Shulou(Shulou.com)06/03 Report--

This article introduces the relevant knowledge of "how to use Java to implement five load balancing algorithms". In the operation of actual cases, many people will encounter this dilemma, so let the editor lead you to learn how to deal with these situations. I hope you can read it carefully and be able to achieve something!

Catalogue

Preface

Concept

Illustrations of several load balancing algorithms

Polling algorithm

Weighted polling method

Weighted random method

Random method

IP_Hash algorithm

Summary

Preface

Load balancer is to solve the concurrency situation, multiple requests to access, the requests are forwarded to each server through the rules agreed in advance. There are several classical algorithms. Before writing these algorithms in java code, let's take a look at the concept of load balancing.

Concept

Cloud load balancer forwards client requests for access to each server through agreed rules in advance. There are several classical algorithms. Let's use Java to implement them.

Illustrations of several load balancing algorithms

The main load balancing algorithms are these in the figure, and before the code is implemented, let's briefly review their concepts.

Polling algorithm

The polling algorithm assigns each new connection request to the next server sequentially, and finally distributes all requests equally to all servers.

Advantages: absolute fairness

Disadvantages: can not be allocated according to the performance of the server, unable to make rational use of server resources.

Package com.monkeyjava.learn.basic.robin;import com.google.common.collect.Lists;import java.util.HashMap;import java.util.List;import java.util.Map;public class TestRound {private Integer index = 0; private List ips = Lists.newArrayList ("192.168.1.1", "192.168.1.2", "192.168.1.3"); public String roundRobin () {String serverIp Synchronized (index) {if (index > = ips.size ()) {index = 0;} serverIp= ips.get (index); / / polling + 1 index + +;} return serverIp;} public static void main (String [] args) {TestRound testRoundRobin = new TestRound (); for (int iTuno

< 10 ;i++){ String serverIp= testRoundRobin.roundRobin(); System.out.println(serverIp); } }} 输出结果: 192.168.1.1 192.168.1.2 192.168.1.3 192.168.1.1 192.168.1.2 192.168.1.3 192.168.1.1 192.168.1.2 192.168.1.3 192.168.1.1 加权轮询法 该算法中，每个机器接受的连接数量是按权重比例分配的。这是对普通轮询算法的改进，比如你可以设定：第三台机器的处理能力是第一台机器的两倍，那么负载均衡器会把两倍的连接数量分配给第3台机器，轮询可以将请求顺序按照权重分配到后端。 package com.monkeyjava.learn.basic.robin;import java.util.ArrayList;import java.util.HashMap;import java.util.List;import java.util.Map;public class TestWeight { private Integer index = 0; static Map ipMap=new HashMap(16); static { // 1.map, key-ip,value-权重 ipMap.put("192.168.1.1", 1); ipMap.put("192.168.1.2", 2); ipMap.put("192.168.1.3", 4); } public List getServerIpByWeight() { List ips = new ArrayList(32); for (Map.Entry entry : ipMap.entrySet()) { String ip = entry.getKey(); Integer weight = entry.getValue(); // 根据权重不同，放入list 中的数量等同于权重，轮询出的的次数等同于权重 for (int ipCount =0; ipCount < weight; ipCount++) { ips.add(ip); } } return ips; } public String weightRobin(){ List ips = this.getServerIpByWeight(); if (index >

= ips.size () {index = 0;} String serverIp= ips.get (index); index + +; return serverIp;} public static void main (String [] args) {TestWeight testWeightRobin=new TestWeight (); for (String server=testWeightRobin.weightRobin ()) (System.out.println (server);})

Output result:

192.168.1.1

192.168.1.3

192.168.1.3

192.168.1.3

192.168.1.3

192.168.1.2

192.168.1.2

192.168.1.1

192.168.1.3

192.168.1.3

Weighted random method

Get the random number with weight, random this kind of thing, can not see absolute, can only look at relative, we do not use index control subscript to poll, only use random to randomly take ip, that is, to achieve the algorithm.

Package com.monkeyjava.learn.basic.robin;import java.util.*;public class TestRandomWeight {static Map ipMap=new HashMap (16); static {/ / 1.map, key-ip,value- weight ipMap.put ("192.168.1.1", 1); ipMap.put ("192.168.1.2", 2); ipMap.put ("192.168.1.3", 4) } public List getServerIpByWeight () {List ips = new ArrayList (32); for (Map.Entry entry: ipMap.entrySet ()) {String ip = entry.getKey (); Integer weight = entry.getValue (); / / according to the weight, the quantity put into list is equal to the weight, and the number of times polled is equal to the weight for (int ipCount = 0; ipCount < weight) IpCount++) {ips.add (ip);}} return ips;} public String randomWeightRobin () {List ips = this.getServerIpByWeight (); / / cyclic random number Random random=new Random (); int index = random.nextInt (ips.size ()); String serverIp = ips.get (index); return serverIp } public static void main (String [] args) {TestRandomWeight testRandomWeightRobin=new TestRandomWeight (); for (int I = 0 political I < 10; iTunes +) {String server= testRandomWeightRobin.randomWeightRobin (); System.out.println (server);}

Output result:

192.168.1.3

192.168.1.3

192.168.1.2

192.168.1.1

192.168.1.2

192.168.1.1

192.168.1.3

192.168.1.2

192.168.1.2

192.168.1.3

Random method

The load balancing method randomly distributes the load to each available server, and selects a server through the random number generation algorithm, which is the simplest, and then the number of calls increases. This algorithm can achieve the number of requests per server close to the average.

Package com.monkeyjava.learn.basic.robin;import com.google.common.collect.Lists;import java.util.List;import java.util.Random;public class TestRandom {private List ips = Lists.newArrayList ("192.168.1.1", "192.168.1.2", "192.168.1.3"); public String randomRobin () {/ / random number Random random=new Random (); int index = random.nextInt (ips.size ()) String serverIp= ips.get (index); return serverIp;} public static void main (String [] args) {TestRandom testRandomdRobin = new TestRandom (); for (int iTunes 0 +) {String serverIp= testRandomdRobin.randomRobin (); System.out.println (serverIp);}

Output

192.168.1.3

192.168.1.3

192.168.1.1

192.168.1.2

192.168.1.1

192.168.1.3

192.168.1.2

192.168.1.3

192.168.1.3

192.168.1.2

IP_Hash algorithm

Hash (ip)% N algorithm, which uses a hash algorithm to assign client source IP requests to different servers according to the hash modulus algorithm.

Advantages: it ensures that the same client IP address will be hashed to the same back-end server until the list of back-end servers is changed. According to this feature, a stateful session session can be established between service consumers and service providers.

Disadvantages: if the server is offline, the source IP routed server IP will become another, if the server does not do session sharing, it will cause session loss.

Package com.monkeyjava.learn.basic.robin;import com.google.common.collect.Lists;import java.util.List;public class TestIpHash {private List ips = Lists.newArrayList ("192.168.1.1", "192.168.1.2", "192.168.1.3"); public String ipHashRobin (String clientIp) {int hashCode=clientIp.hashCode (); int serverListsize=ips.size (); int index = hashCode%serverListsize; String serverIp= ips.get (index) Return serverIp;} public static void main (String [] args) {TestIpHash testIpHash = new TestIpHash (); String servername= testIpHash.ipHashRobin ("192.168.88.2"); System.out.println (servername);}}

Output result

192.168.1.3

The result is the same every time.

This is the end of the tutorial on how to use Java to implement five load balancing algorithms. Thank you for reading. If you want to know more about the industry, you can follow the website, the editor will output more high-quality practical articles for you!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.