分析一例 TX Enqueue contention案例

应用反馈某个业务比较慢,需要紧急处理
查询等待事件

SQL> select * from v$version;

BANNER
----------------------------------------------------------------
Oracle9i Enterprise Edition Release 9.2.0.8.0 - 64bit Production
PL/SQL Release 9.2.0.8.0 - Production
CORE    9.2.0.8.0       Production
TNS for IBM/AIX RISC System/6000: Version 9.2.0.8.0 - Production
NLSRTL Version 9.2.0.8.0 - Production

SQL> select A.INST_ID,count(*), event
  2    from Gv$session_wait a
  3  where event not in ('SQL*Net more data to client',
  4  'rdbms ipc message',
  5               'smon timer',
  6               'pmon timer',
  7               'SQL*Net message from client',
  8               'lock manager wait for remote message',
  9               'ges remote message',
 10               'gcs remote message',
 11               'gcs for action',
 12               'client message',
 13               'pipe get',
 14               'null event',
 15               'PX Idle Wait',
 16               'single-task message',
 17               'PX Deq: Execution Msg',
 18               'KXFQ: kxfqdeq - normal deqeue',
 19               'listen endpoint status',
 20               'slave wait',
 21               'wakeup time manager','jobq slave wait')
 22  group by INST_ID,event
 23  order by 1 desc,2 desc;

   INST_ID   COUNT(*) EVENT
---------- ---------- --------------------------------------
         2          8 enqueue
         2          1 async disk IO
         2          1 db file sequential read
         2          1 SQL*Net message to client
         2          1 PX Deq: reap credit
         1          2 global cache cr request
         1          1 async disk IO
         1          1 PX Deq: reap credit
         1          1 PX Deq: Execute Reply

9 rows selected.

发现enqueue等待有些多,怀疑是TX enquenue,查询阻塞者

SQL> set linesize 100
SQL> set pagesize 66
SQL> col c1 for a15
SQL> col c1 heading "Program Name "
SQL> select l.inst_id,l.SID,program c1,l.TYPE,l.ID1,l.ID2,l.LMODE,l.REQUEST
  2  from gv$lock l,gv$session s
  3  where l.type like 'TX' and l.REQUEST =6
  4  and l.inst_id=s.inst_id and l.sid=s.sid
  5  order by id1
  6  / 

   INST_ID        SID Program Name    TY        ID1        ID2      LMODE    REQUEST
---------- ---------- --------------- -- ---------- ---------- ---------- ----------
         2        295 rtStopMain@zwq_ TX    1441805    2391806          0          6
                      bill_2 (TNS V1-
                      V3)

         2        992 rtStopMain@zwq_ TX    1441805    2391806          0          6
                      bill_2 (TNS V1-
                      V3)

         2       1238 rtStopMain@zwq_ TX    6946827    2546365          0          6
                      bill_2 (TNS V1-
                      V3)

         2       1298 rtStopMain@zwq_ TX    6946827    2546365          0          6
                      bill_2 (TNS V1-
                      V3)

         2       1684 rtStopMain@zwq_ TX    6946827    2546365          0          6
                      bill_2 (TNS V1-
                      V3)

         2       1553 rtStopMain@zwq_ TX    6946827    2546365          0          6
                      bill_2 (TNS V1-
                      V3)

         2         75 rtStopMain@zwq_ TX   12451856     199146          0          6
                      bill_2 (TNS V1-
                      V3)

         2       1125 rtStopMain@zwq_ TX   14352404      63837          0          6
                      bill_2 (TNS V1-
                      V3)

查询持有者

SQL> set linesize 100
SQL> set pagesize 66
SQL> col c1 for a15
SQL> col c1 heading "Program Name "
SQL> select l.inst_id,l.SID,program c1,l.TYPE,l.ID1,l.ID2,l.LMODE,l.REQUEST
  2  from gv$lock l,gv$session s
  3  where l.type like 'TX' and l.LMODE =6 and (l.ID1,l.ID2) in
  4  (select id1,id2 from gv$lock where type like 'TX' and REQUEST =6)
  5  and l.inst_id=s.inst_id and l.sid=s.sid
  6  order by id1
  7  / 

