Browse Prior Art Database

LIST DATA ENCODING EXTENDED FOR DISTRIBUTED QUERY IN DATABASE APPLICATION DOMAIN

IP.com Disclosure Number: IPCOM000238151D
Publication Date: 2014-Aug-05
Document File: 7 page(s) / 77K

Publishing Venue

The IP.com Prior Art Database

Related People

Rola Zhang: AUTHOR [+2]

Abstract

Techniques are presented herein for list data encoding extended for distributed query in database application domain. These techniques achieve low consumption of list data encoding/decoding for distributed query and are independent of the application data structure.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 39% of the total text.

Page 01 of 7

LIST DATA ENCODING EXTENDED FOR DISTRIBUTED QUERY IN DATABASE APPLICATION DOMAIN

AUTHORS:

Rola Zhang

Sunny Liao

CISCO SYSTEMS, INC.

ABSTRACT

    Techniques are presented herein for list data encoding extended for distributed query in database application domain. These techniques achieve low consumption of list data encoding/decoding for distributed query and are independent of the application data structure.

DETAILED DESCRIPTION

    Presented herein are techniques related to distributed application development. These techniques can be used to design a common query host without involving an application data structure.

    In current distributed query systems, one host handles client requests and returns the related data collected from data nodes to the client. The host needs to know the application data structure to do further data decoding/encoding. This is the main consumption of system resources on the host, and greatly affects host performance.

    These techniques extend data encoding/decoding in order to reduce the consumption of encode/decode resources on hosts and improve the host performance. This may help user applications to achieve higher efficiency in the same environment.

    According to these techniques, the host is not involved in the application data structure. On the host, data encoding/decoding is minimized to reduce the resource consumption. A common encoding/decoding operation is extended between hosts and data nodes, and it is independent of the application data structure.

Copyright 2014 Cisco Systems, Inc.

1


Page 02 of 7

  A typical data encode/decode operation includes:
1) On data nodes

A) Get application data with the query request.

B) Encode one application data to a data block. It is a full encode for application data. Memory copy time is related to the number of items of application data.

C) Get key from application data.

D) Encode the key to data stream. If it is a "slight" encode, only do a memory copy.

E) Encode the data block to data stream. If it is also a slight encode, only do a memory copy.

     F) Repeat B, C, D and E. Send data stream to the host.

2) On hosts:

A) Receive data stream from data nodes.

B) Decode a key from data stream. If it is a slight decode, only do a memory copy.

C) Decode a data block form data stream. If it is a slight decode, only do a memory copy.

D) Repeat B and C.

E) Merge sort with key.

F) Encode data block to response. If it is a slight encode, only do a memory copy.

G) Repeat F.

H) Send response to client.

    FIGs. 1 and 2 below illustrate comparisons between current techniques and the techniques presented herein.

Copyright 2014 Cisco Systems, Inc.

2


Page 03 of 7

FIG. 1

FIG. 2

Copyright 2014 Cisco Systems, Inc.

3


Page 04 of 7

    The design of data encode/decode extended reduces the consumption of data encode/decode, and improves the efficiency to achieve better user experience.

    In practical applications, application data can contain one or multiple items. The key can be one item of application data...