Browse Prior Art Database

A Method and System for Optimal Routing of Input/Output Data for Business Intelligence Queries

IP.com Disclosure Number: IPCOM000240794D
Publication Date: 2015-Mar-03
Document File: 3 page(s) / 156K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method and system is disclosed for optimal routing of Input/ Output (I/O) data for Business Intelligence (BI) queries.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 3

A Method and System for Optimal Routing of Input/Output Data for Business Intelligence Queries

Typically, Business Intelligence (BI) queries require access to large amounts of disk-based data. Customers of a large data center have started to run BI workloads on the same systems as online transaction processing (OLTP) workloads. During critical periods, the system performance of OLTP workloads may be negatively impacted if BI query data is dragged through symmetric multi-processing (SMP) node-to-node busses. Typically, computing systems may not monitor data bus usage and may not dynamically alter routing of disk-based data that is moving from and to the I/O system.

Disclosed is a method and system for optimal routing of Input/ Output (I/O) data for Business Intelligence (BI) queries. The method and system utilizes real time data on node-to-node bus with knowledge of an I/O layout to optimally route disk-based data to other I/O components. The other I/O components can include accelerator cards.

The system includes a hardware and/or firmware that monitors critical bus utilization

within a computing server in real-time. The method and system provides algorithms that utilize data to generate updated routing tables which needs to be transmitted to an I/O network switch. When updated routing tables are activated, the updated routing tables enable I/O network switches to route data. Due to this, the system avoids transmitting the data across critical internal bus links. The critical internal bus links may be currently experiencing high utilization.

In an implementation, the method and system includes peripheral component interconnect (PCI) switches designed to allow quick updates to the algorithms. The method and system can be used with other protocols and switches in the I/O network. The hardware is added to critical internal consumer electronics control (CEC) busses to allow periodic samples of the bus utilization. The hardware has the ability to program the length of time used as a basis for the average bus utilization. The average bus utilization by the hardware is 10 microseconds (usec) period (programmable). A

software thread, either in...