   INST_ID        SID Program Name    TY        ID1        ID2      LMODE    REQUEST
---------- ---------- --------------- -- ---------- ---------- ---------- ----------
         2         75 rtStopMain@zwq_ TX    1441805    2391806          6          0
                      bill_2 (TNS V1-
                      V3)

         2        992 rtStopMain@zwq_ TX    6946827    2546365          6          0
                      bill_2 (TNS V1-
                      V3)

         2        295 rtStopMain@zwq_ TX   12451856     199146          6          0
                      bill_2 (TNS V1-
                      V3)

         2       1553 rtStopMain@zwq_ TX   14352404      63837          6          0
                      bill_2 (TNS V1-
                      V3)

通过持有者和阻塞者可以得出:
1.持有者和阻塞者都是在2号实例上
2.持有者75阻塞了295/992的会话
3.持有者992阻塞了1238/1298/1684/1553的会话
4.持有者295阻塞了75的会话
5.持有者1553阻塞了1125的会话
6.同时分析发现,所有的持有者sid也在阻塞者中,也就是持有者阻塞了某个sid,而自身又被其他sid给阻塞,形成了多级阻塞或者环.如:75阻塞了295,而295有阻塞了75;992阻塞了1553,而1553阻塞了1125

查询阻塞和持有者对象

SQL> set linesize 110
SQL> col c0 for 999
SQL> col c0 heading "INS"
SQL> col c1 for a15
SQL> col c1 heading "Program Name "
SQL> select inst_id c0,sid,program c1,ROW_WAIT_OBJ# object_no, ROW_WAIT_FILE# Rfile_no,
  2  ROW_WAIT_BLOCK# Block_no ,ROW_WAIT_ROW# Row_no
  3  from gv$session
  4  where (inst_id,sid) in (select inst_id,sid from gv$session_wait where p1='1415053318')
  5  / 

 INS        SID Program Name     OBJECT_NO   RFILE_NO   BLOCK_NO     ROW_NO
---- ---------- --------------- ---------- ---------- ---------- ----------
   2         75 rtStopMain@zwq_    1323132         13     122601        111
                bill_2 (TNS V1-
                V3)

   2        295 rtStopMain@zwq_    1323132         13     122601        100
                bill_2 (TNS V1-
                V3)

   2        992 rtStopMain@zwq_    1323132         13     122601        101
                bill_2 (TNS V1-
                V3)

   2       1125 rtStopMain@zwq_    1323132         84      38445         70
                bill_2 (TNS V1-
                V3)

   2       1238 rtStopMain@zwq_    1323132         15     255066         41
                bill_2 (TNS V1-
                V3)

   2       1298 rtStopMain@zwq_    1323132         14     118411          8
                bill_2 (TNS V1-
                V3)

   2       1553 rtStopMain@zwq_    1323132         15     255066         19
                bill_2 (TNS V1-
                V3)

   2       1684 rtStopMain@zwq_    1323132         14     118411         21
                bill_2 (TNS V1-
                V3)

8 rows selected.

SQL> set linesize 100
SQL> set pagesize 100
SQL> col owner for a10
SQL> col object_name for a20
SQL> col object_type for a10
SQL> select owner,object_name,object_id,object_type
  2  from dba_objects
  3  where
  4  object_id in (select ROW_WAIT_OBJ# from gv$session
  5  where (inst_id, sid) in (select inst_id,sid from gv$session_wait where p1='1415053318'))
  6  / 
 

OWNER      OBJECT_NAME           OBJECT_ID OBJECT_TYP
---------- -------------------- ---------- ----------
DBACCADM   DCUSTCREDITBALANCE      1323132 TABLE

