测试存储过程性能

create or replace view stats as 
select 'STAT...' || a.name name, b.value       
from v$statname a, v$mystat b      
where a.statistic# = b.statistic#     
union all     
select 'LATCH.' || name,  gets       
from v$latch     
union all     
select 'STAT...Elapsed Time', hsecs from v$timer;
创建查询统计信息视图
create global temporary table
 RUN_STATS (   
  runid VARCHAR2(15),  
  name  VARCHAR2(80),  
  value INTEGER ) 
 on commit delete rows;
1)创建全局临时表
CREATE OR REPLACE PACKAGE runstats_pkg AS
 PROCEDURE print_cputime(p_time_type VARCHAR2 DEFAULT 'start');
 PROCEDURE rs_start;
 PROCEDURE rs_middle;
 PROCEDURE rs_stop(p_difference_threshold IN NUMBER DEFAULT 0);
END; /
2)创建性能统计包
CREATE OR REPLACE PACKAGE BODY runstats_pkg AS
  g_start  INTEGER;
  g_run1   INTEGER;
  g_run2   INTEGER;
  g_startt INTEGER;
  g_runt1  INTEGER;
  g_runt2  INTEGER;

  --获取包处理过程中的cpu时间  
  PROCEDURE print_cputime(p_time_type VARCHAR2) IS
  BEGIN
    CASE lower(p_time_type)
      WHEN 'start' THEN
        dbms_output.put_line(g_start);
      WHEN '1' THEN
        dbms_output.put_line(g_run1);
      WHEN '2' THEN
        dbms_output.put_line(g_run2);
      ELSE
        dbms_output.put_line('parameter is wrong');
    END CASE;
  END;

  --设置比较性能开始比较  
  PROCEDURE rs_start IS
  BEGIN
    DELETE FROM run_stats;
    INSERT INTO run_stats
      SELECT 'before', stats.* FROM stats;
    g_start  := dbms_utility.get_cpu_time;
    g_startt := dbms_utility.get_time;
  END;

  --设置比较性能第二个过程 
  PROCEDURE rs_middle IS
  BEGIN
    g_run1  := (dbms_utility.get_cpu_time - g_start);
    g_runt1 := (dbms_utility.get_time - g_startt);
    INSERT INTO run_stats
      SELECT 'after 1', stats.* FROM stats;
    g_start  := dbms_utility.get_cpu_time;
    g_startt := dbms_utility.get_time;
  END;

  --比较结束,并打印比较信息 
  PROCEDURE rs_stop(p_difference_threshold IN NUMBER DEFAULT 0) IS
    v_rownum INTEGER := 0;
  BEGIN
    g_run2  := (dbms_utility.get_cpu_time - g_start);
    g_runt2 := (dbms_utility.get_time - g_startt);
    dbms_output.put_line('Run1 ran in ' || g_run1 || ' cpu hsecs' ||
                         ' , elapsed time:' || g_runt1);
    dbms_output.put_line('Run2 ran in ' || g_run2 || ' cpu hsecs' ||
                         ' , elapsed time:' || g_runt2);
    IF (g_run2 <> 0) THEN
      dbms_output.put_line('run 1 ran in ' ||
                           round(g_run1 / g_run2 * 100, 2) ||
                           '% of the time');
    END IF;
    dbms_output.put_line(chr(9));
    INSERT INTO run_stats
      SELECT 'after 2', stats.* FROM stats;
    dbms_output.put_line(rpad('linenumber', 20) || rpad('Name', 50) ||
                         lpad('Run1', 12) || lpad('Run2', 12) ||
                         lpad('Diff', 12));
    FOR x IN (SELECT rpad(a.name, 50) ||
                     to_char(b.value - a.value, '999,999,999') ||
                     to_char(c.value - b.value, '999,999,999') ||
                     to_char(((c.value - b.value) - (b.value - a.value)),
                             '999,999,999') data
                FROM run_stats a, run_stats b, run_stats c
               WHERE a.name = b.name
                 AND b.name = c.name
                 AND a.runid = 'before'
                 AND b.runid = 'after 1'
                 AND c.runid = 'after 2'
                 AND abs((c.value - b.value) - (b.value - a.value)) >
                     p_difference_threshold
               ORDER BY abs((c.value - b.value) - (b.value - a.value))) LOOP
      dbms_output.put_line(rpad(to_char(v_rownum), 20) || x.data);
      v_rownum := v_rownum + 1;
    END LOOP;
    v_rownum := 0;
    dbms_output.put_line(chr(9));
    dbms_output.put_line('Run1 latches total versus runs -- difference and pct');
    dbms_output.put_line(rpad('linenumber', 20) || lpad('Run1', 12) ||
                         lpad('Run2', 12) || lpad('Diff', 12) ||
                         lpad('Pct', 10));
    FOR x IN (SELECT to_char(run1, '999,999,999') ||
                     to_char(run2, '999,999,999') ||
                     to_char(diff, '999,999,999') ||
                     to_char(round(run1 /
                                   decode(run2, 0, to_number(0), run2) * 100,
                                   2),
                             '99,999.99') || '%' data
                FROM (SELECT SUM(b.value - a.value) run1,
                             SUM(c.value - b.value) run2,
                             SUM((c.value - b.value) - (b.value - a.value)) diff
                        FROM run_stats a, run_stats b, run_stats c
                       WHERE a.name = b.name
                         AND b.name = c.name
                         AND a.runid = 'before'
                         AND b.runid = 'after 1'
                         AND c.runid = 'after 2'
                         AND a.name LIKE 'LATCH%')) LOOP
      dbms_output.put_line(rpad(to_char(v_rownum), 20) || x.data);
      v_rownum := v_rownum + 1;
    END LOOP;
  END;

