SPARROW 框架redis客户端封装实践
redis 本身有客户端,先抛出来一个问题?为什么要对redis客户端进行二次封装?
大概在11年时侯,第一次接触redis,那时侯研究过redis的各种数据结构,直接拿redis的客户端jedis直接用。公司安排人要对jedis进行封装,当时就很不理解,为什么非要封装一次才可以?
后来自己写框架,意识到一些东西是需要封装的,比如连接的打开和释放,比如一些危险的方法,比如keys * 比如flushdb 等
后来形成了这样的代码结构
T execute(Executor executor, KEY key)throws CacheConnectionException {
ShardedJedis jedis =null;
try {
Long startTime = System.currentTimeMillis();
jedis =this.pool.getResource();.//连接的获取
T result = executor.execute(jedis);
this.pool.returnResource(jedis);//连接的释放
Long endTime = System.currentTimeMillis();
if(this.cacheMonitor!=null) {
this.cacheMonitor.monitor(startTime, endTime, key);
}
return result;
}catch (JedisConnectionException e) {
this.pool.returnBrokenResource(jedis);//出错时连接释放
logger.error(this.getInfo() + SYMBOL.COLON + e.getMessage());
throw new CacheConnectionException(e.getMessage());
}
}
业务层会出现这样的代码
@Override
public Long addToSet(final KEY key,final Object value)throws CacheConnectionException {
//抛出connection exception异常提示业务方捕获处理
return redisPool.execute(new Executor() {
@Override
public Long execute(ShardedJedis jedis) {
return jedis.sadd(key.key(),value.toString());
}
},key);
}
对连接的打开和释放进行了封装,避免业务端忘关链接而导致连接超时。抛出声明式异常,提示业务端处理链接断开的场景。
从业务使用上来讲基本不会有什么问题
但这种结构存在几个问题
1. 业务端存在jedis代码,如果想换jedis客户端,成本很大
2. 如果业务端想换另一种no sql 成本一样很大。
3 监控,当业务足够复杂时,对key的监控,问题排查和热点隔离成为痛点,如果让业务端对每一个 redis的请求点都加代码监控的话,这个成本依然很大。
一般公司业务都是从单体应用到分布式应用,如果某台机器报超时,对业务的key的监控就比较困难,比如
KEY为USER:1 USER:2 ...USER.N 进行监控,如何识别这是一组KEY?
所以对KEY的规范和统一监控就成为痛点
意淫部分
KEY的规范
模块.业务类型:key id
对应数据结构如下:
REDIS 操作接口定义
package com.sparrow.cache;
import com.sparrow.constant.cache.KEY;
import com.sparrow.exception.CacheConnectionException;
import com.sparrow.support.Entity;
import java.util.List;
import java.util.Map;
/**
* @author harry
* @date 2018/1/18
*/
public interface CacheClient {
Map hashGetAll(KEYkey)throws CacheConnectionException;
Map hashGetAll(KEYkey,Class keyClazz, Class dataClazz)throws CacheConnectionException;
Long getHashSize(KEYkey)throws CacheConnectionException;
Long getSetSize(KEYkey)throws CacheConnectionException;
Long removeFromOrderSet(KEYkey, Long from, Long to)throws CacheConnectionException;
Double getScore(KEYkey, Object value)throws CacheConnectionException;
Long getIndexOfOrderSet(KEYkey, Object value)throws CacheConnectionException;
Map getAllWithScore(KEYkey, Class clazz)throws CacheConnectionException;
Long getListSize(KEYkey)throws CacheConnectionException;
String hashGet(KEYkey, String hashKey)throws CacheConnectionException;
T hashGet(KEYkey, String hashKey, Class clazz)throws CacheConnectionException;
Long hashSet(KEYkey, String hashKey, Object value)throws CacheConnectionException;
//order set
Long getOrderSetSize(KEYkey)throws CacheConnectionException;
Long addToSet(KEYkey, Object value)throws CacheConnectionException;
Long addToSet(KEYkey, String... value)throws CacheConnectionException;
Integer addToSet(KEYkey, Iterable values)throws CacheConnectionException;
Long addToList(KEYkey, Object value)throws CacheConnectionException;
Long removeFromList(KEYkey, Object value)throws CacheConnectionException;
Long removeFromSet(KEYkey, Object value)throws CacheConnectionException;
Long addToOrderSet(KEYkey, Object value, Long score)throws CacheConnectionException;
Long removeFromOrderSet(KEYkey, Object value)throws CacheConnectionException;
Boolean existInSet(KEYkey, Object value)throws CacheConnectionException;
Long addToList(KEYkey, String... value)throws CacheConnectionException;
Integer addToList(KEYkey, Iterable values)throws CacheConnectionException;
Long expire(KEYkey, Integer expire)throws CacheConnectionException;
Long delete(KEYkey)throws CacheConnectionException;
Long expireAt(KEYkey, Long expire)throws CacheConnectionException;
String setExpire(KEYkey, Integer seconds, Object value)throws CacheConnectionException;
String set(KEYkey, Entity value)throws CacheConnectionException;
String set(KEYkey, Object value)throws CacheConnectionException;
String get(KEYkey)throws CacheConnectionException;
T get(KEYkey, Class clazz)throws CacheConnectionException;
List getAllOfList(KEYkey)throws CacheConnectionException;
List getAllOfList(KEYkey, Class clazz)throws CacheConnectionException;
Long setIfNotExist(KEYkey, Object value)throws CacheConnectionException;
Long append(KEYkey, Object value)throws CacheConnectionException;
Long decrease(KEYkey)throws CacheConnectionException;
Long decrease(KEYkey, Long count)throws CacheConnectionException;
Long increase(KEYkey, Long count)throws CacheConnectionException;
Long increase(KEYkey)throws CacheConnectionException;
boolean bit(KEYkey, Integer offset)throws CacheConnectionException;
}
业务方必须统一按KEY类型读写redis,类型保护,保证KEY的规范一致。
KEY的定义
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.sparrow.constant.cache;
import com.sparrow.constant.magic.SYMBOL;
import com.sparrow.core.Pair;
import com.sparrow.support.ModuleSupport;
import com.sparrow.utility.StringUtility;
import java.util.Arrays;
/**
* Created by harry on 2018/1/8.
