Search engine for discovering works of Art, research articles, and books related to Art and Culture
ShareThis
Javascript must be enabled to continue!

Design and Implementation of a TCP Long Connection Load Balancing Algorithm Based on Negative Feedback Mechanism

View through CrossRef
Abstract Dynamic load balancing is an important way to improve the resource utilization and parallel computing performance of multi-server system. For the mass of terminals under the Internet of Things scenario high frequency periodic escalation and receiving messages scene, the terminal equipment and message server need to establish a TCP long connection to communicate, the accumulation of messages different degree of weight will cause the load tilt. In this paper, an algorithm based on the socket buffer feedback mechanism is proposed, which takes the server load (CPU, memory, connection rate and socket buffer cache) as the load indicator, collects the load indicator through the load agent and calculates the discrete coefficient of the server socket buffer. The dispatch controller calculates the weight of each server, and the higher the weight, the lighter the server load. The experiments show that the CMTS more realistically feedbacks the server load status than the least connections. As the number of cluster server connections and message sending rate increase, the average cluster response time is shorter.
Title: Design and Implementation of a TCP Long Connection Load Balancing Algorithm Based on Negative Feedback Mechanism
Description:
Abstract Dynamic load balancing is an important way to improve the resource utilization and parallel computing performance of multi-server system.
For the mass of terminals under the Internet of Things scenario high frequency periodic escalation and receiving messages scene, the terminal equipment and message server need to establish a TCP long connection to communicate, the accumulation of messages different degree of weight will cause the load tilt.
In this paper, an algorithm based on the socket buffer feedback mechanism is proposed, which takes the server load (CPU, memory, connection rate and socket buffer cache) as the load indicator, collects the load indicator through the load agent and calculates the discrete coefficient of the server socket buffer.
The dispatch controller calculates the weight of each server, and the higher the weight, the lighter the server load.
The experiments show that the CMTS more realistically feedbacks the server load status than the least connections.
As the number of cluster server connections and message sending rate increase, the average cluster response time is shorter.

Related Results

Single / multi TCP and a fairness issue
Single / multi TCP and a fairness issue
In distributed systems domain a good connection is a structural quest for messages transfer between clusters in today's high performance applications. Initiating a single TCP used ...
Predictors of False-Negative Axillary FNA Among Breast Cancer Patients: A Cross-Sectional Study
Predictors of False-Negative Axillary FNA Among Breast Cancer Patients: A Cross-Sectional Study
Abstract Introduction Fine-needle aspiration (FNA) is commonly used to investigate lymphadenopathy of suspected metastatic origin. The current study aims to find the association be...
Polycomb requires Tcp-1η chaperonin for maintaining gene silencing in Drosophila
Polycomb requires Tcp-1η chaperonin for maintaining gene silencing in Drosophila
Abstract In metazoans, heritable states of cell type specific gene expression patterns linked with specialization of various cell types constitut...
An epistemic justice account of students’ experiences of feedback
An epistemic justice account of students’ experiences of feedback
I am a storyteller. I believe in the power of stories to share experiences and to elucidate thoughts and ideas and to help us to make sense of complex social practices. This thesis...
Thrombocytopenia in neonatal intensive care unit at a tertiary care hospital of South Punjab, Pakistan.
Thrombocytopenia in neonatal intensive care unit at a tertiary care hospital of South Punjab, Pakistan.
Objective: To find out fetal and maternal factors linked with thrombocytopenia (TCP) among neonates admitted in neonatal intensive care unit (NICU) of a tertiary care hospital of S...
Recovery of loss of packet in network using constant packet reordering
Recovery of loss of packet in network using constant packet reordering
When the packet is reordered the most standard implementation of the Transmission control protocol (TCP) gives poor performance. This paper proposes a new version of the TCP which ...
Value-based cost-cognizant test case prioritization for regression testing
Value-based cost-cognizant test case prioritization for regression testing
Software Test Case Prioritization (TCP) is an effective approach for regression testing to tackle time and budget constraints. The major benefit of TCP is to save time through the ...
Designing rich feedback encounters
Designing rich feedback encounters
Feedback is a cornerstone of effective learning, yet it remains one of the most persistently complex challenges in higher education, for educators and students alike. This workshop...

Back to Top