mySeckill.lua
Lua优点：
减少网络开销：这个脚本只要执行一次，能减少网络传输
原子性：Redis将这个脚本作为原子执行要么全部成功或者失败，不担心并发问题，不需要事务，（PS:LUA脚本保证原子性，执行lua脚本时不会同时执行其它脚本或redis命令， 这种语义类似于MULTI/EXEC，这个lua脚本要么执行成功，要么执行失败）
复用性：lua一旦执行就能永久性保存Redis的数据，可以供其它客户端使用?
-- 全局函数: 求阶乘 function factorial(n) if n == 1 then return 1 else return n * fact(n - 1) end end
--商品库存Key
local product_stock_key = KEYS[1]
--商品购买用户记录Key
local buyersKey = KEYS[2]
--用户ID
local uid = KEYS[3]
--校验用户是否重复秒杀
local result = redis.call("sadd" , buyersKey , uid )
if(tonumber(result) == 1)
then
--初次秒杀
local stock = redis.call("lpop" , product_stock_key )
if(stock)
then
--库存>0
return 1
else
--库存不足
return -1
end
else
--重复秒杀
return 2
end
SeckillController?
?PS：@MyAcessLimter(count = 1000,timeout = 1)
redis + lua限流 + AOP实现接口对客户端限流_Zxdwr520的博客-CSDN博客
开始秒杀时整个过程在缓存中进行，秒杀结束后再从缓存中拿数据同步到数据库?
SeckillController
/**
* 秒杀商品设置到Redis中，返回给前端页面展示
* @author fan
* @date 2022/5/10 23:56
* @return java.util.List<com.fan.li.entity.Product>
*/
@RequestMapping(value = "queryAll")
@ResponseBody
@MyAcessLimter(count = 1000,timeout = 1)
public List<Product> queryAll(@RequestParam("seckillDate") String seckillDate) {
DateTimeFormatter dtf = DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:ss");
Gson gson = new Gson();
List<Product> productList = productService.queryAll();//
if (productList == null){
return null;
}
for (Product product: productList) {
long productId = product.getId();
redisTemplate.opsForValue().set("product_" + productId, gson.toJson(product));
// 一个用户只买一件商品
// 商品购买用户Set
redisTemplate.opsForSet().add("product_buyers_" + product.getId(), "");
for (int i = 0; i < product.getStock(); i++) {
redisTemplate.opsForList().leftPush("product_stock_key_" + product.getId(), String.valueOf(i));
}
System.out.println(gson.toJson(product));
}
redisTemplate.opsForValue().set("seckill_plan_" + seckillDate, gson.toJson(productList));//把商品信息存入缓存，列表展示用
return productList;
}
/**
* 开始秒杀商品，并限流--- 这里秒杀时间的操作略
* @author fan
* @date 2022/5/10 1:03
* @param userId
* @param productId
* @return java.lang.String
*/
@RequestMapping(value = "seckillProduct")
@ResponseBody
@MyAcessLimter(count = 1000,timeout = 1)
public String seckillProduct(@RequestParam("uid") long userId, @RequestParam("pid") long productId){
List<String> list = Lists.newArrayList("product_stock_key_" + productId, "product_buyers_" + productId, "" + userId );
Long code = redisTemplate.execute(defaultRedisScript, list, "");
if (code == -1) {
return "库存不足";
} else if (code == 2) {
return "不允许重复秒杀";
} else if (code == 1) {//整个秒杀过程在缓存中进行，秒杀结束后从缓存中拿数据库加入队列同步到数据库中
String productJson = redisTemplate.opsForValue().get("product_" + productId);
Gson gson = new Gson();
Product product = gson.fromJson(productJson, Product.class);
Order order = new Order();
order.setProductId(productId);
order.setUserId(userId);
String id = String.valueOf(UUID.randomUUID());
order.setId(id);
order.setOrderName("抢购" + product.getName());
order.setProductName(product.getName());
redisClient.setString("order_" + id, gson.toJson(order));
return "sueccss";
}
return "error";
}
/**
* 我们可以在夜深人静的时候做点事情，比如同步订单到数据库
* @author fan
* @date 2022/5/10 2:33
* @return java.lang.String
*/
@GetMapping(value = "createOrderProducer")
@ResponseBody
@Scheduled(cron = "0 0 0 * * ?")
