转载自：https://www.cnblogs.com/lanbo203/p/7494587.html
解决方案
一、对强一致要求比较高的，应采用实时同步方案，即查询缓存查询不到再从DB查询，保存到缓存；更新缓存时，先更新数据库，再将缓存的设置过期(建议不要去更新缓存内容，直接设置缓存过期)。
二、对于并发程度较高的，可采用异步队列的方式同步，可采用kafka等消息中间件处理消息生产和消费。
三、使用阿里的同步工具canal，canal实现方式是模拟mysql slave和master的同步机制，监控DB bitlog的日志更新来触发缓存的更新，此种方法可以解放程序员双手，减少工作量，但在使用时有些局限性。
四、采用UDF自定义函数的方式，面对mysql的API进行编程，利用触发器进行缓存同步，但UDF主要是c/c++语言实现，学习成本高。
实时同步
spring3+提供了注解的方式进行缓存编程
@Cacheable(key = "caches[0].name + T(String).valueOf(#userId)",unless = "#result eq null")
@CachePut(key = "caches[0].name + T(String).valueOf(#user.userId)")
@CacheEvict(key = "caches[0].name + T(String).valueOf(#userId)" )
@Caching(evict = {@CacheEvict(key = "caches[0].name + T(String).valueOf(#userId)" ), @CacheEvict(key = "caches[0].name + #result.name" )}) @Cacheable：查询时使用，注意Long类型需转换为Sting类型，否则会抛异常
@CachePut：更新时使用，使用此注解，一定会从DB上查询数据
@CacheEvict：删除时使用；
@Caching：组合用法 具体注解的使用可参考官网 注意：注解方式虽然能使我们的代码简洁，但是注解方式有局限性：对key的获取，以及嵌套使用时注解无效，如下所示
public class User { private Long userId; private String name; private Integer age; private String sex; private String addr; //get set .....} service接口
| 1 2 3 4 5 6 7 | public interface UserService { User getUser(Long userId); User updateUser(User user); User getUserByName(String name); int insertUser(User user); User delete (Long userId); }<br> //实现类<br>//假设有需求是由name查询user的，一般我们是先由name->id,再由id->user，这样会减少redis缓存的冗余信息 |
@Service(value = "userSerivceImpl") @CacheConfig(cacheNames = "user") public class UserServiceImpl implements UserService { private static Logger log = LoggerFactory.getLogger(UserServiceImpl.class); @Autowired UserMapper userMapper; @Cacheable(key = "caches[0].name + T(String).valueOf(#userId)",unless = "#result eq null") public User getUser(Long userId) { User user = userMapper.selectByPrimaryKey(userId); return user; } @Cacheable(key = "caches[0].name + #name") public String getIdByName(String name){ Long userId = userMapper.getIdByName(name); return String.valueOf(userId); } //使用getUserByName方式调用getIdByName 和getUser方法来实现查询，但是如果用此方式在controller中直接调用 //getUserByName方法，缓存效果是不起作用的，必须是直接调用getIdByName和getUser方法才能起作用
public User getUserByName(String name) { //通过name 查询到主键 再由主键查询实体 return getUser(Long.valueOf(getIdByName(name))); } 非注解方式实现
1.先定义一个RedisCacheConfig类用于生成RedisTemplate和对CacheManager的管理
@Configurationpublic class RedisCacheConfig extends CachingConfigurerSupport { /*定义缓存数据 key 生成策略的bean *包名+类名+方法名+所有参数 */ @Bean public KeyGenerator keyGenerator() { return new KeyGenerator() { @Override public Object generate(Object target, Method method, Object... params) { StringBuilder sb = new StringBuilder(); sb.append(target.getClass().getName()); sb.append(method.getName()); for (Object obj : params) { sb.append(obj.toString()); } return sb.toString(); } }; } //@Bean public CacheManager cacheManager( @SuppressWarnings("rawtypes") RedisTemplate redisTemplate) { //RedisCacheManager cacheManager = new RedisCacheManager(redisTemplate); //cacheManager.setDefaultExpiration(60);//设置缓存保留时间（seconds） return cacheManager; } //1.