G. Mitchell, U. Dayal and S. B. Zdonik, Control of an Extensible Query Optimizer: A Planning-Based Approach, Proc. VLDB 1993, Pg. 517-528.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....optimizers with the goal of achieving the efficiency of bottom up optimizers and the extensibility of top down optimizers. The EROC system developed at Bell Labs and NCR [MBH96] combines top down and bottom up approaches. Region based optimizers developed at METU [ONK95] and at Brown University [MDZ93] use different optimization techniques for different phases of optimization in order to achieve increased efficiency. Commercial systems from Microsoft [Gra96] and Tandem [Cel96] are based on Cascades. They include techniques similar to those we present here, but to our knowledge these are the ....

G. Mitchell, U. Dayal and S. B. Zdonik, Control of an Extensible Query Optimizer: A Planning-Based Approach, Proc. VLDB 1993, Pg. 517-528.

....also been suggested for extensible database systems (e.g. EXODUS [4] and GENESIS [3] in order to allow the creation of DBMS s which are tailored to the needs of speci c applications. However, extensibility of language components in such systems has focussed almost exclusively on query optimisers [15, 19, 25, 27, 28] rather than on compilers in general. Some of these optimisers (the Starburst optimiser [19] for example) have been implemented along similar lines to our compiler, with components of the optimiser operating as transformers on a common internal format, while others (such as the Volcano Optimiser ....

G. Mitchell, U. Dayal, and S.B. Zdonik. Control of an Extensible Query Optimizer: A Planning-Based Approach. In R. Agrawal, S. Baker, and D. Bell, editors, Proceedings of the 19th International Conference on Very Large Databases, pages 517-528, Dublin, 1993. Morgan Kaufmann Publishers, Inc.

....optimizers with the goal of achieving the efficiency of bottomup optimizers and the extensibility of top down optimizers. The EROC system developed at Bell Labs and NCR [MBH96] combines top down and bottom up approaches. Region based optimizers developed at METU [ONK95] and at Brown University [MDZ93] use different optimization techniques for different phases of optimization in order to achieve increased efficiency. Commercial systems from Microsoft [Gra96] and Tandem [Cel96] are based on Cascades. They include techniques similar to those we present here, but to our knowledge these are the ....

G. Mitchell, U. Dayal and S. B. Zdonik, Control of an Extensible Query Optimizer: A Planning-Based Approach, Proc. VLDB 1993, Pg. 517-528.

....it. Another one is to build a video query language based on the CVOT model. The spatial, temporal, and spatio temporal queries can be translated into the query calculus and then the query algebra. Until that time it is possible to optimize these queries using object query optimization techniques [MDZ93, OB95] ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proceedings of the 20th International Conference on Very Large Data Bases, pages 517---528, Dublin, Ireland, August 1993.

....[DG94] We also intend to build a video query language based on the CVOT model. The spatial, temporal, and spatiotemporal queries can be translated into the query calculus and then the query algebra. Therefore, it is possible to optimize these queries using object query optimization techniques [MDZ93, OB95] ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proceedings of the 20th International Conference on Very Large Data Bases, pages 517---528, Dublin, Ireland, August 1993.

....implements the extended pattern matching algorithm described in this paper. For efficient rule based rewriting, all the syntactic transformation rules used by the compiler, to obtain canonical form OQL expressions, are coded in C . The architecture of the compiler is modular, as suggested in [Mitchell93]. Each module takes as input an OQL expression, and produces a set of equivalent OQL expressions, based on the knowledge specific to the module and using a specific control strategy. Each module has a goal which characterizes the resulting OQL expressions. This modular architecture allows ....

G. Mitchell, U. Dayal, and S. Zdonik, "Control of an extensible query optimizer: A planning-based approach", Int. Conf. on VLDB, Dublin, August 1993.

....expression represents a user query [Shaw90, Cluet92, Blakeley93] Within the OFL formalism, an OFL program (a functional expression) represents the user query. Thus, equivalent expressions are equivalent OFL programs. Query planning explores the search space of execution plans to find the best one [Valduriez91, Lanzellote92, Mitchell93, Finance94]. Within the OFL formalism, query execution plans are OFL programs. Since the programs describe traversal of collections, equivalent programs correspond to equivalent traversals. The search for the best equivalent traversal includes the following transformations: reducing collection sizes by ....

Mitchell G., Dayal U., Zdonik S.,"Control of an Extensible Query Optimizer: A PlanningBased Approach", Intl. Conf. on VLDB, Dublin, Ireland 1993.

