背景
多线程开启事务处理。每个事务有多个update操作和一个insert操作(都在同一张表)。
DDL(删除了一些不必要的细节)
默认隔离级别:Repeatable Read
CREATE TABLE `list_rate` (
`id` bigint(20) NOT NULL AUTO_INCREMENT,
`hotel_id` bigint(20) NOT NULL,
`rate_date` date NOT NULL,
`room_type_id` bigint(20) NOT NULL,
`rate` decimal(20,2) NOT NULL DEFAULT '0.00',
`create_time` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP,
`update_time` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
`operator_id` int(10) NOT NULL DEFAULT '0',
`is_deleted` tinyint(4) NOT NULL DEFAULT '0',
PRIMARY KEY (`id`) USING BTREE,
KEY `hotel_date_idx` (`hotel_id`,`rate_date`) USING BTREE
) ENGINE=InnoDB;
数据库现有数据
只有hotel_id=2和hotel_id=11111的数据
事务执行的sql(每个事务的区别是hotel_id不同,从10001到10010)
逻辑删除原有数据
UPDATE list_rate
SET is_deleted = 1, update_time = now(), operator_id = 1
WHERE hotel_id = 10007 AND room_type_id = 1 AND is_deleted = 0 AND rate_date IN ( '2018-11-10' , '2018-11-11' )
插入新的数据
INSERT INTO list_rate (hotel_id,rate_date,room_type_id,rate,create_time,update_time,operator_id,is_deleted)
VALUES (10007,'2018-11-10',1,144, now(),now(),1,0) , (10007,'2018-11-11',1,148, now(),now(),1,0)
根据现有数据情况,update的时候没有数据被更新
实际运行结果
报了非常多一样的错
Exception in thread "UpdateListRate-33" org.springframework.dao.DeadlockLoserDataAccessException:
### Error updating database. Cause: com.mysql.jdbc.exceptions.jdbc4.MySQLTransactionRollbackException: Deadlock found when trying to get lock; try restarting transaction
### The error may involve com.oyo.price.mapper.ListRateMapper.presetRate-Inline
### The error occurred while setting parameters
### SQL: INSERT INTO list_rate (hotel_id,rate_date,room_type_id,rate,create_time,update_time,operator_id,is_deleted) VALUES (10005,'2018-11-10',1,148, now(),now(),1,0) , (10005,'2018-11-11',1,123, now(),now(),1,0)
### Cause: com.mysql.jdbc.exceptions.jdbc4.MySQLTransactionRollbackException: Deadlock found when trying to get lock; try restarting transaction
; ]; Deadlock found when trying to get lock; try restarting transaction; nested exception is com.mysql.jdbc.exceptions.jdbc4.MySQLTransactionRollbackException: Deadlock found when trying to get lock; try restarting transaction
at org.springframework.jdbc.support.SQLErrorCodeSQLExceptionTranslator.doTranslate(SQLErrorCodeSQLExceptionTranslator.java:266)
at org.springframework.jdbc.support.AbstractFallbackSQLExceptionTranslator.translate(AbstractFallbackSQLExceptionTranslator.java:72)
at org.mybatis.spring.MyBatisExceptionTranslator.translateExceptionIfPossible(MyBatisExceptionTranslator.java:74)
at org.mybatis.spring.SqlSessionTemplate$SqlSessionInterceptor.invoke(SqlSessionTemplate.java:421)
at com.sun.proxy.$Proxy100.insert(Unknown Source)
at org.mybatis.spring.SqlSessionTemplate.insert(SqlSessionTemplate.java:254)
at org.apache.ibatis.binding.MapperMethod.execute(MapperMethod.java:52)
at org.apache.ibatis.binding.MapperProxy.invoke(MapperProxy.java:53)
at com.sun.proxy.$Proxy109.presetRate(Unknown Source)
at com.oyo.price.repository.impl.ListRateRepositoryImpl.presetRate(ListRateRepositoryImpl.java:36)
at com.oyo.price.repository.impl.ListRateRepositoryImpl$$FastClassBySpringCGLIB$$6a37e66d.invoke(<generated>)
at org.springframework.cglib.proxy.MethodProxy.invoke(MethodProxy.java:204)
at org.springframework.aop.framework.CglibAopProxy$CglibMethodInvocation.invokeJoinpoint(CglibAopProxy.java:746)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:163)
at org.springframework.dao.support.PersistenceExceptionTranslationInterceptor.invoke(PersistenceExceptionTranslationInterceptor.java:139)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:185)
at org.springframework.aop.framework.CglibAopProxy$DynamicAdvisedInterceptor.intercept(CglibAopProxy.java:688)
at com.oyo.price.repository.impl.ListRateRepositoryImpl$$EnhancerBySpringCGLIB$$ccfbbc66.presetRate(<generated>)