项目启动时此方法先被注册成bean被spring管理 @Bean public StringRedisTemplate stringRedisTemplate(RedisConnectionFactory factory) { StringRedisTemplate template = new StringRedisTemplate(factory); Jackson2JsonRedisSerializer jackson2JsonRedisSerializer = new Jackson2JsonRedisSerializer(Object.class); ObjectMapper om = new ObjectMapper(); om.setVisibility(PropertyAccessor.ALL, JsonAutoDetect.Visibility.ANY); om.enableDefaultTyping(ObjectMapper.DefaultTyping.NON_FINAL); jackson2JsonRedisSerializer.setObjectMapper(om); template.setValueSerializer(jackson2JsonRedisSerializer); template.afterPropertiesSet(); return template; } @Bean public RedisTemplate redisTemplate(RedisConnectionFactory connectionFactory) { RedisTemplate template = new RedisTemplate<>(); template.setConnectionFactory(connectionFactory); //使用Jackson2JsonRedisSerializer来序列化和反序列化redis的value值 Jackson2JsonRedisSerializer serializer = new Jackson2JsonRedisSerializer(Object.class); System.out.println("==============obj:"+Object.class.getName()); ObjectMapper mapper = new ObjectMapper(); mapper.setVisibility(PropertyAccessor.ALL, JsonAutoDetect.Visibility.ANY); mapper.enableDefaultTyping(ObjectMapper.DefaultTyping.NON_FINAL); serializer.setObjectMapper(mapper); template.setValueSerializer(serializer); //使用StringRedisSerializer来序列化和反序列化redis的key值 template.setKeySerializer(new StringRedisSerializer()); template.afterPropertiesSet(); return template; }} 2.定义一个redisUtil类用于存取缓存值
@Componentpublic class RedisCacheUtil { @Autowired private StringRedisTemplate stringRedisTemplate; @Autowired private RedisTemplate redisTemplate; /** * 存储字符串 * @param key string类型的key * @param value String类型的value */ public void set(String key, String value) { stringRedisTemplate.opsForValue().set(key, value); } /** * 存储对象 * @param key String类型的key * @param value Object类型的value */ public void set(String key, Object value) { redisTemplate.opsForValue().set(key, value); } /** * 存储对象 * @param key String类型的key * @param value Object类型的value */ public void set(String key, Object value,Long timeOut) { redisTemplate.opsForValue().set(key, value,timeOut, TimeUnit.SECONDS); } /** * 根据key获取字符串数据 * @param key * @return */ public String getValue(String key) { return stringRedisTemplate.opsForValue().get(key); }// public Object getValue(String key) {// return redisTemplate.opsForValue().get(key);// } /** * 根据key获取对象 * @param key * @return */ public Object getValueOfObject(String key) { return redisTemplate.opsForValue().get(key); } /** * 根据key删除缓存信息 * @param key */ public void delete(String key) { redisTemplate.delete(key); } /** * 查询key是否存在 * @param key * @return */ @SuppressWarnings("unchecked") public boolean exists(String key) { return redisTemplate.hasKey(key); }} 3.实现类
/** * Created by yexin on 2017/9/8. * * 在Impl基础上+ 防止缓存雪崩和缓存穿透功能 */@Service(value = "userServiceImpl4")public class UserServiceImpl4 implements UserService { @Autowired UserMapper userMapper; @Autowired RedisCacheUtil redisCacheUtil; @Value("${timeOut}") private long timeOut; @Override public User getUser(Long userId) { String key = "user" + userId; User user = (User) redisCacheUtil.getValueOfObject(key); String keySign = key + "_sign"; String valueSign = redisCacheUtil.getValue(keySign); if(user == null){//防止第一次查询时返回时空结果 //防止缓存穿透 if(redisCacheUtil.exists(key)){ return null; } user = userMapper.selectByPrimaryKey(userId); redisCacheUtil.set(key,user); redisCacheUtil.set(keySign,"1",timeOut *(new Random().nextInt(10) + 1));// redisCacheUtil.set(keySign,"1",0L); //过期时间不能设置为0，必须比0大的数 return user; } if(valueSign != null){ return user; }else { //设置标记的实效时间 Long tt = timeOut * (new Random().nextInt(10) + 1); System.out.println("tt:"+tt); redisCacheUtil.set(keySign,"1",tt); //异步处理缓存更新 应对与高并发的情况，会产生脏读的情况 ThreadPoolUtil.getExecutorService().execute(new Runnable(){ public void run() { // System.out.println("-----执行异步操作-----"); User user1 = userMapper.selectByPrimaryKey(userId); redisCacheUtil.set(key,user1); } });// new Thread(){// public void run() { //应对与高并发的情况，会产生脏读的情况// System.out.println("-----执行异步操作-----");// User user1 = userMapper.selectByPrimaryKey(userId);// redisCacheUtil.set(key,user1);// }// }.start(); } return user; }} 异步实现
