解Bug之路-串包Bug

{"type":"doc","content":[{"type":"heading","attrs":{"align":null,"level":1},"content":[{"type":"text","text":"解Bug之路-串包Bug"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"笔者很热衷于解决Bug,同时比较擅长(网络/协议)部分,所以经常被唤去解决一些网络IO方面的Bug。现在就挑一个案例出来,写出分析思路,以飨读者,希望读者在以后的工作中能够少踩点坑。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":1},"content":[{"type":"text","text":"串包Bug现场"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"前置故障Redis超时"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"由于某个系统大量的hget、hset操作将Redis拖垮,通过监控发现Redis的CPU和IO有大量的尖刺,CPU示意图下图所示:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/4f/4fd22e8304c18ea1c8df87cfe8ec260d.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"CPU达到了100%,导致很多Redis请求处理不及时,其它业务系统都频繁爆出readTimeOut。此时,紧急将这个做大量hget、hset的系统kill,过了一段时间,Redis的CPU恢复平稳。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"一波未平,一波又起"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"就在我们以为事件平息的时候,线上爆出登录后的用户名称不正确。同时错误日志里面也有大量的Redis返回不正确的报错。尤为奇葩的是,系统获取一个已经存在的key,例如get User123456Name,返回的竟然是redis的成功返回OK。示意图如下:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"codeblock","attrs":{"lang":null},"content":[{"type":"text","text":"Jedis.sendCommand:get User123456Name\nJedis.return:OK\n or\nJedis.sendCommand:get User123456Name\nJedis.return:user789\n"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"我们发现此情况时,联系op将Redis集群的所有Key紧急delete,当时监控示意图:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/d2/d28fc449fccb43ebc3ba5ceb7a589f7b.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"当重启后,我们再去线上观察的时候,发现错误依然存在,神奇的是,这种错误发生的频率会随着时间的增加而递减。到最后刷个10分钟页面才会出现这种错,示意图如下所示:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/f7/f7cc61cf84e43a4447966c9a77470d16.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"既然如此,那只能祭出重启大法,把出错的业务系统全部重启了一遍。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"重启之后,线上恢复正常,一切Okay。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"Bug覆盘"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"此次Bug是由Redis本身Server负载太高超时引起的。Bug的现象是通过Jedis去取对应的Key值,得不到预期的结果,简而言之包乱了,串包了。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"缩小Bug范围"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"首先:Redis是全球久经考验的系统,这样的串包不应该是Redis的问题。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"第二:Redis刷新了key后Bug依然存在,而业务系统重启了之后Okay。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"第三:笔者在错误日志中发现一个现象,A系统只可能打印出属于A系统的json串结构(redis存的是json)。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"很明显,是业务系统的问题,如果是Redis本身的问题,那么在很大概率上A系统会接收到B系统的json串结构。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"业务系统问题定位"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"业务系统用的是Jedis,这同样也是一个久经考验的库,出现此问题的可能性不大。那么问题肯定是出在运用Jedis的姿势上。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"于是笔者找到了下面一段代码:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"codeblock","attrs":{"lang":null},"content":[{"type":"text","text":"public Object invoke(Object proxy,Method method,Object[] args) throws Throwable{\n JedisClient jedisClient = jedisPool.getResource(); \n try{\n return method.invoke(jedisClient,args); \n } catch(Exception e){\n logger.error(\"invoke redis error\",e); \n throw e; \n }finally {\n if(jedisClient != null){\n // 问题处在下面这句\n jedisPool.returnResource(jedisClient);\n }\n }\n}\n"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"当时我就觉得很奇怪,笔者自己写的,阅读过的连接池的代码都没有将抛异常的连接放回池里。就以Druid为例,如果是网络IO等fatal级别的异常,直接抛弃连接。这里把jedisClient连接返回去感觉就是出问题的关键。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"Bug推理"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"笔者意识到,之所以串包可能是由于jedisClient里面可能有残余的数据,导致读取的时候读取到此数据,从而造成串包的现象。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"串包原因"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"正常情况下的redis交互"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"先上Jedis源码"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"codeblock","attrs":{"lang":null},"content":[{"type":"text","text":"public String get(final String key) {\n checkIsInMulti();\n client.sendCommand(Protocol.Command.GET, key);\n return client.getBulkReply();\n}\n"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Jedis本身用的是Bio,上述源码的过程示意图如下:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/a5/a51c500eb4f4a4b0b7da5c41ee7b9de5.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"出错的业务系统的redis交互"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/46/46874b44664ee621316ece381217a950.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"由于Redis本身在高负载状态,导致没能及时相应command请求,从而导致readTimeOut异常。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"复用这个出错链接导致出错"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在Redis响应了上一个command后,把数据传到了对应command的socket,进而被inputream给buffer起来。而这个command由于超时失败了。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/32/32438b067c5d1bf511404c902963cd9c.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"这样,inputStream里面就有个上个命令留下来的数据。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"下一次业务操作在此拿到这个连接的时候,就会出现下面的情况。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/3e/3e5f174f75e8ec6c3902a3f1b0813075.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"再下面的命令get user789Key会拿到get user456Key的结果,依次类推,则出现串包的现象。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"串包过程图"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/0f/0f8e2caa8cea8cabf66a0fabd4561fd7.png","alt":"输入图片说明","title":"在这里输入图片标题","style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"上图中相同颜色的矩形对应的数据是一致的。但是由于Redis超时,导致数据串了。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"为什么get操作返回OK"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"上图很明显的解释了为什么一个get操作会返回OK的现象。因为其上一个操作是set操作,它返回的OK被get操作读取到,于是就有了这种现象。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"为什么会随着时间的收敛而频率降低"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"因为在调用Redis出错后,业务系统有一层拦截器会拦截到业务层的出错,同时给这个JedisClient的错误个数+1,当错误个数>3的时候,会将其从池中踢掉。这样这种串了的连接会越来越少,导致Bug原来越难以出现。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"在每次调用之前清理下inputstream可行否"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"不行,因为Redis可能在你清理inputstream后,你下次读取前把数据给传回来。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"怎么避免这种现象?"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"抛出这种IO异常的连接直接给扔掉,不要放到池子里面。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":4},"content":[{"type":"text","text":"怎么从协议层面避免这种现象"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"对每次发送的命令都加一个随机的packetId,然后结果返回回来的时候将这个packetId带回来。在客户端每次接收到数据的时候,获取包中的packetId和之前发出的packetId相比较,如下代码所示:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"codeblock","attrs":{"lang":null},"content":[{"type":"text","text":"if(oldPacketId != packetIdFromData){\n throw new RuntimeException(\"串包\");\n}\n"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"总结"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"至少在笔者遇到的场景中,出现IO异常的连接都必须被抛掉废弃,因为你永远不知道在你复用的那一刻,socket或者inputstream的buffer中到底有没有上一次命令遗留的数据。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"当然如果刻意的去构造协议,能够通过packetId之类的手段把收发状态重新调整为一致也是可以的,这无疑增加了很高的复杂度。所以废弃连接重建是最简单有效的方法。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"公众号"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"关注笔者公众号,获取更多干货文章:"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/04/046ecbf6ada994d1dade8690048697a9.jpeg","alt":"","title":null,"style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}}]}
發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章