relation then [s] u is recursively deco... |

1 | Enumerating consecutive and nested partitions for graphs
- Hwang, Chang
- 1993
(Show Context)
Citation Context .... Each contradictory chain can be completely recursively decomposed into adjacent pairs of connected subchains. Subchains with this property are defined next (similar types of decompositions occur in =-=[3, 11]-=-). Definition 3.5 ((optimal) recursively decomposable subchain)sA recursively decomposable subchain [s] u is inductively defined as follows. 1. If s is a single relation then [s] u is recursively deco... |