通过查询的出来,所有操作的聚焦点都是在DBACCADM.DCUSTCREDITBALANCE表上面

查询相关sql语句

SQL> SQL> set linesize 120
SQL> set pagesize 66
SQL> col c0 for 999
SQL> col c0 heading "INS"
SQL> col c1 for a9
SQL> col c1 heading "OS User"
SQL> col c2 for a9
SQL> col c2 heading "Oracle User"
SQL> col c3 for a15
SQL> col c3 heading "Program Name"
SQL> col b1 for a9
SQL> col b1 heading "Unix PID"
SQL> col b2 for 9999 justify left
SQL> col b2 heading "ORA SID"
SQL> col b3 for 999999 justify left
SQL> col b3 heading "SERIAL#"
SQL> col sql_text for a45
SQL> set space 1
SQL> break on b1 nodup on c0 nodup on c3 nodup on c1 nodup on c2 nodup on b2 nodup on b3 skip 2
SQL> select a.inst_id c0,b.sid b2,c.spid b1, b.program c3, b.username c2,b.serial# b3, a.sql_text
  2    from gv$sql a, gv$session b, gv$process c
  3   where
  4     a.address = b.sql_address
  5     and b.paddr = c.addr
  6     and a.hash_value = b.sql_hash_value
  7     and a.inst_id=b.inst_id and a.inst_id=c.inst_id
  8     and a.inst_id like '&inst_id' and b.sid like '&sid'
  9   order by c.spid,a.hash_value
 10  / 
Enter value for inst_id: 2
Enter value for sid: 75
old   8:    and a.inst_id like '&inst_id' and b.sid like '&sid'
new   8:    and a.inst_id like '2' and b.sid like '75'

 INS ORA SID Unix PID  Program Name    Oracle Us SERIAL# SQL_TEXT
---- ------- --------- --------------- --------- ------- ---------------------------------------------
   2      75 1167392   rtStopMain@zwq_ DBCUSTOPR   42815 update dcustcreditbalance  set limit_owe=:b0,
                       bill_2 (TNS V1-                   unbill_fee=:b1,prepay_fee=:b2,owe_fee=:b3,op_
                       V3)                               time=sysdate where id_no=:b4




SQL> /
Enter value for inst_id: 2
Enter value for sid: 992
old   8:    and a.inst_id like '&inst_id' and b.sid like '&sid'
new   8:    and a.inst_id like '2' and b.sid like '992'

 INS ORA SID Unix PID  Program Name    Oracle Us SERIAL# SQL_TEXT
---- ------- --------- --------------- --------- ------- ---------------------------------------------
   2     992 2760870   rtStopMain@zwq_ DBCUSTOPR   56282 update dcustcreditbalance  set limit_owe=:b0,
                       bill_2 (TNS V1-                   unbill_fee=:b1,prepay_fee=:b2,owe_fee=:b3,op_
                       V3)                               time=sysdate where id_no=:b4




SQL> /
Enter value for inst_id: 2
Enter value for sid: 295
old   8:    and a.inst_id like '&inst_id' and b.sid like '&sid'
new   8:    and a.inst_id like '2' and b.sid like '295'

 INS ORA SID Unix PID  Program Name    Oracle Us SERIAL# SQL_TEXT
---- ------- --------- --------------- --------- ------- ---------------------------------------------
   2     295 1639008   rtStopMain@zwq_ DBCUSTOPR   35740 update dcustcreditbalance  set limit_owe=:b0,
                       bill_2 (TNS V1-                   unbill_fee=:b1,prepay_fee=:b2,owe_fee=:b3,op_
                       V3)                               time=sysdate where id_no=:b4

其他阻塞者和持有者执行sql语句均和该语句相同,省略其他查询.通过这些查询可以确定是因为对dcustcreditbalance表的更新操作导致了这样的现象发生.

