H. M. Dewan. Runtime Reorganization of Parallel and Distributed Expert Database Systems. Technical report, Department of Computer Science, Columbia University, April 1994. Ph.D. Thesis. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Predictive Dynamic Load Balancing of Parallel and Distributed.. - Hasanat Dewan (1994) (4 citations) Self-citation (Dewan) (Correct)
....scheme, issues of data (load) partitioning, both static and dynamic, must be addressed in a comprehensive manner. The PARADISER environment provides these facilities for PARULEL programs. The task partitioning and load balancing techniques used in PARADISER has been the subject of other reports [8, 4, 6]. Here we provide an outline appropriate for the present paper. 4.1 Memory Model The database support for PARULEL programs is provided by an RDBMS, rather than the LISP environment of typical AI rule languages such as OPS5. Within the PARADISER environment, PARULEL rules are compiled into an ....
....We represent the processing potential of the available processing sites by a vector of loading weights . To handle skewed distributions we generalize the copy and constrain scheme as follows. For each rule, a restricted relation and an RA that selects that relation are chosen heuristically [4] at compile time. The database environment is configured automatically by the runtime system to maintain statistics on the RA of each rule by means of appropriately defined triggers. Discrete frequency distributions (DFDs) are maintained by the triggers that are installed for each RA. The weighted ....
[Article contains additional citation context not shown here]
H. M. Dewan. Runtime Reorganization of Parallel and Distributed Expert Database Systems. Technical report, Department of Computer Science, Columbia University, April 1994. Ph.D. Thesis.
Predictive Dynamic Load Balancing of Parallel.. - Dewan, Hernandez.. (1994) (1 citation) Self-citation (Dewan) (Correct)
....buckets (i.e. those with the same hash number) If Prior Work Hetero. Temporal Var DLB Skew Env. Monitoring Meta Data Kitsuregawa [13, 12] No No No No No Schneider DeWitt [15, 8] No No Yes No No Wolf Yu Turek Dias [19, 18] No No Yes No No DeWitt and Naughton [9] No No Yes No Yes Present Work [2] Yes Yes Yes Yes Yes Table 1: Comparison of Parallel Join Methods buckets are small relative to the entire relation, significant savings in computation can result. The joining of corresponding buckets may be carried out either by a nested loop algorithm, or a hash probe method. In the latter, one ....
H. M. Dewan. Runtime Reorganization of Parallel and Distributed Expert Database Systems. Technical report, Dept. of Computer Science, Columbia University, May 1994. Ph.D. Thesis.
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