at com.oyo.price.service.impl.MultithreadServiceImpl.updateListRate(MultithreadServiceImpl.java:60)
at com.oyo.price.service.impl.MultithreadServiceImpl$$FastClassBySpringCGLIB$$e04a972c.invoke(<generated>)
at org.springframework.cglib.proxy.MethodProxy.invoke(MethodProxy.java:204)
at org.springframework.aop.framework.CglibAopProxy$CglibMethodInvocation.invokeJoinpoint(CglibAopProxy.java:746)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:163)
at org.springframework.transaction.interceptor.TransactionAspectSupport.invokeWithinTransaction(TransactionAspectSupport.java:294)
at org.springframework.transaction.interceptor.TransactionInterceptor.invoke(TransactionInterceptor.java:98)
at org.springframework.aop.framework.ReflectiveMethodInvocation.proceed(ReflectiveMethodInvocation.java:185)
at org.springframework.aop.framework.CglibAopProxy$DynamicAdvisedInterceptor.intercept(CglibAopProxy.java:688)
at com.oyo.price.service.impl.MultithreadServiceImpl$$EnhancerBySpringCGLIB$$f81880f4.updateListRate(<generated>)
at com.oyo.price.service.impl.ListRateServiceImpl$UpdateListRateThread.run(ListRateServiceImpl.java:559)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Caused by: com.mysql.jdbc.exceptions.jdbc4.MySQLTransactionRollbackException: Deadlock found when trying to get lock; try restarting transaction
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
at java.lang.reflect.Constructor.newInstance(Constructor.java:423)
at com.mysql.jdbc.Util.handleNewInstance(Util.java:425)
at com.mysql.jdbc.Util.getInstance(Util.java:408)
at com.mysql.jdbc.SQLError.createSQLException(SQLError.java:952)
at com.mysql.jdbc.MysqlIO.checkErrorPacket(MysqlIO.java:3976)
at com.mysql.jdbc.MysqlIO.checkErrorPacket(MysqlIO.java:3912)
at com.mysql.jdbc.MysqlIO.sendCommand(MysqlIO.java:2530)
at com.mysql.jdbc.MysqlIO.sqlQueryDirect(MysqlIO.java:2683)
at com.mysql.jdbc.ConnectionImpl.execSQL(ConnectionImpl.java:2486)
at com.mysql.jdbc.PreparedStatement.executeInternal(PreparedStatement.java:1858)
at com.mysql.jdbc.PreparedStatement.execute(PreparedStatement.java:1197)
at com.alibaba.druid.pool.DruidPooledPreparedStatement.execute(DruidPooledPreparedStatement.java:493)
at sun.reflect.GeneratedMethodAccessor104.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at org.apache.ibatis.logging.jdbc.PreparedStatementLogger.invoke(PreparedStatementLogger.java:59)
at com.sun.proxy.$Proxy152.execute(Unknown Source)
at org.apache.ibatis.executor.statement.PreparedStatementHandler.update(PreparedStatementHandler.java:45)
at org.apache.ibatis.executor.statement.RoutingStatementHandler.update(RoutingStatementHandler.java:73)
at org.apache.ibatis.executor.SimpleExecutor.doUpdate(SimpleExecutor.java:49)
at org.apache.ibatis.executor.BaseExecutor.update(BaseExecutor.java:115)
at org.apache.ibatis.executor.CachingExecutor.update(CachingExecutor.java:75)
at sun.reflect.GeneratedMethodAccessor95.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at org.apache.ibatis.plugin.Plugin.invoke(Plugin.java:63)
at com.sun.proxy.$Proxy113.update(Unknown Source)
at sun.reflect.GeneratedMethodAccessor95.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at org.apache.ibatis.plugin.Invocation.proceed(Invocation.java:49)
at com.oyo.price.mybatis.CatMybatisPlugin.intercept(CatMybatisPlugin.java:84)
at org.apache.ibatis.plugin.Plugin.invoke(Plugin.java:61)
at com.sun.proxy.$Proxy113.update(Unknown Source)
at org.apache.ibatis.session.defaults.DefaultSqlSession.update(DefaultSqlSession.java:170)
at org.apache.ibatis.session.defaults.DefaultSqlSession.insert(DefaultSqlSession.java:157)
at sun.reflect.GeneratedMethodAccessor119.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:498)
at org.mybatis.spring.SqlSessionTemplate$SqlSessionInterceptor.invoke(SqlSessionTemplate.java:408)
... 28 more
发现居然有死锁。
根据常识考虑,我每个线程(事务)更新的数据都不冲突,为什么会产生死锁?
