latch: cache buffers chains故障处理总结

一大早就接到开发商的电话,说数据库的CPU使用率为100%,应用相应迟缓。急匆匆的赶到现场发现进行了基本的检查后发现是latch: cache buffers chains 作祟,处理过程还算顺利,当时忘了记录log,这里总结下处理思路,以便下次查看。

故障分析思路

查看等待事件,判断故障起因

SQL>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;

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

SQL>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

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

SQL> 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

misses/gets的比率是多少

获自spinning的misses的百分比是多少

latch请求了多少次

latch休眠了多少次

查看热点对象和访问信息,TCH列表示对象被访问的次数

SQL> 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

SQL>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,主要

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;

查看引起latch: cache buffers chains的sql

SQL> select * from (select

    count(*),
    sql_id,
    nvl(o.object_name,ash.current_obj#)
objn,
    substr(o.object_type,0,10) otype,
     3    4    5    6       
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;

根据上面得到的sql_id信息查看sql全文

SQL>select sql_fulltext
from v$sqlarea where sql_id='&sqlid';

查看SQL的执行计划
SQL>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

SQL>SELECT  address,
hash_value FROM v$sql
     WHERE sql_id='&sql_id';
SQL>SELECT
operation, options, object_name, cost FROM v$sql_plan
     WHERE address =
'&addr' AND hash_value = 'hash_v';

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

SQL>SELECT
s.sql_hash_value,s.sql_id,s.address, l.name
  FROM V$SESSION s, V$LATCHHOLDER
l
WHERE s.sid = l.sid;

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

关于热块,可以参阅笔者的如下文章

http://czmmiao.iteye.com/blog/1474472

参考至:https://sites.google.com/site/embtdbo/wait-event-documentation/oracle-latch-cache-buffers-chains

http://space.itpub.net/354732/viewspace-697317

http://www.toadworld.com/KNOWLEDGE/KnowledgeXpertforOracle/tabid/648/TopicID/RCBCL/Default.aspx

http://blog.163.com/wghbeyond@126/blog/static/351661812010619073376/

上一篇:CentOS7下安装docker(Docker系列1)


下一篇:C++ Primer 笔记——拷贝控制