其他分享
首页 > 其他分享> > latch: cache buffers chains故障处理总结

latch: cache buffers chains故障处理总结

作者:互联网

latch: cache buffers chains故障处理总结
oracle 版本:10.2.0.5.0
一、故障分析思路 
1、查看等待事件,判断故障起因
select *
from (select sid, event, p1, p2, p3, p1text, WAIT_TIME, SECONDS_IN_WAIT from v$session_wait where wait_class# <> 6 order by wait_time desc)
where rownum <= 10;

2、确认为latch: cache buffers chains引起的故障后,查看latch的命中率

SELECT name, gets, misses, sleeps, immediate_gets, immediate_misses
FROM v$latch
WHERE name = 'cache buffers chains';

各列名称意义如下
NAME:latch名称
IMMEDIATE_GETS:以Immediate模式latch请求数
IMMEDIATE_MISSES:请求失败数
GETS:以Willing to wait请求模式latch的请求数
MISSES:初次尝试请求不成功次数
SPIN_GETS:第一次尝试失败,但在以后的轮次中成功
SLEEP[x]:成功获取前sleeping次数
WAIT_TIME:花费在等待latch的时间
注:这里需要注意MISSES/GETS如果在达10%左右,则说明有比较严重的latch争用,也可以通过查询v$latch_children视图查看其他latch信息 ,语句如下

SELECT *
FROM (SELECT addr,
             child#,
             gets,
             misses,
             sleeps,
             immediate_gets   igets,
             immediate_misses imiss,
             spin_gets        sgets
FROM v$latch_children
WHERE NAME = 'cache buffers chains'
ORDER BY sleeps DESC)
WHERE ROWNUM < 11;

关于latch的统计信息,主要关注以下几部分:

A> misses/gets的比率是多少
B> 获自spinning的misses的百分比是多少
C> latch请求了多少次
D> latch休眠了多少次

3、查看热点对象和访问信息,TCH列表示对象被访问的次数
SELECT *
FROM (SELECT addr, ts#, file#, dbarfil, dbablk, tch
FROM x$bh
ORDER BY tch DESC)
WHERE ROWNUM < 11;
通过对象的文件号和块号查看具体对象信息
select owner, segment_name, partition_name, tablespace_name
from dba_extents
where relative_fno = &v_dba_rfile
and &v_dba_block between block_id and block_id + blocks - 1;

也可以通过如下sql查找热点块
SELECT *
FROM (SELECT O.OWNER, O.OBJECT_NAME, O.OBJECT_TYPE, SUM(TCH) TOUCHTIME
FROM X$BH B, DBA_OBJECTS O
WHERE B.OBJ = O.DATA_OBJECT_ID
AND B.TS# > 0
GROUP BY O.OWNER, O.OBJECT_NAME, O.OBJECT_TYPE
ORDER BY SUM(TCH) DESC)
WHERE ROWNUM <= 10;

4、查看引起latch: cache buffers chains的sql
select *
from (select count(*),
sql_id,
nvl(o.object_name, ash.current_obj#) objn,
substr(o.object_type, 0, 10) otype,
CURRENT_FILE# fn,
CURRENT_BLOCK# blockn
from v$active_session_history ash, all_objects o
where event LIKE 'latch: cache buffers chains'
and o.object_id(+) = ash.CURRENT_OBJ#
group by sql_id,
current_obj#,
current_file#,
current_block#,
o.object_name,
o.object_type
order by count(*) desc)
where rownum <= 10;

5、根据上面得到的sql_id信息查看sql全文
select sql_fulltext from v$sqlarea where sql_id = ’&sqlid’;

查看SQL的执行计划
SELECT * FROM table(DBMS_XPLAN.DISPLAY_CURSOR((’&sql_id’, 0));

在认为sql执行计划不准确的情况也可以通过sql_id查看sql的address和hash_value查看sql的实际执行计划
SELECT address, hash_value FROM v$sql WHERE sql_id = ’&sql_id’;
SELECT operation, options, object_name, cost
FROM v$sql_plan
WHERE address = ’&addr’
AND hash_value = ’hash_v’;

6、当某个会话长时间持有latch时,可以通过联合v$latchholder和v$session视图查看sql信息
SELECT s.sql_hash_value, s.sql_id, s.address, l.name
FROM V$SESSION s, V$LATCHHOLDER l
WHERE s.sid = l.sid;

二、故障处理思路

1、根据sql执行计划判断该执行计划是否正确,sql执行过长往往意味着过长时间的持有latch。


2、优化nested loop join,如果有可能使用hash join代替nested loop join。也可以利用对热块索引进行hash分区,或者使用hash簇的方式减缓热块现象。


3、调整表的pctfree值,将数据尽可能的分布到多个块中


4、调整应用


 

标签:hash,cache,sql,latch,chains,id,SELECT
来源: https://blog.51cto.com/u_12592884/2693881