skip to content »

sushisad.ru

Efficiently updating materialized views dblp

efficiently updating materialized views dblp-61

2.4 Multi-Query Optimization Multiple-query optimization (MQO) [13][27]=-=[24]-=- identifies common subexpressions in query execution plans during optimization, and produces globally-optimal plans.

QPipe follows a “one-operator, many-queries” design philosophy. (throughput and response time) when compared to tuple-by-tuple evaluation engines (iterator model) by saving extraneous procedure calls and by improving temporal locality.Citation Context ...characteristics of each operation in the system independently.• As an extension to the techniques presented in this paper we plan to study dynamic, transparent plan alteration techniques (similar to =-=[19]-=-) to create more opportunities for reusing overlapping work. EPILOGUE Database computing arguably represents the most challenging server computing environment, whereas decision support (DSS) instal... same query, thus avoiding extra work when identical queries execute concurrently, with no previous entries in the result cache.The most commonly used technique is to exhaustively search all possible candidate ...Citation Context ...s of the same query, thus avoiding extra work when identical queries execute concurrently, with no previous entries in the result cache.Although the concept of sharing operators is similar to w... Since it is infeasible for the optimizer to predict the query patterns in a multi-query environment, several algorithms base replacement decisions on the observed importance of different pages.

LRU-K =-=[22]-=- and 2Q [18], for instance, improve the performance of the traditional LRU eviction policy by tracking multiple past-page references, while ARC [21] shows similar performance improvements without rely...

Since queries interact with the buffer pool manager through a page-level interface, it is difficult to develop generic po...

Citation Context ...proposed cache manager [29] dynamically decides on which results to cache, based on result computation costs, sizes, reference frequencies, and maintenance costs due to updates.

Since it is infeasible for the optimizer to predict the query patterns in a multi-query environment,...

Citation Context ..of C code (Berkeley DB itself is around 210,000 lines).

2.4 Multi-Query Optimization Multiple-query optimization (MQO) =-=[13]-=-[27][24] identifies common subexpressions in query execution plans during optimization, and produces globally-optimal plans.