Unlock instant, AI-driven research and patent intelligence for your innovation.

A high-availability cluster system load dynamic balancing method and node device

A cluster system, dynamic balancing technology, applied in transmission systems, digital transmission systems, electrical components, etc., can solve problems such as heavy load, network load of faulty nodes, reduce cluster access concurrency, etc., to reduce the possibility of node failures , the effect of improving concurrency

Active Publication Date: 2017-02-08
INSPUR BEIJING ELECTRONICS INFORMATION IND
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] A high-availability cluster system based on CTDB (such as figure 1 shown) can ensure that the application will not be interrupted when some interface nodes fail, and because it does not have load balancing function, high-performance computing fields, such as petroleum, exploration, earthquake, high-energy physics, spatial information processing and other large-scale, high-concurrent access or Data-intensive applications can easily lead to overloading of individual interface nodes in the cluster, resulting in node downtime, reducing cluster access concurrency and affecting client access performance
[0003] Commonly used load balancing methods are RR-DNS and LVS, in which RR-DNS cannot perceive the load status of cluster nodes, and the scheduling process for faulty nodes will not only cause unnecessary network load, but also increase the system fault-tolerant response time
In addition, if multiple domain name requests are mapped to the same IP address within a TTL, it will cause obvious load imbalance; while LVS only supports Linux applications, it is powerless for a wide range of Windows applications

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
  • A high-availability cluster system load dynamic balancing method and node device
  • A high-availability cluster system load dynamic balancing method and node device
  • A high-availability cluster system load dynamic balancing method and node device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] The inventor of the present application proposes that DLB-HA (Dynamic Load Balancing in High Available cluster) modules can be added to each interface node of the high-availability cluster system to obtain the number of process connections between the interface node and the client in real time. Various resource utilization conditions (various resources include at least CPU, memory, disk utilization and network traffic values), and then use a weighting algorithm to calculate the load index of the interface node, and determine whether the interface client's access request is based on this load index. So that when the client process accesses the high-availability cluster system, it can first select those interface nodes with a smaller load index value (that is, the interface node with a smaller load) for data processing, in order to achieve the purpose of dynamic load balancing, and improve the client Access performance, while ensuring the stability and reliability of the hi...

Embodiment 2

[0047] This embodiment provides a method for dynamic load balancing of a high-availability cluster system, such as figure 2 As shown, it includes the following steps 100 to 300:

[0048] Step 100: Each interface node periodically samples itself, and the sampling information includes acquiring the number of process connections between the interface node and the client and various resource utilization conditions of the interface node;

[0049] The resource utilization of the above interface node includes one or more of the following:

[0050] CPU utilization, memory utilization, disk utilization, network traffic value.

[0051] Step 200: When the interface node receives the access request initiated by the client, it performs a weighted summation of the sampling information of the interface node to calculate the load index of the interface node;

[0052] In this step, it is possible to perform a weighted sum operation on all the sampling information obtained from the most recent sampling ...

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 discloses a dynamic load balancing method of a high-availability cluster system, and node equipment, and relates to a cluster system. The method comprises the steps that each interface node performs self-periodic sampling, wherein sampling information comprises course linking numbers among the interface nodes and clients, and various resource utilization situations of the interface nodes; when the interface nodes receive access requests initiated by the clients, the sampling information of the interface nodes is subjected to weighted summation, and load indexes of the interface nodes are computed; if the computed load indexes exceed a set threshold, the access requests initiated by the clients are rejected; if not, the access requests initiated by the clients are accepted; and connection is established with the clients. The invention further discloses the node equipment in the high-availability cluster system. With the adoption of the technical scheme, the concurrency, the responsiveness and the reliability of the cluster system are improved.

Description

Technical field [0001] The invention relates to a cluster system, in particular to a dynamic load balancing scheme in a high-availability cluster system. Background technique [0002] High-availability cluster system based on CTDB (such as figure 1 (Shown) can ensure that the application will not be interrupted when some interface nodes fail, and because it does not have the load balancing function, high-performance computing fields such as petroleum, exploration, seismic, high-energy physics, spatial information processing and other large-scale, high-concurrent access or Data-intensive applications can easily cause excessive load on individual interface nodes in the cluster, resulting in node downtime, reducing cluster access concurrency and affecting client access performance. [0003] Commonly used load balancing methods are RR-DNS and LVS. RR-DNS cannot perceive the load status of cluster nodes. The scheduling process of faulty nodes will not only cause unnecessary network load...

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): H04L12/803H04L29/08
Inventor 王静范明彬
Owner INSPUR BEIJING ELECTRONICS INFORMATION IND
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com