Browse Prior Art Database

Smart detecting and scheduling Query workload as none-transaction in a Transaction Processing system

IP.com Disclosure Number: IPCOM000244627D
Publication Date: 2015-Dec-30
Document File: 4 page(s) / 78K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method to detect and schedule query workload as none-transaction in a transaction processing system is disclosed. The method includes identifying query workload based upon resource access history, maintaining a query transaction list table, and scheduling the query workload as a none transaction thread where no transactional environment is maintained for this thread, e.g. no logging or locking.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 4

Smart detecting and scheduling Query workload as none

Smart detecting and scheduling Query workload as none-

Transaction processing(TP) system is widely used as a really critical system and platform to support core business around the world. TP system deliberately designs its locking or serialization mechanism to improve the throughput to the maximum extent and maintain data consistency and isolation in the meanwhile.

There are currently two types of workload running in TP system which are treated equally. One type is update workload, the other type is query workload. In a query request, nothing would be changed and the request is idempotent. Running query request as a transaction would consume a big portion of the system resource.

The core idea of the disclosure is to equip TP system with the intelligence to identify query workload by resource access history and then to schedule the query in a none transaction instance.

The advantage of this invention is that existing application assets won't need to be changed and transaction management is still transparent to application developers. As the data is in one single copy, no data synchronization is necessary and data consistency could be guaranteed. Also, for the customers using mature TP systems, system performance can be improved while the cost is reduced.

The implementation of the invention contains two main steps:


1. Query detecting

TP needs to maintain a in-storage table to store a transaction list that belongs to Query workload and the conditions if any. The list is updated during transaction commit process.

A READONLY information is feedback from recoverable resource o...