ORA-00060 Deadlock detected

今天一套10.2.0.4的数据库报ORA-00060 Deadlock detected。

trace部分内容：

Dump file /oracle/oracle/product/10.2.0/admin/APSDB/udump/apsdb_ora_1008062.trc

Oracle Database10gEnterprise Edition Release10.2.0.4.0 - 64bit Production

With the Partitioning, OLAP, Data Mining and Real Application Testing options

ORACLE_HOME = /oracle/oracle/product/10.2.0

System name: AIX

Node name: jzdb-b

Release: 3

Version: 5

Machine: 00C1A5C54C00

Instance name: APSDB

Redo thread mounted by this instance: 1

Oracle process number: 63

Unix process pid: 1008062, image: oracle@jzdb-b

*** 2011-09-26 10:54:37.137

*** SERVICE NAME:(SYS$USERS) 2011-09-26 10:54:37.129

*** SESSION ID:(205.28158) 2011-09-26 10:54:37.129

DEADLOCK DETECTED ( ORA-00060 )

[Transaction Deadlock]

The following deadlock is not an ORACLE error. It is a

deadlock due to user error in the design of an application

or from issuing incorrect ad-hoc SQL. The following

information may aid in determining the deadlock:

Deadlock graph:

---------Blocker(s)-------- ---------Waiter(s)---------

Resource Name process session holds waits process session holds waits

TX-00010023-00004fdb 63 205 X 62 179 X

TX-000a0025-00009f87 62 179 X 63 205 X

session 205: DID 0001-003F-000095E7session 179: DID 0001-003E-00009E3A

session 179: DID 0001-003E-00009E3Asession 205: DID 0001-003F-000095E7

Rows waited on:

Session 179: obj - rowid = 00002AD8 - AAACrYAAEAAAIkRABP

(dictionary objn - 10968, file - 4, block - 35089, slot - 79)

Session 205: obj - rowid =00002A16 - AAACoWAAEAAAJisAAF

(dictionary objn - 10774, file - 4, block - 39084, slot - 5)

Information on the OTHER waiting sessions:

Session 179:

pid=62 serial=42457 audsid=12628802 user: 28/CNAPS

O/S info: user: , term: , ospid: 1234, machine: rcca-web

program:

Current SQL Statement:

update JJ_SHENGQINGJIAN5 set YINHANG_NO=:1, SAVING=:2, STOCKS=:3, HOUSEREGION=:4, HOUSEAREA=:5, HOUSEPRICE=:6, OTHER_CREDITLIMIT=:7, BRANCHSUGGESTLIMIT=:8, VEHICLETYPE=:9, VEHICLEPRICE=:10, VEHICLEWEIGHT=:11, VEHICLEREGISTDATE=:12, REGISTFUND=:13, REGISTDATE=:14, REFTEL1BYCREDIT=:15, REFTEL2BYCREDIT=:16, REFERENCETEL=:17, CONFIRMCONTENT=:18, COMMENTINFOBYBRANCH=:19, YEARINCOMEBYJJ=:20, INCOMEBYCREDIT=:21, BUSINESSLICENSE=:22, NOTERESULT=:23, NOTEOVERFLG=:24, OVER_RECHECK_FLAG=:25, RISK_CHECK_FLAG=:26, REFUSE=:27, PRIMARYLIMIT=:28, SECONDARYLIMITPERCENT=:29, SECONDARYLIMIT=:30, REASONCD1=:31, REASONCD2=:32, REASONCD3=:33, REASONCD4=:34, REASONCD5=:35, EXPLAIN=:36, RFE_TIME=:37, PINION=:38, VERFLG=:39, THERS=:40, PINE=:41, REFTELOPINE=:42, CREATER=:43, INSERTTIME=:44, UPDATOR=:45, UPDATETIME=:46, ID=:47, STOCKSBYCREDIT=:48, SAVINGBYCREDIT=:49, HOLD_CARD_FLG=:50, CHECKCREDIT_FLG=:51, HUNTER_FLG=:52, REGISTFUNDBYCREDIT=:53, VEHICLEPRICEBYCREDIT=:54, HOUSEPRICEBYCREDIT=:55, CANCEL_REASON=:56, MASTER_AUDIT_REASON=:57, FINAL_REASON=:58, FINAL_TIME=:59, LCOPINE=:60, income_flag=:61, inspection_flag=:62, TVCODE=:63, SCCODE=:64, CTCODE=:65, ACCODE=:66, ICCODE=:67, CCODE=:68, PCCODE=:69, RCCODE=:70, FTECODE=:71, FINALFLAG=:72, docstart=:73, docend=:74 where SHENQINGJIAN_ID=:75

End of information on OTHER waiting sessions.

Current SQL statement for this session:

update JBPM_TASKINSTANCE set NAME_=:1, DESCRIPTION_=:2, ACTORID_=:3, CREATE_=:4, START_=:5, END_=:6, DUEDATE_=:7, PRIORITY_=:8, ISCANCELLED_=:9, ISSUSPENDED_=:10, ISOPEN_=:11, ISSIGNALLING_=:12, ISBLOCKING_=:13, TASK_=:14, TOKEN_=:15, SWIMLANINSTANCE_=:16, TASKMGMTINSTANCE_=:17 where ID_=:18

简单说明一下，从Deadlock graph中，确认以下信息：

---------Blocker(s)-------- ---------Waiter(s)---------

Resource Name process session holds waits process session holds waits

TX-00010023-00004fdb 63 205 X 62 179 X

TX-000a0025-00009f87 62 179 X 63 205 X

从这里可以看到，session 179以排他方式(X)拥有资源TX-00010023-00004fdb，阻塞了session 205对该资源的申请。Session 205以排他方式(X)拥有资源TX-000a0025-00009f87，阻塞了session 179对该资源的申请，从而导致了死锁。