处理方案
1.临时处理方案:kill掉持有者
2.永久处理方案:修改这部分程序业务逻辑

发表在 Oracle性能优化 | 标签为 | 评论关闭

关闭数据库出现ORA-00379错误

关闭数据库出现ORA-00379错误

SQL> shutdown immediate
ORA-00604: 递归 SQL 级别 1 出现错误
ORA-00379: 缓冲池 DEFAULT 中无法提供 8K 块大小的空闲缓冲区

查看内存分配

SQL> show parameter sga;

NAME                                 TYPE        VALUE
------------------------------------ ----------- -----------
lock_sga                             boolean     FALSE
pre_page_sga                         boolean     FALSE
sga_max_size                         big integer 412M
sga_target                           big integer 0

SQL> select * from v$sgainfo;

NAME                                  BYTES RES
-------------------------------- ---------- ---
Fixed SGA Size                      1333676 No
Redo Buffers                        6078464 No
Buffer Cache Size                 104857600 Yes
Shared Pool Size                  142606336 Yes
Large Pool Size                     4194304 Yes
Java Pool Size                     12582912 Yes
Streams Pool Size                         0 Yes
Shared IO Pool Size                       0 Yes
Granule Size                        4194304 No
Maximum SGA Size                  431038464 No
Startup overhead in Shared Pool    46137344 No
Free SGA Memory Available         159383552

--spfile中分配情况
orcl.__db_cache_size=104857600
orcl.__java_pool_size=12582912
orcl.__large_pool_size=4194304
orcl.__pga_aggregate_target=104857600
orcl.__sga_target=281018368
orcl.__shared_io_pool_size=0
orcl.__shared_pool_size=142606336
orcl.__streams_pool_size=0

--初始化参数
*.sga_max_size=0
*.sga_target=536870912
*.memory_max_target=536870912
*.memory_target=536870912

alert日志

Mon Jul 02 11:30:19 2012
DIA0 started with pid=8, OS id=1520 
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_dia0_1520.trc  (incident=10883):
ORA-04030: out of process memory when trying to allocate 29916 bytes (heap_ksdhngreq,msg_body:ksdhng)
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_dia0_1520.trc  (incident=10884):
ORA-04030: out of process memory when trying to allocate 8204 bytes (diag pga,dbgtbDefaultBucket)
ORA-04030: out of process memory when trying to allocate 29916 bytes (heap_ksdhngreq,msg_body:ksdhng)
ORA-4030 : opidrv aborting process DIA0 ospid (1348_1520)
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_dia0_1520.trc  (incident=12013):
ORA-04030: out of process memory when trying to allocate 8204 bytes (diag pga,dbgtbDefaultBucket)
ORA-04030: out of process memory when trying to allocate 8204 bytes (diag pga,dbgtbDefaultBucket)
ORA-04030: out of process memory when trying to allocate 29916 bytes (heap_ksdhngreq,msg_body:ksdhng)
Process debug not enabled via parameter _debug_enable

Mon Jul 02 11:33:19 2012
Trace dumping is performing id=[cdmp_20120702113319]
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_smon_1820.trc:
ORA-00379: no free buffers available in buffer pool DEFAULT for block size 8K
ORA-00379: no free buffers available in buffer pool DEFAULT for block size 8K
Mon Jul 02 11:33:49 2012
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_smon_1820.trc:
ORA-00379: no free buffers available in buffer pool DEFAULT for block size 8K
Mon Jul 02 11:34:38 2012
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_smon_1820.trc:
ORA-00379: no free buffers available in buffer pool DEFAULT for block size 8K
Mon Jul 02 11:37:05 2012
Errors in file e:\app\administrator\diag\rdbms\orcl\orcl\trace\orcl_ora_2400.trc:
ORA-00379: 缓冲池 DEFAULT 中无法提供 8K 块大小的空闲缓冲区
ORA-00379: 缓冲池 DEFAULT 中无法提供 8K 块大小的空闲缓冲区
ORA-00379: 缓冲池 DEFAULT 中无法提供 8K 块大小的空闲缓冲区

Tue Jul 03 09:58:06 2012
WARNING: sga_target 432013312 cannot be more than memory_target (432013312) - 
pga_aggregate_target (104857600/0) or untunable pga 104857600, 73783296

通过这里可以看出,系统的data buffe和pga都有内存不足的报错.

解决问题
问题的原因是由于内存分配不多,导致sga组件被消耗完,现在数据库不能正常关闭,修改了相关的内存参数的配置[避免该bug采用asmm内存管理]也无法生效,现在需要做的任务是重启数据库.导致数据库不能被关闭的原因是因为data buffer中的脏数据不能写入新数据.查询MOS发现是Bug 7702085.正常关闭库解决办法手工刷sga组件,然后升级数据库到11.2.0.1 (Base Release)/11.1.0.7.3 (Patch Set Update)/11.1.0.7 Patch 25 on Windows Platforms

SQL>  alter system flush BUFFER_CACHE;

System altered.

SQL> alter system flush  SHARED_POOL;

System altered.

SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
发表在 ORA-xxxxx | 标签为 | 一条评论

使用copy实现long类型转移表空间

在一次8.1.6的数据库恢复过程中,发现一个表空间的数据文件损坏,在转移该表空间相关表时,遇到让人郁闷的long类型.不能使用ctas和move来实现转移,最后通过古老的copy来实现该项工作.
模拟LONG类型表

SQL> create table chf.t_long (id number,name long) tablespace ts_xifenfei;

Table created.

SQL> insert into chf.t_long  select object_id,object_name from dba_objects where rownum<10;

9 rows created.

SQL> commit;

Commit complete.

SQL> desc chf.t_long
 Name                                      Null?    Type
 ----------------------------------------- -------- -----------------
 ID                                                 NUMBER
 NAME                                               LONG

测试ctas和move

SQL> create table chf.t_long_bak 
  2  as
  3  select * from chf.t_long;
select * from chf.t_long
       *
ERROR at line 3:
ORA-00997: illegal use of LONG datatype


SQL> alter table chf.t_long move tablespace users;
alter table chf.t_long move tablespace users
*
ERROR at line 1:
ORA-00997: illegal use of LONG datatype

使用copy实现LONG表跟换表空间

SQL> SET LONG 1000
SQL> select dbms_metadata.get_ddl('TABLE','T_LONG','CHF') from dual;

DBMS_METADATA.GET_DDL('TABLE','T_LONG','CHF')
--------------------------------------------------------------------------------

  CREATE TABLE "CHF"."T_LONG"
   (    "ID" NUMBER,
        "NAME" LONG
   ) SEGMENT CREATION IMMEDIATE
  PCTFREE 10 PCTUSED 40 INITRANS 1 MAXTRANS 255
 NOCOMPRESS LOGGING
  STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645
  PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1
  BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)
  TABLESPACE "TS_XIFENFEI"


SQL>   CREATE TABLE "CHF"."T_LONG_BAK"
  2     (    "ID" NUMBER,
  3          "NAME" LONG
  4     ) SEGMENT CREATION IMMEDIATE
  5    PCTFREE 10 PCTUSED 40 INITRANS 1 MAXTRANS 255
  6   NOCOMPRESS LOGGING
  7    STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645
  8    PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1
  9    BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)
 10    TABLESPACE "USERS";

Table created.

SQL> copy from chf/xifenfei@ora11g_d INSERT chf.t_long_bak using select * from chf.t_long;

Array fetch/bind size is 15. (arraysize is 15)
Will commit when done. (copycommit is 0)
Maximum long size is 80. (long is 80)
   9 rows selected from chf@ora11g_d.
   9 rows inserted into CHF.T_LONG_BAK.
   9 rows committed into CHF.T_LONG_BAK at DEFAULT HOST connection.




SQL> alter table t_long rename to t_long_old;

Table altered.

SQL> alter table t_long_bak rename to t_long;

Table altered.

SQL> select tablespace_name,table_name from dba_tables where table_name like 'T_LONG%';

TABLESPACE_NAME TABLE_NAME
--------------- ---------------
TS_XIFENFEI     T_LONG_OLD
USERS           T_LONG

SQL> DROP TABLE T_LONG_OLD PURGE;

Table dropped.
发表在 Oracle备份恢复 | 一条评论