Redis实现布隆过滤器解析
布隆过滤器原理介绍
【1】概念说明
1)布隆过滤器(Bloom Filter)是1970年由布隆提出的。它实际上是一个很长的二进制向量和一系列随机映射函数。布隆过滤器可以用于检索一个元素是否在一个集合中。它的优点是空间效率和查询时间都远远超过一般的算法,缺点是有一定的误识别率和删除困难。
【2】设计思想
1)BF是由一个长度为m比特的位数组(bit array)与k个哈希函数(hash function)组成的数据结构。位数组均初始化为0,所有哈希函数都可以分别把输入数据尽量均匀地散列。
2)当要插入一个元素时,将其数据分别输入k个哈希函数,产生k个哈希值。以哈希值作为位数组中的下标,将所有k个对应的比特置为1。
3)当要查询(即判断是否存在)一个元素时,同样将其数据输入哈希函数,然后检查对应的k个比特。如果有任意一个比特为0,表明该元素一定不在集合中。如果所有比特均为1,表明该集合有(较大的)可能性在集合中。为什么不是一定在集合中呢?因为一个比特被置为1有可能会受到其他元素的影响,这就是所谓“假阳性”(false positive)。相对地,“假阴性”(false negative)在BF中是绝不会出现的。
【3】图示
【4】优缺点
1)优点
1.不需要存储数据本身,只用比特表示,因此空间占用相对于传统方式有巨大的优势,并且能够保密数据;
2.时间效率也较高,插入和查询的时间复杂度均为O(k);
3.哈希函数之间相互独立,可以在硬件指令层面并行计算。
2)缺点
1.存在假阳性的概率,不适用于任何要求100%准确率的情境;
2.只能插入和查询元素,不能删除元素,这与产生假阳性的原因是相同的。我们可以简单地想到通过计数(即将一个比特扩展为计数值)来记录元素数,但仍然无法保证删除的元素一定在集合中。(因此只能进行重建)
guava框架如何实现布隆过滤器
【1】引入依赖
<dependency> <groupId>com.google.guava</groupId> <artifactId>guava</artifactId> <version>28.0-jre</version> </dependency>
【2】简单使用
//布隆过滤器-数字指纹存储在当前jvm当中 public class LocalBloomFilter { private static final BloomFilter<String> bloomFilter = BloomFilter.create(Funnels.stringFunnel(StandardCharsets.UTF_8),1000000,0.01); /** * 谷歌guava布隆过滤器 * @param id * @return */ public static boolean match(String id){ return bloomFilter.mightContain(id); } public static void put(Long id){ bloomFilter.put(id+""); } }
【3】源码分析(由上面的三个主要方法看起,create方法,mightContain方法,put方法)
1)create方法深入分析
@VisibleForTesting static <T> BloomFilter<T> create(Funnel<? super T> funnel, long expectedInsertions, double fpp, Strategy strategy) { //检测序列化器 checkNotNull(funnel); //检测存储容量 checkArgument(expectedInsertions >= 0, "Expected insertions (%s) must be >= 0", expectedInsertions); //容错率应该在0-1之前 checkArgument(fpp > 0.0, "False positive probability (%s) must be > 0.0", fpp); checkArgument(fpp < 1.0, "False positive probability (%s) must be < 1.0", fpp); //检测策略 checkNotNull(strategy); if (expectedInsertions == 0) { expectedInsertions = 1; } // 这里numBits即底下LockFreeBitArray位数组的长度,可以看到计算方式就是外部传入的期待数和容错率 long numBits = optimalNumOfBits(expectedInsertions, fpp); int numHashFunctions = optimalNumOfHashFunctions(expectedInsertions, numBits); try { return new BloomFilter<T>(new BitArray(numBits), numHashFunctions, funnel, strategy); } catch (IllegalArgumentException e) { throw new IllegalArgumentException("Could not create BloomFilter of " + numBits + " bits", e); } } private BloomFilter(BitArray bits, int numHashFunctions, Funnel<? super T> funnel, Strategy strategy) { //检测hash函数个数应该在0-255之间 checkArgument(numHashFunctions > 0, "numHashFunctions (%s) must be > 0", numHashFunctions); checkArgument(numHashFunctions <= 255, "numHashFunctions (%s) must be <= 255", numHashFunctions); this.bits = checkNotNull(bits); this.numHashFunctions = numHashFunctions; this.funnel = checkNotNull(funnel); this.strategy = checkNotNull(strategy); } //计算容量大小 @VisibleForTesting static long optimalNumOfBits(long n, double p) { if (p == 0) { p = Double.MIN_VALUE; } return (long) (-n * Math.log(p) / (Math.log(2) * Math.log(2))); } //计算满足条件时,应进行多少次hash函数 @VisibleForTesting static int optimalNumOfHashFunctions(long n, long m) { // (m / n) * log(2), but avoid truncation due to division! return Math.max(1, (int) Math.round((double) m / n * Math.log(2))); }
2)mightContain方法深入分析
