[20250128]21C下测试CBC latch相关测试5.txt

[20250128]21C下测试CBC latch相关测试5.txt

--//元旦前别人问的问题，看vage的<oracle内核技术揭秘>，里面提到的各种情况，实际上从11.2.0.4开始全表扫描，唯一索引，rowid
--//模式以及快速全索引扫描的读读模式下(共享模式)基本不会出现cbc latch，如果出现也许不巧遇到读取cbc latch链时其它 sql 语
--//句修改或者某种操作读取其链表上其它数据块，出现cbc latch。

--//我看了我以前的测试笔记，在21c验证看看是否存在一些变化。
--//最后的结论：21c下基本在读读模式下基本不会出现cbc latch等待事件。

--//前几天测试写的脚本是顺序执行的，不具有通用性，修改一下顺便测试看看。

1.环境：
SCOTT@book01p> @ ver2
==============================
PORT_STRING                   : x86_64/Linux 2.4.xx
VERSION                       : 21.0.0.0.0
BANNER                        : Oracle Database 21c Enterprise Edition Release 21.0.0.0.0 - Production
BANNER_FULL                   : Oracle Database 21c Enterprise Edition Release 21.0.0.0.0 - Production
Version 21.3.0.0.0
BANNER_LEGACY                 : Oracle Database 21c Enterprise Edition Release 21.0.0.0.0 - Production
CON_ID                        : 0
PL/SQL procedure successfully completed.

2.建立测试环境：
SCOTT@book01p> create table job_times (sid number, time_ela number,method varchar2(20));
Table created.

SCOTT@book01p> create table t as select rownum id ,cast('test' as varchar2(10)) name from dual ;
Table created.

SCOTT@book01p> alter table t modify ( id  not null );
Table altered.

SCOTT@book01p> create index i_t_id on t(id);
Index created.
--//create unique index i_t_id on t(id);
--//分析表略。

SCOTT@book01p> select rowid from t;
ROWID
------------------
AAAYbIAAMAAAAEDAAA

SCOTT@book01p> @ rowid AAAYbIAAMAAAAEDAAA

DATA_OBJECT_ID       FILE      BLOCK        ROW ROWID_DBA            DBA                  TEXT
-------------- ---------- ---------- ---------- -------------------- -------------------- --------------------------------------------------
        100040         12        259          0  0x3000103           12,259               alter system dump datafile 12 block 259 ;

$ nl loop.txt
     1  declare
     2  v_id number;
     3  v_d date;
     4  begin
     5      for i in 1 .. &&1 loop
     6          --select /*+ full(t) &&3 */ count (name) into v_id from t ;
     7          --select /*+ ROWID(t) &&3 */ count (name) into v_id from t where rowid='AAAYbIAAMAAAAEDAAA';
     8          --select /*+ index(t) &&3 */ count (name) into v_id from t ;
     9          --select /*+ index(t) &&3 */ count (name) into v_id from t where id=1;
    10          --select /*+ index_ffs(t) &&3 */ count (*) into v_id from t ;
    11          --select /*+ index_ffs(t) &&3 */ count (name) into v_id from t ;
    12          --select /*+ ROWID(t) &&3 */ count (name) into v_id from t where rowid between 'AAAYbIAAMAAAAEDAAA' and 'AAAYbIAAMAAAAEDAAB';
    13      end loop;
    14  end ;
    15  /

$ cp loop.txt loop.org
--//做1个备份。

$ cat record.txt
set verify off
variable v_method varchar2(20)
exec :v_method := '&&2';

define t=&&1;

@@ loop.txt 5 &&2

insert into job_times values ( sys_context ('userenv', 'sid') ,dbms_utility.get_time ,'&&2') ;
commit ;

@@ loop.txt &&t &&2

update job_times set time_ela = dbms_utility.get_time - time_ela where sid=sys_context ('userenv', 'sid') and method= :v_method;
commit;
quit

--//简单说明：
--//每次测试前我都单独执行要测试sql语句5次。
--//注：唯一索引与非唯一索引分开测试。在提示中加入 &&3，主要避免出现cursor: pin S等待事件。
--//以前测试时分别注解上面的部分，这次通过编写脚本1次完成，事后分析。

$ cat cbc2.sh

#!/bin/bash

v_conn=scott/book@book01p
v_test="p1full p1rowid p1index_fs p1not_uniindex p1indexffs p1indexffs_name p1rowidbetween"
v_line=6

sqlplus -s -l ${v_conn} <<EOF > /dev/null
drop index i_t_id;
create index i_t_id on t(id);
quit
EOF
sleep 1

for ii in $v_test
do
    sed  -e "${v_line}s/ --select/ select/" loop.org >| loop.txt
    sleep 3
    zzdate
    seq 50 | xargs -I{} -P 50 sqlplus -s -l ${v_conn} @record.txt 1e5 ${ii}=50 {} >/dev/null
    zzdate
    ((v_line++))
    sleep 3
done

sqlplus -s -l ${v_conn} <<EOF > /dev/null
drop index i_t_id;
create unique index i_t_id on t(id);
quit
EOF
sleep 1

v_conn=scott/book@book01p
v_test="p2full p2rowid p2index_fs p2uniindex p2indexffs p2indexffs_name p2rowidbetween"
v_line=6
for ii in $v_test
do
    sed  -e "${v_line}s/ --select/ select/" loop.org >| loop.txt
    sleep 3
    zzdate
    seq 50 | xargs -I{} -P 50 sqlplus -s -l ${v_conn} @record.txt 1e5 ${ii}=50 {} >/dev/null
    zzdate
    ((v_line++))
    sleep 3
done

