Host group decision system in high availability cluster and switching method for host group decision system

A group decision-making and intra-cluster technology, applied in the field of data processing, can solve problems such as complex paxos algorithm, difficult program implementation, and inability to effectively solve livelocks, so as to ensure healthy operation and data consistency, and realize host switching and failure Isolate and resolve cluster split-brain effects

Inactive Publication Date: 2014-03-19
SHANGHAI STOCK EXCHANGE
View PDF4 Cites 47 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problem of collective split-brain in the host state monitoring cluster decision-making and switching process in the prior art, but the paxos algorithm is more complicated, the program implementation is difficult, and the problem of livelock cannot be effectively solved. Improve the overall performance of the server, suitable for fault detection, diagnostic decision-making, fault isolation and swi

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Host group decision system in high availability cluster and switching method for host group decision system
  • Host group decision system in high availability cluster and switching method for host group decision system
  • Host group decision system in high availability cluster and switching method for host group decision system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] figure 1 It shows a typical deployment method based on high-availability multi-machine backup technology in which multiple machines are hot standby for each other. In the figure, there are 9 hosts forming a cluster, which consists of sequence server hosts, communication server hosts, application server hosts, and the hosts. Data and control messages are exchanged through TCP / IP network interconnection, and access to disks is shared between hosts through a storage area network.

Embodiment 2

[0032] Such as figure 2 As shown, it is a schematic diagram of the host state of the present invention. The task can have multiple states. The high-availability module responsible for the conversion of the host state is completed by three modules, including: a probe module, a decision module and a routing module. The probe module periodically accesses the idle status of the host’s applications and resources, and broadcasts the health status of the host in the high-availability cluster; the decision-making module receives the health heartbeat message sent by the probe module, through active and passive ways, Determine the health status of the host; the routing module loads the static routing table and maintains the update of the dynamic routing table. The probe module broadcasts the detected health status of the local machine through the TCP / IP network in the high-availability cluster, and the decision-making module receives the health heartbeat message sent by the probe modul...

Embodiment 3

[0036] image 3 Shows the main functional modules of the transaction host. Decision-making module: Receive the health heartbeat message from the probe module, and modify the status and routing rules of the host according to the health of the host; Send healthy heartbeat messages by means of routing module: Load static routing information when starting, and dynamically update routing information according to the instructions of the decision-making module when the cluster is running, and complete the switching and recovery of the host.

[0037] Between each module of the same host, adopt the way of inter-process communication to transfer messages, and use TCP / IP mode between different hosts to communicate between hosts.

[0038] According to the particularity of the securities itself, the products that enter the transaction host are divided into different product sets. For a product set, there can be one host as the main processor, and multiple hosts as backup processors, which...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention relates to the data processing field and particularly relates to a host group decision system in a high availability cluster and a switching method for the host group decision system. The host group decision system is characterized in that the host group decision system is at a system background and comprises a plurality of transaction hosts, an integral structure comprises three levels of a sequencing-level server host, a communication-level server host and a processing-level server host, communication among the various hosts formed the cluster can be carried out through a network, reading-writing access to a file system is carried out through sharing storing equipment, real-time synchronization of data among the hosts is carried out to realize consistency, a high availability module used for converting host states is cooperatively accomplished through three modules which are a probe module, a decision module and a route module. Compared with the prior art, the host group decision system improves integral performance of a server and is advantaged in that physical isolation switching of a fault point is carried out through a two dimensional table cross collective vote decision mode, so host switching and fault isolation are realized, and a problem of cluster brain cracking is effectively solved.

Description

[technical field] [0001] The invention relates to the field of data processing, in particular to a host group decision system and switching method in a highly available cluster. [Background technique] [0002] The stock exchange provides a public trading platform for the securities market. The securities core trading system, as a platform system for real-time bidding and matching of securities products, is also called a bidding and matching platform. It is a key business system, and its safety and reliability are directly related to the domestic financial system. The market is prosperous and stable, so the bidding platform needs to ensure stability and high availability. [0003] According to Gartner's statistics, the causes of unplanned system downtime are mainly application problems (40%), operation problems (40%), operating system failures (10%) and hardware failures (10%), and these problems are basically in a It can be solved in the running system, and only a few probl...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): H04L12/24
Inventor 郭鹏光武剑锋王泊张佳岭朱星垠黄寅飞白硕
Owner SHANGHAI STOCK EXCHANGE
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap