[20191125]探究等待事件的本源.txt
--//当工作中遇到oracle的性能问题时,查看awr报表提供很好的解决问题途径.但是有时候很容易想当然.
--//比如以前我一看到 log file sync等待事件就很主观的认为redo 磁盘IO不行,实际上真实的情况可能有许多原因.
--//比如提交进程太多,cpu调度不过来.总之需要wait chains确定问题的本源.
--//上午看了大师Tanel Poder的dash_wait_chains.sql与ash_wait_chains.sql脚本,简单学习它的使用.
--//dash_wait_chains.sql 是基于DBA_HIST_ACTIVE_SESS_HISTORY视图.而ash_wait_chains.sql基于V$ACTIVE_SESSION_HISTORY视图.
--//二者命令执行格式差不多,仅仅学习ash_wait_chains.sql就ok了.
--//实际上根本不需要学习与记忆,看看脚本的开头就知道如何执行与使用:
-- Usage:
-- @ash_wait_chains <grouping_cols> <filters> <fromtime> <totime>
--
-- Example:
-- @ash_wait_chains username||':'||program2||event2 session_type='FOREGROUND' sysdate-1/24 sysdate
--
-- Other:
-- This script uses only the in-memory V$ACTIVE_SESSION_HISTORY, use
-- @dash_wait_chains.sql for accessiong the DBA_HIST_ACTIVE_SESS_HISTORY archive
--
-- Oracle 10g does not have the BLOCKING_INST_ID column in ASH so you'll need
-- to comment out this column in this script. This may give you somewhat
-- incorrect results in RAC environment with global blockers.
--
--------------------------------------------------------------------------------
--//我做一点点简单的修改:
COL wait_chain FOR A300 WORD_WRAP
--//修改为:
COL wait_chain FOR A200 WORD_WRAP
--//我的工作环境最大显示宽度是271列.如果设置再小,字体显示就太小,我同事许多设置才237.
$ echo $COLUMNS
271
--//因为我还需要管理10g的数据库,根据前面的说明,可以拷贝脚本命名为ash_wait_chains10g.sql.
--//注解如下,当然对于rac环境可能不正确.
-- AND PRIOR d.blocking_inst_id = d.inst_id
--//作者还重新定义字段program,event为program2,event2.这样显示更加直观,占用宽度更小一些.
--//以下是测试部分:
1.环境:
qqqqqqqqqqq> @ ver1
PORT_STRING VERSION BANNER
------------------- ---------- ----------------------------------------------------------------------------
x86_64/Linux 2.4.xx 11.2.0.4.0 Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
qqqqqqqqqqq> @ tpt/ash/ash_wait_chains program2||':'||event2 1=1 trunc(sysdate)+9/24 trunc(sysdate)+10/24
-- Display ASH Wait Chain Signatures script v0.2 BETA by Tanel Poder ( http://blog.tanelpoder.com )
%This SECONDS AAS WAIT_CHAIN
------ ---------- ---------- -------------------------------------------------------------------
33% 4362 .1 -> (XXXYYY.EXE) :ON CPU
10% 1368 0 -> (NSSn) :LNS wait on SENDREQ
10% 1323 0 -> (LGWR) :LGWR-LNS wait on channel
6% 735 0 -> (XXXYYY.EXE) :log file sync -> (LGWR) :LGWR-LNS wait on channel
5% 605 0 -> (CAPAA-PIPE) :ON CPU
5% 604 0 -> (XXXYYY.exe) :ON CPU
4% 573 0 -> (wnwp.exe) :ON CPU
3% 378 0 -> (DIAn) :ON CPU
2% 274 0 -> (sqlplus) :ON CPU
2% 246 0 -> (httpd.exe) :ON CPU
2% 230 0 -> (PSPn) :ON CPU
1% 115 0 -> (CKPT) :ON CPU
1% 111 0 -> (LMSn) :ON CPU
1% 111 0 -> (wnwp.exe) :log file sync -> (LGWR) :LGWR-LNS wait on channel
1% 89 0 -> (routine.exe) :ON CPU
1% 87 0 -> (oracle) :ON CPU
1% 86 0 -> (sqlplus) :control file sequential read
1% 83 0 -> (LGWR) :ON CPU
1% 76 0 -> (routine.exe) :log file sync -> (LGWR) :LGWR-LNS wait on channel
1% 72 0 -> (ARCn) :ON CPU
1% 69 0 -> (XXXYYY.EXE) :gc current block 2-way
1% 68 0 -> (XXXYYY.EXE) :gc cr block 2-way
0% 63 0 -> (LGWR) :log file parallel write
0% 55 0 -> (LMON) :ON CPU
0% 55 0 -> () :null event
0% 47 0 -> (XXXYYY.EXE) :log file sync
0% 46 0 -> (nnnn.exe) :log file sync -> (LGWR) :LGWR-LNS wait on channel
0% 39 0 -> (LMSn) :gcs log flush sync -> (LGWR) :LGWR-LNS wait on channel
0% 38 0 -> (XXXYYY.EXE) :direct path read
0% 38 0 -> (XXXYYY.EXE) :gc cr block busy
30 rows selected.
--//开始不理解AAS表示什么.表示如下
ROUND(COUNT(*) / ((CAST(&4 AS DATE) - CAST(&3 AS DATE)) * 86400), 1) AAS
--//相当于每秒取样几次.
--//参数3,参数4是时间范围,我建议使用类似例子的格式.你仅仅需要知道1/24表示1小时,1/1440表示1分钟就ok了.