3.测试：

--//delete from job_times ;
--//commit ;

$ ./cbc2.sh
trunc(sysdate)+15/24+51/1440+41/86400 1738309901.866128439     --//p1full=50
trunc(sysdate)+15/24+52/1440+15/86400 1738309935.589019338     --//p1full=50
trunc(sysdate)+15/24+52/1440+21/86400 1738309941.602397560     --//p1rowid=50
trunc(sysdate)+15/24+52/1440+44/86400 1738309964.033334966     --//p1rowid=50
trunc(sysdate)+15/24+52/1440+50/86400 1738309970.056889505     --//p1index_fs=50
trunc(sysdate)+15/24+53/1440+17/86400 1738309997.670372914     --//p1index_fs=50
trunc(sysdate)+15/24+53/1440+23/86400 1738310003.698719931     --//p1not_uniindex=50
trunc(sysdate)+15/24+53/1440+51/86400 1738310031.984141857     --//p1not_uniindex=50
trunc(sysdate)+15/24+53/1440+58/86400 1738310038.006265731     --//p1indexffs=50
trunc(sysdate)+15/24+54/1440+28/86400 1738310068.625749964     --//p1indexffs=50
trunc(sysdate)+15/24+54/1440+34/86400 1738310074.649780188     --//p1indexffs_name=50
trunc(sysdate)+15/24+55/1440+02/86400 1738310102.426901359     --//p1indexffs_name=50
trunc(sysdate)+15/24+55/1440+08/86400 1738310108.459568675     --//p1rowidbetween=50
trunc(sysdate)+15/24+56/1440+01/86400 1738310161.221741315     --//p1rowidbetween=50
trunc(sysdate)+15/24+56/1440+08/86400 1738310168.904871134     --//p2full=50
trunc(sysdate)+15/24+56/1440+42/86400 1738310202.628110965     --//p2full=50
trunc(sysdate)+15/24+56/1440+48/86400 1738310208.650691446     --//p2rowid=50
trunc(sysdate)+15/24+57/1440+10/86400 1738310230.620899334     --//p2rowid=50
trunc(sysdate)+15/24+57/1440+16/86400 1738310236.645815987     --//p2index_fs=50
trunc(sysdate)+15/24+57/1440+44/86400 1738310264.315145405     --//p2index_fs=50
trunc(sysdate)+15/24+57/1440+50/86400 1738310270.344066349     --//p2uniindex=50
trunc(sysdate)+15/24+58/1440+15/86400 1738310295.138286658     --//p2uniindex=50
trunc(sysdate)+15/24+58/1440+21/86400 1738310301.170380645     --//p2indexffs=50
trunc(sysdate)+15/24+58/1440+51/86400 1738310331.735310144     --//p2indexffs=50
trunc(sysdate)+15/24+58/1440+57/86400 1738310337.764016288     --//p2indexffs_name=50
trunc(sysdate)+15/24+59/1440+25/86400 1738310365.409580747     --//p2indexffs_name=50
trunc(sysdate)+15/24+59/1440+31/86400 1738310371.434929589     --//p2rowidbetween=50
trunc(sysdate)+16/24+00/1440+23/86400 1738310423.884724199     --//p2rowidbetween=50

SYS@book> @ ashtop event 1=1 trunc(sysdate)+15/24+51/1440+41/86400 trunc(sysdate)+16/24+00/1440+23/86400
    Total                                                                                    Distinct Distinct    Distinct
  Seconds     AAS %This   EVENT                    FIRST_SEEN          LAST_SEEN           Execs Seen  Tstamps Execs Seen1
--------- ------- ------- ------------------------ ------------------- ------------------- ---------- -------- -----------
    18621    35.7  100% |                          2025-01-31 15:51:41 2025-01-31 16:00:22      11454      447       11899
       26      .0    0% | log file sync            2025-01-31 15:51:43 2025-01-31 16:00:19          1       25          25
        8      .0    0% | log file parallel write  2025-01-31 15:52:43 2025-01-31 15:58:58          1        8           8
        2      .0    0% | latch: shared pool       2025-01-31 15:57:41 2025-01-31 15:59:01          1        2           2
        1      .0    0% | oracle thread bootstrap  2025-01-31 16:00:02 2025-01-31 16:00:02          1        1           1
        1      .0    0% | os thread creation       2025-01-31 16:00:02 2025-01-31 16:00:02          1        1           1
6 rows selected.
--//没有latch: cache buffers chains等待事件。

COTT@book01p> Select method,count(*),round(avg(TIME_ELA),2),sum(TIME_ELA) from job_times  group by method order by 3 ;
METHOD                 COUNT(*) ROUND(AVG(TIME_ELA),2) SUM(TIME_ELA)
-------------------- ---------- ---------------------- -------------
p2rowid=50                   50                1720.24         86012
p1rowid=50                   50                1765.52         88276
p2uniindex=50                50                1984.92         99246
p2index_fs=50                50                2280.64        114032
p1index_fs=50                50                2280.86        114043
p2indexffs_name=50           50                2285.90        114295
p1indexffs_name=50           50                2295.26        114763
p1not_uniindex=50            50                2337.68        116884
p2indexffs=50                50                2565.56        128278
p1indexffs=50                50                2571.90        128595
p1full=50                    50                2856.22        142811
p2full=50                    50                2862.62        143131
p2rowidbetween=50            50                4689.38        234469
p1rowidbetween=50            50                4768.18        238409
14 rows selected.
--//2次测试结果基本一致。pXindex_fs，pxindexffs_name都是index full scan，执行计划一样，交错出现很正常。

--//你可以反复测试，注意几点测试前删除job_times历史记录.