上文讲到了高水位推进的情况,当并发会话同时进行insert时,极易引起高水位争用enq: HW - contention,那么发生此类争用时,该如何诊断呢?
查看v$session_wait,应该会有如下等待事件:
SQL>select event,p1,p2,p3 from v$session_wait;
EVENT P1 P2 P3
---------------------- -------- ------- ----------
enq: HW - contention 1213661190 4 17005691
通过P3进行DBMS_UTILITY转换可以获知发生争用的文件和block
SQL> select DBMS_UTILITY.DATA_BLOCK_ADDRESS_FILE(17005691) FILE#,
2 DBMS_UTILITY.DATA_BLOCK_ADDRESS_BLOCK(17005691) BLOCK#
3 from dual;
FILE# BLOCK#
---------- ----------
4 228475
进而通过file#和block#定位对象
SQL> select owner, segment_type, segment_name
2 from dba_extents
3 where file_id = 4
4 and 228475 between block_id and block_id + blocks - 1;
OWNER SEGMENT_TYPE SEGMENT_NAME
--------------- --------------- ------------------------------
SCOTT LOBSEGMENT EMP_DATA_LOB
我们知道enqueue锁的p2,p3值与v$lock的id1,id2值相同,同样通过id2,也可以知道发生争用的文件和block
SQL> select DBMS_UTILITY.DATA_BLOCK_ADDRESS_FILE(ID2) FILE#,
2 DBMS_UTILITY.DATA_BLOCK_ADDRESS_BLOCK(ID2) BLOCK#
3 from v$lock
4 where type = 'HW';
FILE# BLOCK#
---------- ----------
4 228475
通过p1值可以知道锁类型和模式
SQL> select chr(bitand(1213661190,-16777216)/16777215)||
2 chr(bitand(1213661190,16711680)/65535) "Lock",to_char( bitand(1213661190, 65535) ) "Mode" from dual;
Lock Mode
---------- ----------
HW 6
当知道了,lob对象发生高水位争用时,该怎么办呢?metalink(740075.1)为我们提供了几种解决办法,仅供参考
查看v$session_wait,应该会有如下等待事件:
SQL>select event,p1,p2,p3 from v$session_wait;
EVENT P1 P2 P3
---------------------- -------- ------- ----------
enq: HW - contention 1213661190 4 17005691
通过P3进行DBMS_UTILITY转换可以获知发生争用的文件和block
SQL> select DBMS_UTILITY.DATA_BLOCK_ADDRESS_FILE(17005691) FILE#,
2 DBMS_UTILITY.DATA_BLOCK_ADDRESS_BLOCK(17005691) BLOCK#
3 from dual;
FILE# BLOCK#
---------- ----------
4 228475
进而通过file#和block#定位对象
SQL> select owner, segment_type, segment_name
2 from dba_extents
3 where file_id = 4
4 and 228475 between block_id and block_id + blocks - 1;
OWNER SEGMENT_TYPE SEGMENT_NAME
--------------- --------------- ------------------------------
SCOTT LOBSEGMENT EMP_DATA_LOB
我们知道enqueue锁的p2,p3值与v$lock的id1,id2值相同,同样通过id2,也可以知道发生争用的文件和block
SQL> select DBMS_UTILITY.DATA_BLOCK_ADDRESS_FILE(ID2) FILE#,
2 DBMS_UTILITY.DATA_BLOCK_ADDRESS_BLOCK(ID2) BLOCK#
3 from v$lock
4 where type = 'HW';
FILE# BLOCK#
---------- ----------
4 228475
通过p1值可以知道锁类型和模式
SQL> select chr(bitand(1213661190,-16777216)/16777215)||
2 chr(bitand(1213661190,16711680)/65535) "Lock",to_char( bitand(1213661190, 65535) ) "Mode" from dual;
Lock Mode
---------- ----------
HW 6
当知道了,lob对象发生高水位争用时,该怎么办呢?metalink(740075.1)为我们提供了几种解决办法,仅供参考
引用
1. When using Automatic Segment Space Management (ASSM)
a) As temporary workaround, manually add extra space to the LOB segment
ALTER TABLE <lob_table>
MODIFY LOB (<column_name>) (allocate extent (size <extent size>));
OR
b) It may related Bug 6376915.
Refer to Note 6376915.8 "Bug 6376915 HW enqueue contention for ASSM LOB segments"
In 10.2.0.4 or above, this fix has been included, and can be enabled by setting event 44951 to a value
between 1 and 1024. A higher value would be more beneficial in reducing contention.
EVENT="44951 TRACE NAME CONTEXT FOREVER, LEVEL < 1 - 1024 >"
OR
c) Consider partitioning the LOB in a manner that will evenly distribute concurrent DML across multiple partitions
2. When using Manual Segment Space Management(MSSM)
a) As temporary workaround, manually add extra space to the LOB segment
ALTER TABLE <lob_table>
MODIFY LOB (<column_name>) (allocate extent (size <extent size>));
OR
b) Consider partitioning the LOB in a manner that will evenly distribute concurrent DML across multiple partitions
a) As temporary workaround, manually add extra space to the LOB segment
ALTER TABLE <lob_table>
MODIFY LOB (<column_name>) (allocate extent (size <extent size>));
OR
b) It may related Bug 6376915.
Refer to Note 6376915.8 "Bug 6376915 HW enqueue contention for ASSM LOB segments"
In 10.2.0.4 or above, this fix has been included, and can be enabled by setting event 44951 to a value
between 1 and 1024. A higher value would be more beneficial in reducing contention.
EVENT="44951 TRACE NAME CONTEXT FOREVER, LEVEL < 1 - 1024 >"
OR
c) Consider partitioning the LOB in a manner that will evenly distribute concurrent DML across multiple partitions
2. When using Manual Segment Space Management(MSSM)
a) As temporary workaround, manually add extra space to the LOB segment
ALTER TABLE <lob_table>
MODIFY LOB (<column_name>) (allocate extent (size <extent size>));
OR
b) Consider partitioning the LOB in a manner that will evenly distribute concurrent DML across multiple partitions