带着这个问题,打印mysql最近一次的死锁信息
show engine innodb status
显示如下
=====================================
2018-12-26 11:58:07 0x4994 INNODB MONITOR OUTPUT
=====================================
Per second averages calculated from the last 21 seconds
-----------------
BACKGROUND THREAD
-----------------
srv_master_thread loops: 35631 srv_active, 0 srv_shutdown, 478534 srv_idle
srv_master_thread log flush and writes: 514165
----------
SEMAPHORES
----------
OS WAIT ARRAY INFO: reservation count 73218
OS WAIT ARRAY INFO: signal count 76026
RW-shared spins 0, rounds 149557, OS waits 72069
RW-excl spins 0, rounds 30643, OS waits 72
RW-sx spins 3137, rounds 22537, OS waits 108
Spin rounds per wait: 149557.00 RW-shared, 30643.00 RW-excl, 7.18 RW-sx
------------------------
LATEST DETECTED DEADLOCK
------------------------
2018-12-26 11:31:01 0x4e84
*** (1) TRANSACTION:
TRANSACTION 4348374, ACTIVE 0 sec inserting
mysql tables in use 1, locked 1
LOCK WAIT 3 lock struct(s), heap size 1136, 2 row lock(s), undo log entries 1
MySQL thread id 142, OS thread handle 16456, query id 279197 localhost 127.0.0.1 root update
INSERT INTO list_rate (hotel_id,rate_date,room_type_id,rate,create_time,update_time,operator_id,is_deleted)
VALUES
(10007,'2018-11-10',1,139,
now(),now(),1,0)
,
(10007,'2018-11-11',1,105,
now(),now(),1,0)
*** (1) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 65 page no 5656 n bits 736 index hotel_date_idx of table `price`.`list_rate` trx id 4348374 lock_mode X locks gap before rec insert intention waiting
Record lock, heap no 457 PHYSICAL RECORD: n_fields 3; compact format; info bits 0
0: len 8; hex 8000000000002b67; asc +g;;
1: len 3; hex 8fc56b; asc k;;
2: len 8; hex 8000000000000164; asc d;;
*** (2) TRANSACTION:
TRANSACTION 4348375, ACTIVE 0 sec inserting
mysql tables in use 1, locked 1
3 lock struct(s), heap size 1136, 2 row lock(s), undo log entries 1
MySQL thread id 146, OS thread handle 20100, query id 279199 localhost 127.0.0.1 root update
INSERT INTO list_rate (hotel_id,rate_date,room_type_id,rate,create_time,update_time,operator_id,is_deleted)
VALUES
(10009,'2018-11-10',1,129,
now(),now(),1,0)
,
(10009,'2018-11-11',1,88,
now(),now(),1,0)
*** (2) HOLDS THE LOCK(S):
RECORD LOCKS space id 65 page no 5656 n bits 736 index hotel_date_idx of table `price`.`list_rate` trx id 4348375 lock_mode X locks gap before rec
Record lock, heap no 457 PHYSICAL RECORD: n_fields 3; compact format; info bits 0
0: len 8; hex 8000000000002b67; asc +g;;
1: len 3; hex 8fc56b; asc k;;
2: len 8; hex 8000000000000164; asc d;;
*** (2) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 65 page no 5656 n bits 736 index hotel_date_idx of table `price`.`list_rate` trx id 4348375 lock_mode X locks gap before rec insert intention waiting
Record lock, heap no 457 PHYSICAL RECORD: n_fields 3; compact format; info bits 0
0: len 8; hex 8000000000002b67; asc +g;;
1: len 3; hex 8fc56b; asc k;;
2: len 8; hex 8000000000000164; asc d;;
*** WE ROLL BACK TRANSACTION (2)
------------
TRANSACTIONS
------------
Trx id counter 4348700
Purge done for trx's n:o < 4348700 undo n:o < 0 state: running but idle
History list length 11
LIST OF TRANSACTIONS FOR EACH SESSION:
---TRANSACTION 283745059479304, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059478432, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059477560, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059476688, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059475816, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059474944, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059474072, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
---TRANSACTION 283745059473200, not started
0 lock struct(s), heap size 1136, 0 row lock(s)
--------
FILE I/O
--------
I/O thread 0 state: wait Windows aio (insert buffer thread)
I/O thread 1 state: wait Windows aio (log thread)
I/O thread 2 state: wait Windows aio (read thread)
I/O thread 3 state: wait Windows aio (read thread)
I/O thread 4 state: wait Windows aio (read thread)
I/O thread 5 state: wait Windows aio (read thread)
I/O thread 6 state: wait Windows aio (write thread)
