An evaluation of Starburst's memory resident storage component

Lehman, Tobin J.; Shekita, Eugene J.; Cabrera, Luis-Felipe

doi:10.1109/69.180606

Cited by 59 publications

(21 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The arguments by some researchers stated that the whole database will soon fit in memory for certain applications because of the declining price of memory for their higher capacity counter-parts [9,15]. So there are various memory-oriented database techniques have been studied such as memory-oriented record format, page structures, indexing techniques and various memoryresident DBMSs have been developed [10,11,12,13]. They perform significantly better than disk-oriented DBMSs even when the disk-oriented DBMSs have all data in memory, because they have data structures and algorithms which are fitted only into the memory.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Two-Tier Storage DBMS for High-Performance Query Processing

Eo¹,

Li²,

Kim³

et al. 2008

Journal of Information Processing Systems

View full text Add to dashboard Cite

This paper describes the design and implementation of a two-tier DBMS for handling massive data and providing faster response time. In the present day, the main requirements of DBMS are figured out using two aspects. The first is handling large amounts of data. And the second is providing fast response time. But in fact, Traditional DBMS cannot fulfill both the requirements. The disk-oriented DBMS can handle massive data but the response time is relatively slower than the memory-resident DBMS. On the other hand, the memory-resident DBMS can provide fast response time but they have original restrictions of database size. In this paper, to meet the requirements of handling large volumes of data and providing fast response time, a two-tier DBMS is proposed. The cold-data which does not require fast response times are managed by disk storage manager, and the hot-data which require fast response time among the large volumes of data are handled by memory storage manager as snapshots. As a result, the proposed system performs significantly better than diskoriented DBMS with an added advantage to manage massive data at the same time.

show abstract

Section: Introductionmentioning

confidence: 99%

“…These problems can be solved using a multi-level storage system [13,14]. This system stores data in several deferent levels.…”

Section: Introductionmentioning

confidence: 99%

Two-Tier Storage DBMS for High-Performance Query Processing

Eo¹,

Li²,

Kim³

et al. 2008

Journal of Information Processing Systems

View full text Add to dashboard Cite

show abstract

“…Starburst IMS attachment [9] allows links between tuples. If the main-memory storage method [32] is exploited, such links are main-memory pointers that provide comparable performance to that of OODBMSs. However, in RDBMSs, the type of the link, disk or main memory, does not change, regardless of the data being on disk or in main memory.…”

Section: R Elated Workmentioning

confidence: 99%

Composite-object views in relational DBMS: An implementation perspective

Pirahesh

Mitschang

Südkamp

et al. 1994

Advances in Database Technology — EDBT '94

View full text Add to dashboard Cite

-We present a novel approach for supporting Composite Objects (CO) as an abstraction over the relational data. This approach brings the advanced CO model to existing relational databases and applications, without requiring an expensive migration to other DBMSs which support CO. The concept of views in relational DBMSs (RDBMS) gives the basis for providing the CO abstraction. This model is strictly an extension to the relational model, and it is fully upward compatible with it. We present an overview of the data model. We put emphasis in this paper on showing how we have made the extensions to the architecture and implementation of an RDBMS (Starburst) to support this model. We show that such a major extension to the data model is in fact quite attractive both in terms of implementation cost and query performance. We introduce a CO cache for navigation through components of a CO. With this technique, the performance of navigation through COs, which has been of a concern in RDBMSs in the past, is in fact quite satisfactory. We present our practical experience in using this facility. We show that our work on CO enables existing RDBMSs to incorporate efficient CO facilities at a low cost and at a high degree of application reusability and database sharability. MOTIVATIONIt is widely agreed now that complex applications, such as design applications, multi-media and AI applications, and even advanced business applications can benefit significantly from database interfaces that support composite (or complex) objects (shortly, CO) [15,23]. A generally accepted characterization defines a CO consisting of several components (possibly from different types) with relationships in between [2, 5, 11, 26, 36, 19}. Interestingly, object oriented DBMSs (OODBMSs) have adopted a very similar model [1,43,14,27]. This is particularly true for OODBMSs used in practice. Especially 00 programming environments have made advances in handling of COs.Such environments facilitate the growth of complex applications. As a result, there is considerable pressure on RDBMSs for better support of COs. To respond to this demand, several systems today are bridging 00 environments with relational. An example of such a system is the Persistence DBMS [20], which builds a layer on the top of RDBMSs, providing better support for COs. These systems essentially extract the data from the relational and port data to 00 environments. · One straightforward way of extracting data with complex structure is to follow the parent/child relationships: for each parent instance, execute a query to get the children; repeat the same thing for each child to get its children, and so on. However, this style of data extraction leads to numerous queries, and does not lend itself to effective set-oriented processing. Essentially, the process of data extraction is broken into fragmented queries where the number of fragments is in the order of number of instances of parent components in the extracted data. A better approach is to employ much more powerful set-oriented p...

show abstract

“…It has been observed that specialized memory-resident techniques allow substantial performance gains over buffer-resident data in a disk-based system [9,13,14]. More recently, [2] describes ways to organize pages in a disk-based database system so that database operations give good CPU performance when the pages are memory resident in the database buffer.…”

Section: Introductionmentioning

confidence: 99%

Conjunctive selection conditions in main memory

Ross

2002

Proceedings of the Twenty-First ACM SIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems - PODS '02

View full text Add to dashboard Cite

We consider the fundamental operation of applying a conjunction of selection conditions to a set of records. With large main memories available cheaply, systems may choose to keep the data entirely in main memory, in order to improve query and/or update performance.The design of a data-intensive algorithm in main memory needs to take into account the architectural characteristics of modern processors, just as a disk-based method needs to consider the physical characteristics of disk devices. An important architectural feature that influences the performance of main memory algorithms is the branch misprediction penalty. We demonstrate that branch misprediction has a substantial impact on the performance of an algorithm for applying selection conditions. We describe a space of "query plans" that are logically equivalent, but differ in terms of performance due to variations in their branch prediction behavior. We propose a cost model that takes branch prediction into account, and develop a query optimization algorithm that chooses a plan with optimal estimated cost. We also develop an efficient heuristic optimization algorithm.We provide experimental results for a case study based on an event notification system. Our results show the effectiveness of the proposed optimization techniques. Our results also demonstrate that significant improvements in performance can be obtained by applying a methodology that takes branch misprediction latency into account.

show abstract

An evaluation of Starburst's memory resident storage component

Cited by 59 publications

References 19 publications

Two-Tier Storage DBMS for High-Performance Query Processing

Two-Tier Storage DBMS for High-Performance Query Processing

Composite-object views in relational DBMS: An implementation perspective

Conjunctive selection conditions in main memory

Contact Info

Product

Resources

About