....alternative plans for star queries. Yet, certain other types of queries might benefit a lot from considering all possible plans. The space of alternative plans must therefore be adjustable for each type of query. This leads to the concept of region based query optimization. In [Mit 93, MZD 92, MDZ 93, MDZ 94] an architecture for region based extensible processing and optimization of queries is proposed. An Epoq optimizer is a collection of concurrently available region modules, each of which embodies one strategy for the optimization of query expressions. Different regions often accomplish ....

Mitchell, G., Dayal, U., and Zdonik, B.S., "Control of an Extensible Query Optimizer: A PlanningBased Approach", Proc. of Intl. Conf. on Very Large Databases, 1993.

....dans un langage d eclaratif mais cod ees directement dans l optimiseur. Pour simplifier le probl eme du controle de l application des r egles dans le contexte objet o u le nombre de transformations est tr es grand, nous avons opt e pour une architecture modulaire, comme celle propos ee dans [MZD93] D autre part, le choix du meilleur plan d ex ecution est effectu e en fonction d un mod ele de cout. Un autre objectif important a et e de construire une bonne plateforme pour exp erimenter les el ements principaux d un optimiseur: les transformations, les heuristiques, les strat egies de ....

....d une assertion, une nouvelle transformation s emantique est ajout ee a la biblioth eque des transformations el ementaires. Pour eviter le controle complexe de l application des transformations el ementaires, celles ci sont regroup ees dans des modules, comme dans l optimiseur Epoch [MZD93] Les transformations utilisant le meme type de connaissances sont donc regroup ees dans le meme module. Par exemple, les principaux modules de l optimiseur Flora sont la simplification, la r e ecriture s emantique, l ordonnancement d op erations, la r e ecriture alg ebrique et l optimisation ....

G. Mitchell, S. Zdonick, and U. Dayal. Control of an extensible query optimizer: A planning-based appoach. In Proc. Int. Conf. on Very Large Data Bases, Dublin, Ireland, August 1993.

.... world: GD87, Bat86, BMG93, GM93] are proposals for optimizer generators; Fre87] has made rule based optimization popular in the relational context, which was adopted by the object oriented community [OS90, KM90, CD92] as well as by researchers in extended relational systems [BG92, Loh88] [MDZ93, KMP93] proposed architectural frameworks for query optimization. While this (rather incomplete) list indicates the numerous efforts invested in query optimization, it is interesting to observe that all proposals are based on the assumption that selection predicates are conjunctively combined. If not, ....

....query evaluation plans for bypass processing. This optimizer is an instantiation of our previously developed architectural framework for object oriented query optimization [KMP93] This architecture is centered around a blackboard structure divided into regions an idea also proposed by [MDZ93] which contain incomplete query evaluation plans (QEPs) QEPs are composed from basic building blocks which are, step by step, augmented to complete evaluation plans. This building block approach also used for global query optimization [Sel86] incorporates the factorization of common ....

G. Mitchell, U. Dayal, and S.B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 517--528, Dublin, Ireland, Aug 1993.

....performance) or to generate a monolithic optimizer. The algorithm for the latter is not described, so the performance of the resulting optimizer is hard to predict. Also, it is not evident whether the general nature of the framework makes it hard and difficult to use. Mitchell, Dayal, and Zdonik [16, 17] propose a framework called Epoq in which optimizers are constructed using extensible regions. A region is defined by a stated goal (e.g. lower cost, join reorder, etc. Each region defines a control strategy that transforms a query into alternative forms based on its internal transformation ....

Gail Mitchell, Umeshwar Dayal, and Stanley B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proceedings 19th International Conference on Very Large Data Bases, pages 517--528, Dublin, August 1993.

....enables spatio temporal reasoning. Another future work is to build a video query language based on the CVOT model. The queries can be translated into the TIGUKAT query calculus and then query algebra. Therefore, it is possible to optimize these queries using object query optimization techniques [MDZ93, OB95] ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an extensible query optimizer: A planningbased approach. In Proceedings of the 20th International Conference on Very Large Data Bases, pages 517---528, Dublin, Ireland, August 1993.

....in a rule body) This ability is useful when we have a number of rule systems controlled by another rule system. In fact we may have a hierarchy of rule systems where a non leaf rule system controls the evaluation of its children. This framework is similar to the one described by Mitchell, et al. [11]. 6 An Optimizer for an Object Oriented Language In this section we specify an optimizer for object oriented queries expressed in the ZQL[C ] language [1] ZQL[C ] is an SQL based object query language designed to be well integrated with C . Our approach to solving this optimization problem is ....