I/O thread 7 state: wait Windows aio (write thread)
I/O thread 8 state: wait Windows aio (write thread)
I/O thread 9 state: wait Windows aio (write thread)
Pending normal aio reads: [0, 0, 0, 0] , aio writes: [0, 0, 0, 0] ,
ibuf aio reads:, log i/o's:, sync i/o's:
Pending flushes (fsync) log: 0; buffer pool: 0
3705 OS file reads, 393906 OS file writes, 287002 OS fsyncs
0.00 reads/s, 0 avg bytes/read, 0.86 writes/s, 0.76 fsyncs/s
-------------------------------------
INSERT BUFFER AND ADAPTIVE HASH INDEX
-------------------------------------
Ibuf: size 1, free list len 789, seg size 791, 0 merges
merged operations:
insert 0, delete mark 0, delete 0
discarded operations:
insert 0, delete mark 0, delete 0
Hash table size 34679, node heap has 1 buffer(s)
Hash table size 34679, node heap has 2 buffer(s)
Hash table size 34679, node heap has 0 buffer(s)
Hash table size 34679, node heap has 0 buffer(s)
Hash table size 34679, node heap has 2 buffer(s)
Hash table size 34679, node heap has 0 buffer(s)
Hash table size 34679, node heap has 0 buffer(s)
Hash table size 34679, node heap has 1 buffer(s)
0.24 hash searches/s, 0.10 non-hash searches/s
---
LOG
---
Log sequence number 1153015126
Log flushed up to 1153015126
Pages flushed up to 1153015126
Last checkpoint at 1153015117
0 pending log flushes, 0 pending chkp writes
180065 log i/o's done, 0.48 log i/o's/second
----------------------
BUFFER POOL AND MEMORY
----------------------
Total large memory allocated 137297920
Dictionary memory allocated 164031
Buffer pool size 8192
Free buffers 3424
Database pages 4762
Old database pages 1737
Modified db pages 0
Pending reads 0
Pending writes: LRU 0, flush list 0, single page 0
Pages made young 2, not young 0
0.00 youngs/s, 0.00 non-youngs/s
Pages read 3615, created 1147, written 177183
0.00 reads/s, 0.00 creates/s, 0.29 writes/s
Buffer pool hit rate 1000 / 1000, young-making rate 0 / 1000 not 0 / 1000
Pages read ahead 0.00/s, evicted without access 0.00/s, Random read ahead 0.00/s
LRU len: 4762, unzip_LRU len: 0
I/O sum[0]:cur[0], unzip sum[0]:cur[0]
--------------
ROW OPERATIONS
--------------
0 queries inside InnoDB, 0 queries in queue
0 read views open inside InnoDB
Process ID=3504, Main thread ID=4272, state: sleeping
Number of rows inserted 15970, updated 46294, deleted 15313, read 1720295
0.00 inserts/s, 0.10 updates/s, 0.00 deletes/s, 0.24 reads/s
----------------------------
END OF INNODB MONITOR OUTPUT
============================
发现事务1在等待一个锁
*** (1) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 65 page no 5656 n bits 736 index hotel_date_idx of table `price`.`list_rate` trx id 4348374 lock_mode X locks gap before rec insert intention waiting
事务2也在等待一个锁
*** (2) WAITING FOR THIS LOCK TO BE GRANTED:
RECORD LOCKS space id 65 page no 5656 n bits 736 index hotel_date_idx of table `price`.`list_rate` trx id 4348375 lock_mode X locks gap before rec insert intention waiting
而且事物2持有了事物1需要的锁
*** (2) HOLDS THE LOCK(S):
RECORD LOCKS space id 65 page no 5656 n bits 736 index hotel_date_idx of table `price`.`list_rate` trx id 4348375 lock_mode X locks gap before rec
关于锁的描述,出现了lock_mode,gap before rec,insert intention等字眼,看不懂说明了什么?说明我关于mysql的锁相关的知识储备还不够。那就开始调查mysql的锁相关知识。
通过搜索引擎,
获取到如下知识:
InnoDB是一个支持行锁的存储引擎,锁的类型有:共享锁(S)、排他锁(X)、意向共享(IS)、意向排他(IX)
InnoDB还将锁细分为如下几种子类型:
-
record lock(RK)
锁直接加在索引记录上面,锁住的是key -
gap lock(GK)
间隙锁,锁定一个范围,但不包括记录本身。GAP锁的目的,是为了防止同一事务的两次当前读,出现幻读的情况 -
next key lock(NK)
RK+GK -
insert intention lock(IK)
如果插入前,该间隙已经由gap锁,那么Insert会申请插入意向锁。因为了避免幻读,当其他事务持有该间隙的间隔锁,插入意向锁就会被阻塞。
锁的持有兼容程度如下表
请求锁\持有锁 | gap lock | insert intention lock | record lock | next key lock |
---|---|---|---|---|
gap lock | 兼容 | 兼容 | 兼容 | 兼容 |
insert intention lock | 冲突 | 兼容 | 兼容 | 冲突 |
record lock | 兼容 | 兼容 | 冲突 | 冲突 |
next key lock | 兼容 | 兼容 | 冲突 | 冲突 |
那么再回到死锁日志,可以知道 :
事务1正在获取插入意向锁
事务2正在获取插入意向锁,持有排他gap锁
再看我们上面的锁兼容表格,可以知道,gap lock和insert intention lock是不兼容的
那么就可以推断出:事务1持有gap lock,等待事务2的insert intention lock释放;事务2持有gap lock,等待事务1的insert intention lock释放,从而导致死锁。
那么新的问题就来了,事务1的intention lock 为什么会和事务2的gap lock 有交集,或者说,事务1要插入的数据的位置为什么会被事务2给锁住?