异步实现通过kafka作为消息队列实现，异步只针对更新操作，查询无需异步，实现类如下
1.pom文件需依赖
org.springframework.cloud spring-cloud-starter-stream-kafka
2.生产着代码
@EnableBinding(Source.class)public class SendService { @Autowired private Source source; public void sendMessage(String msg) { try{ source.output().send(MessageBuilder.withPayload(msg).build()); } catch (Exception e) { e.printStackTrace(); } }//接受的是一个实体类，具体配置在application.yml public void sendMessage(TransMsg msg) { try { //MessageBuilder.withPayload(msg).setHeader(KafkaHeaders.TOPIC,"111111").build(); source.output().send(MessageBuilder.withPayload(msg).build()); } catch (Exception e) { e.printStackTrace(); } }} 3.消费者代码
@EnableBinding(Sink.class)public class MsgSink { @Resource(name = "userSerivceImpl3") UserService userService; @StreamListener(Sink.INPUT) public void process(TransMsg msg) throws NoSuchMethodException, InvocationTargetException, IllegalAccessException, ClassNotFoundException { System.out.println("sink......"+msg); System.out.println("opt db strat ----"); userService.updateUser((User) msg.getParams()); System.out.println("执行db结束------"); }} 4.application.yml配置
spring: application: name: demo-provider redis: database: 0 host: 192.168.252.128 #host: localhost port: 6379 password: pool: max-active: 50 max-wait: -1 max-idle: 50 timeout: 0#kafka cloud: stream: kafka: binder: brokers: 192.168.252.128:9092 zk-nodes: 192.168.252.128:2181 minPartitionCount: 1 autoCreateTopics: true autoAddPartitions: true bindings: input: destination: topic-02# content-type: application/json content-type: application/x-java-object #此种类型配置在消费端接受到的为一个实体类 group: t1 consumer: concurrency: 1 partitioned: false output: destination: topic-02 content-type: application/x-java-object producer: partitionCount: 1 instance-count: 1 instance-index: 0
5.实现类
@Service(value = "userServiceImpl2")public class UserServiceImpl2 implements UserService{ @Autowired UserMapper userMapper; @Autowired RedisCacheUtil redisCacheUtil; private static Logger log = LoggerFactory.getLogger(UserServiceImpl.class); @Autowired SendService sendService; public User updateUser(User user) { System.out.println(" impl2 active "); String key = "user"+ user.getUserId(); System.out.println("key:"+key); //是否存在key if(!redisCacheUtil.exists(key)){ return userMapper.updateByPrimaryKeySelective(user) == 1 ? user : null; } /* 更新key对应的value 更新队列 */ User user1 = (User)redisCacheUtil.getValueOfObject(key); try { redisCacheUtil.set(key,user); TransMsg msg = new TransMsg (key,user,this.getClass().getName(),"updateUser",user); sendService.sendMessage(msg); }catch (Exception e){ redisCacheUtil.set(key,user1); } return user; }} 注意:kafka与zookeeper的配置在此不介绍
canal实现方式
先要安装canal，配置canal的example文件等，配置暂不介绍
package org.example.canal;import com.alibaba.fastjson.JSONObject;import com.alibaba.otter.canal.client.CanalConnector;import com.alibaba.otter.canal.client.CanalConnectors;import com.alibaba.otter.canal.common.utils.AddressUtils;import com.alibaba.otter.canal.protocol.Message;import com.alibaba.otter.canal.protocol.CanalEntry.Column;import com.alibaba.otter.canal.protocol.CanalEntry.Entry;import com.alibaba.otter.canal.protocol.CanalEntry.EntryType;import com.alibaba.otter.canal.protocol.CanalEntry.EventType;import com.alibaba.otter.canal.protocol.CanalEntry.RowChange;import com.alibaba.otter.canal.protocol.CanalEntry.RowData;import org.example.canal.util.RedisUtil;import java.net.InetSocketAddress;import java.util.List;public class CanalClient { public static void main(String[] args) { // 创建链接 CanalConnector connector = CanalConnectors.newSingleConnector(new InetSocketAddress(AddressUtils.getHostIp(), 11111), "example", "", ""); int batchSize = 1000; try { connector.connect(); connector.subscribe(".