*/
public class KEY {
private Stringbusiness;
private ObjectbusinessId;
private Stringmodule;
private KEY(){}
private KEY(Builder builder) {
this.business = builder.business.getKey();
this.module=builder.business.getModule();
if (builder.businessId !=null) {
this.businessId = StringUtility.join(Arrays.asList(builder.businessId), SYMBOL.DOT);
}
}
public static KEY parse(String key){
if(StringUtility.isNullOrEmpty(key)){
return null;
}
KEY k=new KEY();
Pair businessWithId=Pair.split(key,SYMBOL.COLON);
k.businessId=businessWithId.getSecond();
String[] businessArray=businessWithId.getFirst().split("\\.");
k.module=businessArray[0];
k.business=businessWithId.getFirst();
return k;
}
public String key() {
if (StringUtility.isNullOrEmpty(this.businessId)) {
return this.business;
}
return this.business + SYMBOL.COLON +this.businessId;
}
public String getBusiness() {
return business;
}
public String getModule() {
return module;
}
public static class Business {
private Stringmodule;
private Stringkey;
public Business(ModuleSupport module, String... business) {
this.module = module.name();
this.key =this.module;
if (business !=null && business.length >0) {
this.key += SYMBOL.DOT + StringUtility.join(Arrays.asList(business), SYMBOL.DOT);
}
}
public String getKey() {
return key;
}
public String getModule() {
return module;
}
}
public static class Builder {
private Businessbusiness;
private Object[]businessId;
public Builder business(Business business) {
this.business = business;
return this;
}
public Builder businessId(Object... businessId) {
this.businessId = businessId;
return this;
}
public KEY build() {
return new KEY(this);
}
}
}
业务方可通过实现CacheMonitor 接口,对KEY进行统一监控
/**
* Created by harry on 2018/1/25.
*/
public class SparrowCacheMonitorimplements CacheMonitor{
@Override
public void monitor(Long startTime, Long endTime, KEY key) {
可以对module 或business维护对KEY进行监控
System.out.println("module-"+key.getModule()+" business.type-"+key.getBusiness()+" key-"+key.key()+" start.time-"+startTime+" end.time-"+endTime);
}
}
DEMO实例
/**
* @author by harry
*/
public class RedisTest {
public static void main(String[] args)throws CacheConnectionException {
Container container =new SparrowContainerImpl();
//定义模块,一个业务会存在多个模块
ModuleSupport OD=new ModuleSupport() {
@Override
public String code() {
return "01";
}
@Override
public String name() {
return "OD";
}
};
//相同模块下会存在多个业务
KEY.Business od=new KEY.Business(OD,"POOL");
container.init();
CacheClient client = container.getBean("cacheClient");
//相同业务下存在多个KEY
KEY key =new KEY.Builder().business(od).businessId("BJS","CHI","HU").build();
client.set(key,"test");
KEY k2=KEY.parse("OD.POOL:BJS.CHI.HU");
System.out.println("key:"+k2.key()+",module:"+k2.getModule()+" business:"+k2.getBusiness());
}
}
运行结果:
容器初始化...
module-OD business.type-OD.POOL key-OD.POOL:BJS.CHI.HU start.time-1516877958682 end.time-1516877958714
key:OD.POOL:BJS.CHI.HU,module:OD business:OD.POOL
源码下载
https://github.com/sparrowzoo/sparrow
上一篇: SPARROW架构介绍
下一篇: tomcat 日志那点事