这里的TX-00010023-00004fdb，TX代表锁的类型，常见的有：

TM - DML enqueue

TX - Transaction enqueue

UL - User supplied

00010023和00004fdb分别对应的是v$lock中ID1和ID2字段。

对于TX类型锁，通过ID1可以换算出事务的回滚段号、slot号，ID2代表事务的序列号。

先看00010023，0001代表回滚段号，换算成十进制为1，0023为事务的slot号，换算成十进制为35。而00004fdb换算成十进制为20443。

所以TX-00010023-00004fdb对应v$transaction中XIDUSN=1，XIDSLOT=35，XIDSQN=20443的事务。

当然也可以这么换算，00010023对应十进制的65571。

SQL> select trunc(65571/65536),mod(65571,65536) from dual;

TRUNC(65571/65536) MOD(65571,65536)

------------------ ----------------

1 35

得到XIDUSN=1，XIDSLOT=35。

对于TM类型锁，ID1代表object id，ID2总是为0。

Lock type	TX
ID1	ID1 indicates the rollback segment and the slot number. On 32-bit machines, you must convert ID1 into hexadecimal and left pad with zeros up to 8 characters. The high order 2 bytes give the rollback segment number, and the low order 2 bytes give the rollback slot number. The values can be seen in the XIDUSN and XIDSLOT columns of the V$TRANSACTION view.
ID2	Rollback segment wrap or sequence number. This value can be seen in the XIDSQN column of V$TRANSACTION view.
Lock type	TM
ID1	ID1 indicates the object ID of the table, which can be found in DBA_OBJECTS.OBJECT_ID.
ID2	Always 0.

接下来trace信息记录了DID信息：

session 205: DID 0001-003F-000095E7session 179: DID 0001-003E-00009E3A

session 179: DID 0001-003E-00009E3Asession 205: DID 0001-003F-000095E7

DID即deadlock id，这里以0001-003F-000095E7为例：

0001：换算成十进制为1，代表的是实例号

003F：换算为十进制为63，代表ORACLE PID，即ORACLE进程号

000095E7：换算为十进制为38375，类似于v$session.serial#，ORACLE PID的一个串行值。

同时，下面的trace信息，描述了死锁发生时对应会话的object id和rowid。

Session 179: obj - rowid = 00002AD8 - AAACrYAAEAAAIkRABP

(dictionary objn - 10968, file - 4, block - 35089, slot - 79)

Session 205: obj - rowid =00002A16 - AAACoWAAEAAAJisAAF

(dictionary objn - 10774, file - 4, block - 39084, slot - 5)

这里的会话179，对象号是00002AD8，rowid是AAACrYAAEAAAIkRABP。

后面的括号内容(dictionary objn - 10968, file - 4, block - 35089, slot - 79)已经很清楚地翻译了这些信息，这里不再详细说明了。

最后，我们可以看到死锁时会话执行的sql语句。

Session 179:

Session 205:

有了以上信息，就很方便我们定位死锁的原因了。

ORA-00060 Deadlock detected相关推荐

Oracle ORA-00060：Deadlock detected. More info in file /var/oracle/app/diag/rdbms/...记一次位图索引误用引发卡顿
Oracle ORA-00060:Deadlock detected. More info in file /var/oracle/app/diag/rdbms/-记一次位图索引误用带来的数据库卡顿死 ...
Linux ora-00060,解决 ORA-00060: Deadlock detected 小例
数据库版本: SQL > select * from v$version; BANNER ---------------------------------------------------- ...
ORA-00060: Deadlock detected（场景模拟)
from:http://www.cnblogs.com/killkill/archive/2010/09/12/1824650.html ORA-00060: Deadlock detected(场景 ...
追根溯源-数据库deadlock重启
每周日必出故障系列,似乎是看不惯我周末休息,老美的圣诞节也没镇住老美的数据库. 数据库在圣诞节最后一刻2节点宕机了!!!!当然按照重启法则似乎能解决一切,但是这里并不适用,重启之后故障重现,给了我足够 ...
Know more about Enqueue Deadlock Detection
今天在 ORACLE ALLSTAR群里讨论了一个关于队列锁死锁检测的问题,原帖子在这里. 有同学指出对于enqueue lock的死锁检测应当是每3 秒钟检测一次,这样说的依据是通过一个简单可重复的 ...
Identify ksusetxn DID:An Deadlock ID
我们在查看10704 event trace(Print out information about what enqueues are being obtained)或deadlock detect ...
Global Enqueue Services Deadlock --WFG的诊断
Normal 0 7.8 磅 0 2 false false false MicrosoftInternetExplorer4 今天客户数据库报了一个死锁错误,alert.log里信息比较简单 Nor ...
关于Oracle死锁的知识点总结
源地址 : https://www.cnblogs.com/kerrycode/p/6928263.html 李华荣总结的死锁 1 BLOG文档结构图 wpsA550.tmp 2 前言部分 2.1 导 ...
归档—监控ORACLE数据库告警日志
ORACLE的告警日志里面包含许多有用的信息,尤其是一些ORACLE的ORA错误信息,所以有必要及时归档.监控数据库告警日志的ORA错误,及时提醒数据库管理员DBA处理这些错误信息,那么我们首先来看看 ...

ORA-00060 Deadlock detected

ORA-00060 Deadlock detected相关推荐

最新文章

热门文章