让我回顾一下gap lock的定义:
间隙锁,锁定一个范围,但不包括记录本身。GAP锁的目的,是为了防止同一事务的两次当前读,出现幻读的情况
那为什么是gap lock,gap lock到底是基于什么逻辑锁的记录?发现自己相关的知识储备还不够。那就开始调查。
调查后发现,当当前索引是一个普通索引的时候,会加一个gap lock来防止幻读,此gap lock 会锁住一个左开右闭的区间。假设索引为xx_idx(xx_id),数据分布为1,4,6,8,12,当更新xx_id=9的时候,这个时候gap lock的锁定记录区间就是(8,12],也就是锁住了xxid in (9,10,11,12)的数据,当有其他事务要插入xxid in (9,10,11,12)的数据时,就会处于等待获取锁的状态。
ps:当前索引不是普通索引,而且是唯一索引等其他情况,请参考下面资料
MySQL 加锁处理分析
回到我自己的案例中,重新屡一下事务1的执行过程:
UPDATE list_rate
SET is_deleted = 1, update_time = now(), operator_id = 1
WHERE hotel_id = 10007 AND room_type_id = 1 AND is_deleted = 0 AND rate_date IN ( '2018-11-10' , '2018-11-11' )
因为普通索引
KEY
hotel_date_idx(
hotel_id,
rate_date)
的关系 这段sql会获取一个gap lock,范围(2,11111]
INSERT INTO list_rate (hotel_id,rate_date,room_type_id,rate,create_time,update_time,operator_id,is_deleted)
VALUES (10007,'2018-11-10',1,144, now(),now(),1,0) , (10007,'2018-11-11',1,148, now(),now(),1,0)
这段sql会获取一个insert intention lock (waiting)
再看事务2的执行过程
UPDATE list_rate
SET is_deleted = 1, update_time = now(), operator_id = 1
WHERE hotel_id = 10009 AND room_type_id = 1 AND is_deleted = 0 AND rate_date IN ( '2018-11-10' , '2018-11-11' )
因为普通索引
KEY
hotel_date_idx(
hotel_id,
rate_date)
的关系 这段sql也会获取一个gap lock,范围也是(2,11111](根据前面的知识,gap lock之间会互相兼容,可以一起持有锁的)
INSERT INTO list_rate (hotel_id,rate_date,room_type_id,rate,create_time,update_time,operator_id,is_deleted)
VALUES (10009,'2018-11-10',1,144, now(),now(),1,0) , (10009,'2018-11-11',1,148, now(),now(),1,0)
这段sql也会获取一个insert intention lock (waiting)
看到这里,基本也就破案了。因为普通索引的关系,事务1和事务2的gap lock的覆盖范围太广,导致其他事务无法插入数据。
重新梳理一下:
transaction1 | transaction2 |
---|---|
update (10007,'2018-11-11') gap lock (2,11111] | |
update (10009,'2018-11-11') gap lock (2,11111] | |
insert wait lock | |
insert wait lock | |
dead lock roll back | |
done |
所以从结果来看,一堆事务被回滚,只有10007数据被更新成功
结论
gap lock 导致了并发处理的死锁
处理
在mysql默认的事务隔离级别(repeatable read)下,无法避免这种情况。只能把并发处理改成同步处理。或者从业务层面做处理。
知识库levelup
共享锁、排他锁、意向共享、意向排他
record lock、gap lock、next key lock、insert intention lock
show engine innodb status