*\\..*"); connector.rollback(); while (true) { Message message = connector.getWithoutAck(batchSize); // 获取指定数量的数据 long batchId = message.getId(); int size = message.getEntries().size(); if (batchId == -1 || size == 0) { try { Thread.sleep(1000); } catch (InterruptedException e) { e.printStackTrace(); } } else { printEntry(message.getEntries()); } connector.ack(batchId); // 提交确认 // connector.rollback(batchId); // 处理失败, 回滚数据 } } finally { connector.disconnect(); } } private static void printEntry( List entrys) { for (Entry entry : entrys) { if (entry.getEntryType() == EntryType.TRANSACTIONBEGIN || entry.getEntryType() == EntryType.TRANSACTIONEND) { continue; } RowChange rowChage = null; try { System.out.println("tablename:"+entry.getHeaderOrBuilder().getTableName()); rowChage = RowChange.parseFrom(entry.getStoreValue()); } catch (Exception e) { throw new RuntimeException("ERROR ## parser of eromanga-event has an error , data:" + entry.toString(), e); } EventType eventType = rowChage.getEventType(); System.out.println(String.format("================> binlog[%s:%s] , name[%s,%s] , eventType : %s", entry.getHeader().getLogfileName(), entry.getHeader().getLogfileOffset(), entry.getHeader().getSchemaName(), entry.getHeader().getTableName(), eventType)); for (RowData rowData : rowChage.getRowDatasList()) { if (eventType == EventType.DELETE) { redisDelete(rowData.getBeforeColumnsList()); } else if (eventType == EventType.INSERT) { redisInsert(rowData.getAfterColumnsList()); } else { System.out.println("-------> before"); printColumn(rowData.getBeforeColumnsList()); System.out.println("-------> after"); redisUpdate(rowData.getAfterColumnsList()); } } } } private static void printColumn( List columns) { for (Column column : columns) { System.out.println(column.getName() + " : " + column.getValue() + " update=" + column.getUpdated()); } } private static void redisInsert( List columns){ JSONObject json=new JSONObject(); for (Column column : columns) { json.put(column.getName(), column.getValue()); } if(columns.size()>0){ RedisUtil.stringSet("user:"+ columns.get(0).getValue(),json.toJSONString()); } } private static void redisUpdate( List columns){ JSONObject json=new JSONObject(); for (Column column : columns) { json.put(column.getName(), column.getValue()); } if(columns.size()>0){ RedisUtil.stringSet("user:"+ columns.get(0).getValue(),json.toJSONString()); } } private static void redisDelete( List columns){ JSONObject json=new JSONObject(); for (Column column : columns) { json.put(column.getName(), column.getValue()); } if(columns.size()>0){ RedisUtil.delKey("user:"+ columns.get(0).getValue()); } }} package org.example.canal.util;import redis.clients.jedis.Jedis;import redis.clients.jedis.JedisPool;import redis.clients.jedis.JedisPoolConfig;public class RedisUtil { // Redis服务器IP private static String ADDR = "192.168.252.128"; // Redis的端口号 private static int PORT = 6379; // 访问密码 //private static String AUTH = "admin"; // 可用连接实例的最大数目，默认值为8； // 如果赋值为-1，则表示不限制；如果pool已经分配了maxActive个jedis实例，则此时pool的状态为exhausted(耗尽)。 private static int MAX_ACTIVE = 1024; // 控制一个pool最多有多少个状态为idle(空闲的)的jedis实例，默认值也是8。 