--//比如你需要查看当天8:15到9:10的情况,时间可以写成:
trunc(sysdate)+8/24+15/1440 trunc(sysdate)+9/24+10/1440
--//另外注意一点:V$ACTIVE_SESSION_HISTORY保留信息有限制,时间太久的可能查询不到.要改用dash_wait_chains.sql脚本查询.
--//参数2可以加入一些过滤例子:session_type='FOREGROUND' 或者使用 1=1 表示全部.
--//你可以看到这台服务器的log file sync等待事件实际上由于(LGWR) :LGWR-LNS wait on channel.
--//问题在与安装配置dg时的参数:
qqqqqqqqqqq> show parameter log_archive_dest_2
NAME TYPE VALUE
------------------ ------ ----------------------------------------------------------------------------------------------------
log_archive_dest_2 string service=rzdbra lgwr sync reopen=15 max_failure=10 net_timeout=30 optional noaffirm db_unique_name=XXXXXX
--//对方配置了sync,这样在提交时要确定dg方已经写入日志文件才算提交.
--//如果改成async,相关等待LNS wait on SENDREQ,LGWR-LNS wait on channel,log file sync都会消失或者减少.
--//再来看看另外1个10g的数据库:
WWWWWWWWW> @ ver1
PORT_STRING VERSION BANNER
------------------- -------------- ----------------------------------------------------------------
x86_64/Linux 2.4.xx 10.2.0.4.0 Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bi
WWWWWWWWW> @ tpt/ash/ash_wait_chains10g program2||':'||event2 1=1 trunc(sysdate)+9/24 trunc(sysdate)+10/24
-- Display ASH Wait Chain Signatures script v0.2 BETA by Tanel Poder ( http://blog.tanelpoder.com )
%This SECONDS AAS WAIT_CHAIN
------ ---------- ---------- ------------------------------------------------------------------------
30% 3602 .1 -> (xxxzzzz.exe) :db file sequential read
19% 2244 0 -> (DBWn) :ON CPU
11% 1321 0 -> (LGWR) :log file parallel write
9% 1113 0 -> (xxxzzzz.exe) :ON CPU
7% 890 0 -> (xxxzzzz.exe) :db file scattered read
6% 784 0 -> (xxxzzzz.exe) :log file sync -> (LGWR) :log file parallel write
5% 633 0 -> (aaatobbbb.exe) :log file sync -> (LGWR) :log file parallel write
4% 490 0 -> (CKPT) :control file parallel write
1% 106 0 -> (wnwp.exe) :db file sequential read
1% 88 0 -> (JDBC Thin Client) :db file sequential read
1% 86 0 -> (aaatobbbb.exe) :db file scattered read
1% 82 0 -> (mnnn) :db file sequential read
0% 59 0 -> (xxxzzzz.exe) :read by other session
0% 55 0 -> (wnwp.exe) :ON CPU
0% 54 0 -> (JDBC Thin Client) :ON CPU
0% 45 0 -> (plsqldev.exe) :enq: TX - row lock contention
0% 44 0 -> (sqlplus) :ON CPU
0% 27 0 -> (wnwp.exe) :log file sync -> (LGWR) :log file parallel write
0% 27 0 -> (aaatobbbb.exe) :ON CPU
0% 25 0 -> (xxxzzzz.exe) :SQL*Net more data from dblink
0% 25 0 -> (mnnn) :ON CPU
0% 24 0 -> (ARCn) :log file sequential read
0% 19 0 -> (xxxzzzz.exe) :db file parallel read
0% 13 0 -> (CTWR) :ON CPU
0% 13 0 -> (aaatobbbb.exe) :db file sequential read
0% 12 0 -> (Jnnn) :db file sequential read
0% 10 0 -> (aaatobbbb.exe) :log file sync
0% 10 0 -> (SMON) :db file sequential read
0% 10 0 -> (ARCn) :ON CPU
0% 10 0 -> (CJQn) :ON CPU
30 rows selected.
--//可以发现这台系统磁盘io已经不行.如果负荷在增加就不行.顺便说一下这台机器硬件配置很差.
WWWWWWWWW> @ tpt/ash/ash_wait_chains10g event2 1=1 trunc(sysdate)+7/24 sysdate
-- Display ASH Wait Chain Signatures script v0.2 BETA by Tanel Poder ( http://blog.tanelpoder.com )
%This SECONDS AAS WAIT_CHAIN
------ ---------- ---------- -----------------------------------------------------------
32% 4910 .1 -> db file sequential read
32% 4887 .1 -> ON CPU
12% 1896 0 -> log file sync -> log file parallel write
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
11% 1761 0 -> log file parallel write
7% 1078 0 -> db file scattered read
4% 650 0 -> control file parallel write
0% 59 0 -> read by other session
0% 45 0 -> enq: TX - row lock contention
0% 34 0 -> SQL*Net more data to client
0% 34 0 -> SQL*Net more data from dblink
0% 24 0 -> log file sequential read
0% 23 0 -> db file parallel read
0% 21 0 -> log file sync
0% 7 0 -> log file switch completion
0% 6 0 -> null event
0% 5 0 -> change tracking file synchronous write
0% 4 0 -> SQL*Net more data from client
0% 4 0 -> control file sequential read
0% 3 0 -> os thread startup
0% 2 0 -> enq: CF - contention -> control file parallel write
0% 1 0 -> SQL*Net break/reset to client
0% 1 0 -> Log archive I/O
0% 1 0 -> latch: library cache
0% 1 0 -> log file sync -> ON CPU
24 rows selected.
--//注意看下划线,log file sync主要是由于log file parallel write太慢造成的,与前面的不同.这样定位问题就不会太盲目乱猜.