....system for query optimizers is more comprehensive and more declarative than other systems to date. We intend to produce additional optimizer specifications to validate this hypothesis, particularly ones that use attributes to control optimization context, in the manner of Mitchell, et al. [11]. We also intend to experiment with different search engines that support cost based pruning. Another interesting research area is the development of specialized libraries of support functions to aid the optimizer specification process within a specific domain of applications. Our specification ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an Extensible Query Optimizer: A Planning-Based Approach. In Proceedings of the 19th VLDB Conference, August 1993. To appear.

....be changed or replaced. For example, a System R style bottom up optimizer cannot be implemented using the Cascades Framework. Various architectures have been proposed to allow extensible control over the search strategy of an optimizer. The region based optimizer architecture of Mitchell et al. MDZ93] the modular optimizer architecture by Sciore and Sieg [SJ90] the blackboard architecture of Kemper, Moerkotte and Peithner [KMP93] are all based on the concept of dividing an optimizer into regions that carry out different parts of the optimization. A query then has to pass through these ....

....regions that carry out different parts of the optimization. A query then has to pass through these various regions to be optimized. These architectures differ in the methods used to pass control between the various regions. In [SJ90] control passes from one region to another in a fixed sequence. MDZ93] uses a hierarchy of regions in which the parent region dynamically controls the sequence of regions through which the query passes while being optimized. In the blackboard approach [KMP93] knowledge sources are responsible for moving the queries between regions. All these architectures describe ....

Gail Mitchell, Umeshwar Dayal, and Stanley B. Zdonik. "Control of an Extensible Query Optimizer: A Planning Based Approach". In Proc. of the 19th VLDB Conf., Dublin, Ireland, 1993.

.... in the object oriented context, as e.g. OS90, KM90, CD92] Many researchers have worked on optimizer architectures that facilitate flexibility: Bat86, GD87, BMG93, GM93] are proposals for optimizer generators; HFLP89, BG92] describe extensible optimizers in the extended relational context; MDZ93, KMP93] propose architectural frameworks for query optimization in object bases. Besides these works on optimizer architectures, optimization strategies for both traditional and next generation database systems are being developed. LMS94] introduces a technique for moving predicates across ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 517--528, Dublin, Ireland, 1993.

....is not described, so the performance of the resulting optimizer is hard to predict. Also, it is not evident whether the general nature of the framework makes it hard and unwieldy to use. If so, then it defeats the purpose of an extensible optimizer generator paradigm. Mitchell, Dayal, and Zdonik [39, 40] propose a framework called Epoq in which optimizers are constructed using extensible regions. A region is defined by a stated goal (e.g. lower cost, join reorder, etc. Each region defines a control strategy that transforms a query into alternative forms based on its internal transformation ....

....encapsulation of rule sets in layers, we can foresee each layer encapsulating a rule engine. Thus, the need for a global search strategy disappears, and each layer implements its own search strategy (in addition to its own search space and cost model) This is similar to the frameworks proposed in [39, 40, 44]. One interesting problem is the compaction of layers namely, how can we generate a monolithic rule set with a search strategy that is semantically equivalent to the composition of the search strategies in the different layers Since each layer can have a different rule engine, any compaction ....

Gail Mitchell, Umeshwar Dayal, and Stanley B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proceedings 19th International Conference on Very Large Data Bases, pages 517--528, Dublin, August 1993.

.... adopted in the object oriented context [OS90, KM90, CD92] Many researchers have worked on optimizer architectures that facilitate flexibility: GD87, Bat86, BMG93, GM93] are proposals for optimizer generators; BG92, HFLP89] described extensible optimizers in the extended relational context; MDZ93, KMP93] proposed architectural frameworks for query optimization in object bases. This work was supported by the German Research Council under contracts DFG Ke 401 6 1 and SFB 346. HS93] pointed out that the ordering of the selection predicate evaluation is particularly important in the ....

....query evaluation plans for bypass processing. This optimizer is an instantiation of our previously developed architectural framework for object oriented query optimization [KMP93] This architecture is centered around a blackboard structure divided into regions an idea also proposed by [MDZ93] which contain incomplete query evaluation plans (QEPs) QEPs are composed from basic building blocks [Loh88] which are, step by step, augmented to complete evaluation plans. This building block approach also used for global query optimization [Sel86] incorporates the factorization of ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an extensible query optimizer: A planningbased approach. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 517--528, Dublin, Ireland, August 1993.