private static int MAX_IDLE = 200; // 等待可用连接的最大时间，单位毫秒，默认值为-1，表示永不超时。如果超过等待时间，则直接抛出JedisConnectionException； private static int MAX_WAIT = 10000; // 过期时间 protected static int expireTime = 60 * 60 *24; // 连接池 protected static JedisPool pool; static { JedisPoolConfig config = new JedisPoolConfig(); //最大连接数 config.setMaxTotal(MAX_ACTIVE); //最多空闲实例 config.setMaxIdle(MAX_IDLE); //超时时间 config.setMaxWaitMillis(MAX_WAIT); // config.setTestOnBorrow(false); pool = new JedisPool(config, ADDR, PORT, 1000); } /** * 获取jedis实例 */ protected static synchronized Jedis getJedis() { Jedis jedis = null; try { jedis = pool.getResource(); } catch (Exception e) { e.printStackTrace(); if (jedis != null) { pool.returnBrokenResource(jedis); } } return jedis; } /** * 释放jedis资源 * @param jedis * @param isBroken */ protected static void closeResource(Jedis jedis, boolean isBroken) { try { if (isBroken) { pool.returnBrokenResource(jedis); } else { pool.returnResource(jedis); } } catch (Exception e) { } } /** * 是否存在key * @param key */ public static boolean existKey(String key) { Jedis jedis = null; boolean isBroken = false; try { jedis = getJedis(); jedis.select(0); return jedis.exists(key); } catch (Exception e) { isBroken = true; } finally { closeResource(jedis, isBroken); } return false; } /** * 删除key * @param key */ public static void delKey(String key) { Jedis jedis = null; boolean isBroken = false; try { jedis = getJedis(); jedis.select(0); jedis.del(key); } catch (Exception e) { isBroken = true; } finally { closeResource(jedis, isBroken); } } /** * 取得key的值 * @param key */ public static String stringGet(String key) { Jedis jedis = null; boolean isBroken = false; String lastVal = null; try { jedis = getJedis(); jedis.select(0); lastVal = jedis.get(key); jedis.expire(key, expireTime); } catch (Exception e) { isBroken = true; } finally { closeResource(jedis, isBroken); } return lastVal; } /** * 添加string数据 * @param key * @param value */ public static String stringSet(String key, String value) { Jedis jedis = null; boolean isBroken = false; String lastVal = null; try { jedis = getJedis(); jedis.select(0); lastVal = jedis.set(key, value); jedis.expire(key, expireTime); } catch (Exception e) { e.printStackTrace(); isBroken = true; } finally { closeResource(jedis, isBroken); } return lastVal; } /** * 添加hash数据 * @param key * @param field * @param value */ public static void hashSet(String key, String field, String value) { boolean isBroken = false; Jedis jedis = null; try { jedis = getJedis(); if (jedis != null) { jedis.select(0); jedis.hset(key, field, value); jedis.expire(key, expireTime); } } catch (Exception e) { isBroken = true; } finally { closeResource(jedis, isBroken); } }} 附redis关于缓存雪崩和缓存穿透，热点key
穿透
穿透：频繁查询一个不存在的数据，由于缓存不命中，每次都要查询持久层。从而失去缓存的意义。
解决办法： 持久层查询不到就缓存空结果，查询时先判断缓存中是否exists(key) ,如果有直接返回空，没有则查询后返回，
注意insert时需清除查询的key，否则即便DB中有值也查询不到(当然也可以设置空缓存的过期时间）
雪崩
雪崩：缓存大量失效的时候，引发大量查询数据库。
解决办法：①用锁/分布式锁或者队列串行访问②缓存失效时间均匀分布
热点key
热点key:某个key访问非常频繁，当key失效的时候有打量线程来构建缓存，导致负载增加，系统崩溃。
解决办法：
①使用锁，单机用synchronized,lock等，分布式用分布式锁。
②缓存过期时间不设置，而是设置在key对应的value里。如果检测到存的时间超过过期时间则异步更新缓存。
③在value设置一个比过期时间t0小的过期时间值t1，当t1过期的时候，延长t1并做更新缓存操作。
4设置标签缓存，标签缓存设置过期时间，标签缓存过期后，需异步地更新实际缓存 具体参照userServiceImpl4的处理方式
总结
一、查询redis缓存时，一般查询如果以非id方式查询，建议先由条件查询到id,再由id查询pojo
二、异步kafka在消费端接受信息后，该怎么识别处理那张表，调用哪个方法，此问题暂时还没解决
三、比较简单的redis缓存，推荐使用canal
参考文档
http://blog.csdn.net/fly_time2012/article/details/50751316
http://blog.csdn.net/kkgbn/article/details/60576477
http://www.cnblogs.com/fidelQuan/p/4543387.html