public String createOrderProducer(){
Set<String> orderKeys = redisTemplate.keys("order_*");
for (String orderKey: orderKeys) {//从缓存拿订单数据加入队列
if (orderKey != null) {
String orderJson = redisClient.getString(orderKey);
Gson gson = new Gson();
Order order = gson.fromJson(orderJson, Order.class);//从缓存取出订单
orderToRedisProducer.createOrderToRedisProducer("fanoutQueueOrder", order);
}
}
return "sueccss";
}
生产者
@Component
@Slf4j
public class OrderToRedisProducer {
@Autowired
private RabbitTemplate rabbitTemplate;
@Autowired
private RedisTemplate redisTemplate;
@Autowired
private RedisClient redisClient;
@Autowired
private DefaultRedisScript<Long> defaultRedisScript;
public String createOrderToRedisProducer(String queueName, Order order){
JSONObject jsonObject = new JSONObject();
String messageId = String.valueOf(UUID.randomUUID());
jsonObject.put("timestamp", System.currentTimeMillis());
if (order != null) {
String productJson = (String) redisTemplate.opsForValue().get("product_" + order.getProductId());
jsonObject.put("productJson", productJson);
jsonObject.put("id", order.getId());
jsonObject.put("productId", order.getProductId() + "");
jsonObject.put("userId", order.getUserId() + "");
String jsonString = jsonObject.toJSONString();
Message message = MessageBuilder.withBody(jsonString.getBytes())
.setDeliveryMode(MessageDeliveryMode.PERSISTENT)
.setContentType(MessageProperties.CONTENT_TYPE_JSON)
.setDeliveryTag(System.currentTimeMillis())
.setContentEncoding("utf-8")
.setMessageId(messageId)
.build();
rabbitTemplate.convertAndSend(queueName, message);
return "success";
}else {
return "error";
}
}
}消费者
@Slf4j
@Component
public class OrderToRedisConsumer {
@Autowired
private RedisClient redisClient;
@Autowired
private OrderService orderService;
@Autowired
private ProductService productService;
@RabbitListener(queues = "fanoutQueueOrder")
public void createOrderConsumer(Message message, Channel channel) throws Exception {
log.info("OrderConsumer 消费者收到消息:{}" , JSONObject.toJSONString(message));
String messageId = message.getMessageProperties().getMessageId();
String msg = new String(message.getBody(), "UTF-8");
String messageIdRedis = //(String) redisTemplate.opsForValue().get("messageId");
redisClient.getString("messageId");
if (messageId != null) {//避免消费者消息重复
if (!messageId.equals(messageIdRedis)) {
redisClient.setString(messageId,messageId,3_000L * 600);// 写入缓存
System.out.println("number==" + channel.getChannelNumber());
JSONObject jsonObject = JSONObject.parseObject(msg);
long deliverTag = message.getMessageProperties().getDeliveryTag();
try {
//具体业务
String id = (String) jsonObject.get("id");
String productId = (String) jsonObject.get("productId");
String userId = (String) jsonObject.get("userId");
String productJson = (String) jsonObject.get("productJson");
if (productJson != null) {
Gson gson = new Gson();
Product product = gson.fromJson(productJson, Product.class);
Order order = new Order();
order.setProductId(Long.parseLong(productId));
order.setUserId(Long.parseLong(userId));
order.setId(id);
order.setOrderName("抢购" + product.getName());
order.setProductName(product.getName());
int p = productService.updateProduct(product.getId());//开启事务
if (p > 0) {
int i = orderService.insert(order);//开启事务
//loggerService.saveLog(jsonObject);//日志记录
if (i > 0 && p > 0) {
log.info("创建订单成功：i=" + i);
log.info("商品库存减-1成功：p=" + p);
LocalDateTime localDateTime = LocalDateTime.now();
Date dataTime = Date.from(localDateTime.atZone(ZoneId.systemDefault()).toInstant());
log.info("dataTime=" + dataTime);
channel.basicAck(deliverTag, true);//手动设置ack，消费成功，确认消息
}else {
channel.basicNack(deliverTag, false, true);
}
}
}
}catch (Exception e){
try {
channel.basicNack(deliverTag, false, true);
} catch (Exception ioException) {
log.error("重新放入队列失败，失败原因:{}",e.getMessage(),e);
}
log.error("TopicConsumer消费者出错,mq参数:{}，错误信息：{}",message,e.getMessage(),e);
}
System.out.println("消费消息jsonObject：" + jsonObject + ",messageId：" + messageId);
}
}
}
}?效果图：