....facilitates porting the query module to work on top of other OODBMSs. The Open OODB query processor includes a query execution engine containing efficient implementations of scan, indexed scan, hybrid hash join [Shapiro 1986] and complex object assembly [Keller et al. 1991] The EPOQ project [Mitchell et al. 1993] is another approach to query optimization extensibility, where the search space is divided into regions. Each region corresponds to an equivalent family of query expressions that are reachable from each other. The regions are not necessarily mutually exclusive (Figure 2) and differ in the queries ....

....The general strategies are not likely to change, but the tuning of the parameters and the definition of the space of acceptable solutions should be expected to change. It is also interesting to note the surface similarity between randomized search algorithms and the regions approach proposed by Mitchell, et al. [1993]. Further studies are required to establish the relationship more firmly. Cost Function Typical cost functions used in query optimization take into account the various costs that are incurred in processing the query. In non distributed systems, this is typically the I O and CPU cost, while in ....

MITCHELL, G., DAYAL, U., AND ZDONIK, S. 1993. Control of an extensible query optimizer: A planning-based approach. In Proc. 19th Int. Conf. on Very Large Databases, 517--528.

.... in the object oriented context, as e.g. OS90, KM90, CD92] Many researchers have worked on optimizer architectures that facilitate flexibility: Bat86, GD87, BMG93, GM93] are proposals for optimizer generators; HFLP89, BG92] described extensible optimizers in the extended relational context; MDZ93, KMP93] proposed architectural frameworks for query optimization in object bases. Besides these works on optimizer architectures, optimization strategies for both traditional and nextgeneration database systems are being developed. LMS94] introduces a technique for moving predicates across ....

G. Mitchell, U. Dayal, and S. B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proc. of the Conf. on Very Large Data Bases (VLDB), pages 517-- 528, Dublin, Ireland, 1993.

....Dayal. The EQUAL query algebra and related results in query equality and equivalence were developed under the supervision of Zdonik. My optimization results were developed in collaboration with Zdonik and Dayal. The presentation in this dissertation is based in part on the published papers [105, 108, 132, 133, 134] 1 , an invited report [154] and Brown University reports [104, 107] The publications [132, 133, 134, 154] report on my work with Zdonik; the others on my work with Dayal and Zdonik. Mitchell (Shaw) is the primary author on all papers with the exception of [154] which describes primarily the ....

....paper on query representation is the basis for the presentation in Chapter 7. 108] This paper, presented at the 1992 Hawaii International Conference on System Sciences, is a preliminary description of our work on query optimization. This work maps to the presentation of Sections 5.1 and 5.2. [105] This paper describing the control problem and planning based approach to control has been accepted for publication in the Proceedings of the 19th International Conference on Very 1 Most of the research related to the EQUAL query algebra is published under my former name, Gail Mitchell Shaw. 2 ....

[Article contains additional citation context not shown here]

Gail Mitchell, Umeshwar Dayal, and Stanley B. Zdonik. Control of an Extensible Query Optimizer: A Planning-Based Approach. In Proceedings of the 19th VLDB Conference, August 1993. To appear.

....of list, and we believe that we will have little difficulty simulating the ODMG arrays with AQUA lists. Our view of predicates, however, is significantly more powerful. We are currently developing a cost model and incorporating the list and tree algebras into the EPOQ extensible query optimizer [22] being developed at Brown University. Acknowledgements We would like to thank Catriel Beeri, Mitch Cherniack, Darryn Lavery, Gail Mitchell, Arnold Rosenberg, Dennis Shasha, Hagit Shatkay, Sairam Subramanian, Bennet Vance, and Jason Wang for useful discussions. ....

G. Mitchell, U. Dayal, and S. B. Zdonik, "Control of an Extensible Query Optimizer: A Planning-Based Approach ," Proc. VLDB (1993), 517--528.

No context found.

G. Mitchell, U. Dayal, and S.B. Zdonik. Control of an extensible query optimizer: A planning-based approach. In Proc. 19th Int. Conf. on Very Large Data Bases, pages 517--528, 1993.

No context found.

Gail Mitchell, Umeshwar Dayal, and Stanley B. Zdonik. Control of an Extensible Query Optimizer: A Planning-Based Approach. In Proc. Intl. Conference on Very Large Databases, VLDB, August 1993.

No context found.

Gail Mitchell, Umeshwar Dayal, and Stanley B. Zdonik, "Control of an Extensible Query Optimizer: A Planning-Based Approach", Proc. 19th VLDB Conf., 517528, August, 1993.

No context found.

G. Mitchell, U. Dayal, and S. Zdonik, Control of an extensible query optimizer: a planning-based approach, Proc. International Conference on Very Large Data Bases, Dublin, (1993).

First 50 documents

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