『互联网架构』软件架构-redis的通信协议（protocol）（52）
redis的通信协议是什么？双方约定了一种编码方式，客户端将要发送的命令进行编码，然后服务端收到后，使用同样的协议进行解码，服务端处理完成后，再次编码返回给客户端，客户端解码拿到返回结果，这样就完成了一次通信。
（一）协议介绍
https://redis.io/topics/protocol
- 易于实现
- 可以高效地被计算机分析（parse）
- 可以很容易地被人类读懂
（二）编码后的字符标记解释
- For Simple Strings the first byte of the reply is “+” 回复
- For Errors the first byte of the reply is “-” 错误
- For Integers the first byte of the reply is “:” 整数
- For Bulk Strings the first byte of the reply is “$” 字符串
- For Arrays the first byte of the reply is “*” 数组
（三）模拟Redis客户端&分片
-
客户端
> Jedis跟redis通讯很简单发送了socket，然后发送了resp这种‘暗语’进行通讯。
import java.io.IOException;
import java.io.InputStream;
import java.net.Socket;
public class RedisClient {
private Socket socket;
private String host;
private int port;
public RedisClient() throws IOException {
this.socket = new Socket("192.168.79.100",6379);
}
public RedisClient(String host, int port) {
this.host = host;
this.port = port;
try {
this.socket = new Socket(host,port);
} catch (IOException e) {
e.printStackTrace();
}
}
//set wk 2018
public String set( String key, String value) throws IOException {
//"*3/r/n$3/r/nset/r/n$2/r/nwk/r/n$4/r/n2018"
StringBuilder stringBuilder=new StringBuilder();
stringBuilder.append("*3").append("/r/n");
stringBuilder.append("$").append(CommandRedis.set.name().length()).append("/r/n");
stringBuilder.append(CommandRedis.set).append("/r/n");
stringBuilder.append("$").append(key.getBytes().length).append("/r/n");
stringBuilder.append(key).append("/r/n");
stringBuilder.append("$").append(value.getBytes().length).append("/r/n");
stringBuilder.append(value).append("/r/n");
socket.getOutputStream().write(stringBuilder.toString().getBytes());
byte[] b=new byte[2048];
socket.getInputStream().read(b);
return new String(b);
}
public String get( String key) throws IOException {
//"*3/r/n$3/r/nset/r/n$2/r/nwk/r/n$4/r/n2018"
StringBuilder stringBuilder=new StringBuilder();
stringBuilder.append("*2").append("/r/n");
stringBuilder.append("$").append(CommandRedis.get.name().length()).append("/r/n");
stringBuilder.append(CommandRedis.get).append("/r/n");
stringBuilder.append("$").append(key.getBytes().length).append("/r/n");
stringBuilder.append(key).append("/r/n");
socket.getOutputStream().write(stringBuilder.toString().getBytes());
byte[] b=new byte[2048];
socket.getInputStream().read(b);
return new String(b);
}
public String setnx(String key, String value) throws Exception {
StringBuffer stringBuffer = new StringBuffer();
stringBuffer.append("*3").append("/r/n");
stringBuffer.append("$5").append("/r/n");
stringBuffer.append("setnx").append("/r/n");
stringBuffer.append("$").append(key.getBytes().length).append("/r/n");
stringBuffer.append(key).append("/r/n");
stringBuffer.append("$").append(value.getBytes().length).append("/r/n");
stringBuffer.append(value).append("/r/n");
socket.getOutputStream().write(stringBuffer.toString().getBytes());
byte[] b = new byte[2048];
socket.getInputStream().read(b );
return new String(b);
}
// 管道
public void pipeline(String key)throws Exception {
StringBuffer stringBuffer = new StringBuffer();
stringBuffer.append("*2").append("/r/n");
stringBuffer.append("$"+CommandRedis.subscribe.name().length()).append("/r/n");
stringBuffer.append(CommandRedis.subscribe).append("/r/n");
stringBuffer.append("$").append(key.getBytes().length).append("/r/n");
stringBuffer.append(key).append("/r/n");
socket.getOutputStream().write(stringBuffer.toString().getBytes());
InputStream inputStream = socket.getInputStream();