public boolean mightContain(T object) { return strategy.mightContain(object, funnel, numHashFunctions, bits); } public <T> boolean mightContain(T object, Funnel<? super T> funnel, int numHashFunctions, BloomFilterStrategies.BitArray bits) { long bitSize = bits.bitSize(); long hash64 = Hashing.murmur3_128().hashObject(object, funnel).asLong(); int hash1 = (int)hash64; int hash2 = (int)(hash64 >>> 32); for(int i = 1; i <= numHashFunctions; ++i) { int combinedHash = hash1 + i * hash2; if (combinedHash < 0) { combinedHash = ~combinedHash; } if (!bits.get((long)combinedHash % bitSize)) { return false; } } return true; }
3)put方法深入分析
@CanIgnoreReturnValue public boolean put(T object) { return strategy.put(object, funnel, numHashFunctions, bits); } //策略实现填入bits public <T> boolean put(T object, Funnel<? super T> funnel, int numHashFunctions, BloomFilterStrategies.BitArray bits) { long bitSize = bits.bitSize(); long hash64 = Hashing.murmur3_128().hashObject(object, funnel).asLong(); int hash1 = (int)hash64; int hash2 = (int)(hash64 >>> 32); boolean bitsChanged = false; for(int i = 1; i <= numHashFunctions; ++i) { int combinedHash = hash1 + i * hash2; if (combinedHash < 0) { combinedHash = ~combinedHash; } bitsChanged |= bits.set((long)combinedHash % bitSize); } return bitsChanged; }
采用Redis实现布隆过滤器
【1】抽出guava框架中部分核心逻辑方法形成工具类
/** * 算法过程: * 1. 首先需要k个hash函数,每个函数可以把key散列成为1个整数 * 2. 初始化时,需要一个长度为n比特的数组,每个比特位初始化为0 * 3. 某个key加入集合时,用k个hash函数计算出k个散列值,并把数组中对应的比特位置为1 * 4. 判断某个key是否在集合时,用k个hash函数计算出k个散列值,并查询数组中对应的比特位,如果所有的比特位都是1,认为在集合中。 * @description: 布隆过滤器,摘录自Google-guava包 **/ public class BloomFilterHelper<T> { private int numHashFunctions; private int bitSize; private Funnel<T> funnel; public BloomFilterHelper(Funnel<T> funnel, int expectedInsertions, double fpp) { Preconditions.checkArgument(funnel != null, "funnel不能为空"); this.funnel = funnel; // 计算bit数组长度 bitSize = optimalNumOfBits(expectedInsertions, fpp); // 计算hash方法执行次数 numHashFunctions = optimalNumOfHashFunctions(expectedInsertions, bitSize); } public int[] murmurHashOffset(T value) { int[] offset = new int[numHashFunctions]; //有点类似于hashmap中采用高32位与低32位相与获得hash值 long hash64 = Hashing.murmur3_128().hashObject(value, funnel).asLong(); int hash1 = (int) hash64; int hash2 = (int) (hash64 >>> 32); //采用对低32进行变更以达到随机哈希函数的效果 for (int i = 1; i <= numHashFunctions; i++) { int nextHash = hash1 + i * hash2; if (nextHash < 0) { nextHash = ~nextHash; } offset[i - 1] = nextHash % bitSize; } return offset; } /** * 计算bit数组长度 * Math.log(2) = 0.6931471805599453;(取0.693147来用) * (Math.log(2) * Math.log(2)) = 0.48045237; * 假设传入n为1,000,000 , p为0.01; * Math.log(0.01) = -4.605170185988091; * 则返回值为9,585,071 ,即差不多是预设容量的10倍 * * 要知道 1MB = 1024KB , 1KB = 1024B ,1B=8bit。 * 也就是对一百万数据预计花费的内存为1.143MB的内存 */ private int optimalNumOfBits(long n, double p) { if (p == 0) { // 设定最小期望长度 p = Double.MIN_VALUE; } return (int) (-n * Math.log(p) / (Math.log(2) * Math.log(2))); } /** * 计算hash方法执行次数 */ private int optimalNumOfHashFunctions(long n, long m) { return Math.max(1, (int) Math.round((double) m / n * Math.log(2))); } }
【2】构建Redis实现布隆过滤器的服务类