while (true) {
byte[] b = new byte[2048];
inputStream.read(b );
System.out.println(new String(b));
}
}
//subscribe
public void subscribe(String key)throws Exception {
StringBuffer stringBuffer = new StringBuffer();
stringBuffer.append("*2").append("/r/n");
stringBuffer.append("$"+CommandRedis.subscribe.name().length()).append("/r/n");
stringBuffer.append(CommandRedis.subscribe).append("/r/n");
stringBuffer.append("$").append(key.getBytes().length).append("/r/n");
stringBuffer.append(key).append("/r/n");
socket.getOutputStream().write(stringBuffer.toString().getBytes());
InputStream inputStream = socket.getInputStream();
while (true) {
byte[] b = new byte[2048];
inputStream.read(b );
System.out.println(new String(b));
}
}
public void close(){
if(socket != null){
try {
socket.close();
} catch (IOException e) {
e.printStackTrace();
}
}
}
}
public enum CommandRedis {
set,get,incr,subscribe
}
-
分片
> RedisClient分片方式，简单模拟，理解概念分片。
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
public class TestRedisClient {
public static void main(String[] args) throws IOException {
/* RedisClient redisClient=new RedisClient();
System.out.println(redisClient.set("wk","悟空"));
System.out.println(redisClient.get("wk"));*/
List<RedisClient> pool=new ArrayList<>();
pool.add(new RedisClient("192.168.0.12",6379));
pool.add(new RedisClient("192.168.0.12",6380));
pool.add(new RedisClient("192.168.0.12",6381));
Crc16Sharding crc16Sharding=new Crc16Sharding(pool);
for (int i=0;i<100;i++){
String key="xx"+i;
RedisClient redisClient=crc16Sharding.crc16(key);
redisClient.set(key,i+"");
System.out.println(redisClient.get(key));
}
}
}
import com.sun.org.apache.xerces.internal.impl.xpath.regex.Match;
import java.util.List;
public class Crc16Sharding {
List<RedisClient> pool;
public Crc16Sharding(List<RedisClient> pool) {
this.pool = pool;
}
/**
* 通过一个key可以定位到一块 节点
* @param key
* @return
*/
public RedisClient crc16(String key){
int num=Math.abs(key.hashCode()%pool.size());
return pool.get(num);
/* if(key.length()<3){
return pool.get(0);
}else if(key.length()<6){
return pool.get(1);
}else{
return pool.get(2);
}*/
}
}
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
public class TestRedisClient {
public static void main(String[] args) throws IOException {
List<RedisClient> pool=new ArrayList<>();
pool.add(new RedisClient("192.168.0.12",6379));
pool.add(new RedisClient("192.168.0.12",6380));
pool.add(new RedisClient("192.168.0.12",6381));
Crc16Sharding crc16Sharding=new Crc16Sharding(pool);
for (int i=0;i<100;i++){
String key="xx"+i;
RedisClient redisClient=crc16Sharding.crc16(key);
redisClient.set(key,i+"");
System.out.println(redisClient.get(key));;
}
}
}
（四）redis使用技巧
- 如果是订单表，取的话也是通过唯一的订单号来进行，不要通过循环去redis里面取。
- 开关类型的放到本地的jvm内存中，尽量减少redis的压力。
- 避免使用慢查询 hgetall
- 存储redis的购物车，有的人都是很多的好几M的，几百个商品，redis存储的内容的压缩。减少通信的宽带。优化存储内容。
- 少用字符串存储，可以存储map格式放入redis。获取redis里面的map，直接通过key就可以取到对应的值。
（五）监控&运维
-
Open-falcon
> http://open-falcon.org/
>open-falcon是小米开源监控系统，强大灵活的数据采集、采集、传输、绘制、查询、报警等。采用全部golang编写，portal和dashboard使用python编写。
-
Redis-migrate-tool
>https://github.com/vipshop/redis-migrate-tool
>2年多没更新了，可以了解下
>Redis-migrate-tool是唯品会开源针对redis数据运维工具。
支持单独、twemproxy集群、redis custer、aof、rdb文件迁移。支持实时迁移、异构迁移、过滤等功能。
PS：介绍了运维工具，redis通信的内容拼接，redis存储的技巧，redis正常的时候没问题，并发量大的时候一定要注意，在大流量的情况下，代码的细节至关重要！
百度未收录
>>原创文章，欢迎转载。转载请注明：转载自IT人故事会，谢谢！
>>原文链接地址：上一篇:已是最新文章
正文到此结束
热门推荐
相关文章
Loading...
![[HBLOG]公众号](https://www.liuhaihua.cn/img/qrcode_gzh.jpg)