public class BloomRedisService { private RedisTemplate<String, Object> redisTemplate; private BloomFilterHelper bloomFilterHelper; public void setBloomFilterHelper(BloomFilterHelper bloomFilterHelper) { this.bloomFilterHelper = bloomFilterHelper; } public void setRedisTemplate(RedisTemplate<String, Object> redisTemplate) { this.redisTemplate = redisTemplate; } /** * 根据给定的布隆过滤器添加值 * 这里可以考虑LUA脚本进行优化,减少传输次数 * 如 eval "redis.call('setbit',KEYS[1],ARGV[1],1) redis.call('setbit',KEYS[1],ARGV[2],1) " 1 mybool 243 5143 * 但是又需要衡量操作的时间,与如果次数很多导致的传输的数据量很大容易阻塞问题 */ public <T> void addByBloomFilter(String key, T value) { Preconditions.checkArgument(bloomFilterHelper != null, "bloomFilterHelper不能为空"); int[] offset = bloomFilterHelper.murmurHashOffset(value); for (int i : offset) { redisTemplate.opsForValue().setBit(key, i, true); } } /** * 根据给定的布隆过滤器判断值是否存在 */ public <T> boolean includeByBloomFilter(String key, T value) { Preconditions.checkArgument(bloomFilterHelper != null, "bloomFilterHelper不能为空"); int[] offset = bloomFilterHelper.murmurHashOffset(value); for (int i : offset) { if (!redisTemplate.opsForValue().getBit(key, i)) { return false; } } return true; } }
【3】编辑配置类
@Slf4j @Configuration public class BloomFilterConfig implements InitializingBean{ @Autowired private PmsProductService productService; @Autowired private RedisTemplate template; @Bean public BloomFilterHelper<String> initBloomFilterHelper() { return new BloomFilterHelper<>((Funnel<String>) (from, into) -> into.putString(from, Charsets.UTF_8) .putString(from, Charsets.UTF_8), 1000000, 0.01); } // 布隆过滤器bean注入 @Bean public BloomRedisService bloomRedisService(){ BloomRedisService bloomRedisService = new BloomRedisService(); bloomRedisService.setBloomFilterHelper(initBloomFilterHelper()); bloomRedisService.setRedisTemplate(template); return bloomRedisService; } @Override public void afterPropertiesSet() throws Exception { List<Long> list = productService.getAllProductId(); log.info("加载产品到布隆过滤器当中,size:{}",list.size()); if(!CollectionUtils.isEmpty(list)){ list.stream().forEach(item->{ //LocalBloomFilter.put(item); bloomRedisService().addByBloomFilter(RedisKeyPrefixConst.PRODUCT_REDIS_BLOOM_FILTER,item+""); }); } } }
【4】构建布隆过滤器的拦截器
//拦截器,所有需要查看商品详情的请求必须先过布隆过滤器 @Slf4j public class BloomFilterInterceptor implements HandlerInterceptor { @Autowired private BloomRedisService bloomRedisService; @Override public boolean preHandle(HttpServletRequest request, HttpServletResponse response, Object handler) throws Exception { String currentUrl = request.getRequestURI(); PathMatcher matcher = new AntPathMatcher(); //解析出pathvariable Map<String, String> pathVariable = matcher.extractUriTemplateVariables("/pms/productInfo/{id}", currentUrl); //布隆过滤器存储在redis中 if(bloomRedisService.includeByBloomFilter(RedisKeyPrefixConst.PRODUCT_REDIS_BLOOM_FILTER,pathVariable.get("id"))){ return true; } /* * 不在布隆过滤器当中,直接返回验证失败 * 设置响应头 */ response.setHeader("Content-Type","application/json"); response.setCharacterEncoding("UTF-8"); String result = new ObjectMapper().writeValueAsString(CommonResult.validateFailed("产品不存在!")); response.getWriter().print(result); return false; } }
【5】将拦截器注册进SpringMVC中
@Configuration public class IntercepterConfiguration implements WebMvcConfigurer { @Override public void addInterceptors(InterceptorRegistry registry) { //注册拦截器 registry.addInterceptor(authInterceptorHandler()) .addPathPatterns("/pms/productInfo/**"); } @Bean public BloomFilterInterceptor authInterceptorHandler(){ return new BloomFilterInterceptor(); } }