END;
/
3)创建性能统计包体
CREATE OR REPLACE PROCEDURE proc_fortest1(i_x INTEGER DEFAULT 10,
                                          i_y INTEGER DEFAULT 10,
                                          i_z INTEGER DEFAULT 10) IS
  PRAGMA AUTONOMOUS_TRANSACTION;
  l_i         INTEGER := i_x;
  l_j         INTEGER := i_y;
  l_k         INTEGER := i_z;
  l_sql       VARCHAR2(1024) := 'insert into fort(i,j,k)values(:x,:y,:z)';
  l_starttime DATE := SYSDATE;
BEGIN
  FOR i IN 1 .. l_i LOOP
    FOR j IN 1 .. l_j LOOP
      FOR k IN 1 .. l_k LOOP
        EXECUTE IMMEDIATE l_sql
          USING i, j, k;
      END LOOP;
    END LOOP;
  END LOOP;
  COMMIT;
  dbms_output.put_line('proc_fortest1 run times(s):' ||
                       (SYSDATE - l_starttime) * 24 * 60 * 60);
END proc_fortest1;
/

CREATE OR REPLACE PROCEDURE proc_fortest2(i_x INTEGER DEFAULT 10,
                                          i_y INTEGER DEFAULT 10,
                                          i_z INTEGER DEFAULT 10) IS
  PRAGMA AUTONOMOUS_TRANSACTION;
  l_i         INTEGER := i_x;
  l_j         INTEGER := i_y;
  l_k         INTEGER := i_z;
  l_starttime DATE := SYSDATE;
  l_sql       VARCHAR2(1024) := 'INSERT INTO fort   (i, j, k)   SELECT i.rn, j.rn, k.rn     FROM (SELECT rownum rn FROM dual CONNECT BY rownum <= :l_i) i,       (SELECT rownum rn FROM dual CONNECT BY rownum <= :l_j) j,       (SELECT rownum rn FROM dual CONNECT BY rownum <= :l_k) k';
BEGIN
  EXECUTE IMMEDIATE l_sql
    USING l_i, l_j, l_k;
  COMMIT;
  dbms_output.put_line('proc_fortest2 run times(s):' ||
                       (SYSDATE - l_starttime) * 24 * 60 * 60);
END proc_fortest2;
/

CREATE OR REPLACE PROCEDURE proc_fortest(i_limite_value INTEGER DEFAULT 0,
                                         i_x            INTEGER DEFAULT 10,
                                         i_y            INTEGER DEFAULT 10,
                                         i_z            INTEGER DEFAULT 10) IS
BEGIN
  runstats_pkg.rs_start();
  proc_fortest1(i_x, i_y, i_z);
  runstats_pkg.rs_middle();
  proc_fortest2(i_x, i_y, i_z);
  runstats_pkg.rs_stop(i_limite_value);
END proc_fortest;
/
测试用例--存储过程

5、输出结果:

SQL>set serveroutput on
SQL>set linesize 300
SQL>exec proc_fortest(0,10,20,30);

proc_fortest1 run times(s):0 proc_fortest2 run times(s):0 
Run1 ran in 22 cpu hsecs , elapsed time:28
Run2 ran in 2 cpu hsecs , elapsed time:3 run 1 ran in 1100% of the time

linenumber  Name   
 Run1      Run2 Diff 0  LATCH.library cache    pin allocation                          
  1            0           -1 1                  STAT...change write time                                     
  1            0        -1 2                  STAT...physical read total multi block requests              
  0            1         1 3                  STAT...messages sent                                         
  3            2        -1 4                  LATCH.OS process allocation                                  
  1            0        -1 5                  LATCH.SGA IO buffer pool latch                               
  0            1         1 6                  STAT...heap block compress                                   
  5            6         1 7                  LATCH.flashback mapping                                      
  0            1         1 8                  LATCH.KMG MMAN ready and startup request latch               
  1            0        -1 9                  LATCH.lgwr LWN SCN                                           
  3            1        -2 10                 LATCH.mostly latch-free SCN                                  
  3            1        -2 11                 LATCH.multiblock read objects                                
  8            6        -2 12                 STAT...consistent gets - examination 
  15          17        2 13                  STAT...cleanout - number of ktugct calls  
  15          17        2 14                  STAT...active txn count during cleanout
  15          17        2 15                  STAT...deferred (CURRENT) block cleanout applicati
  2           4         2 16                  STAT...commit cleanouts successfully completed 
  12          14        2 17                  STAT...commit cleanouts
  12          14        2 18                  LATCH.Consistent RBA 
  3           1        -2 19                  LATCH.undo global data 
  21          18       -3 20                  STAT...calls to kcmgcs  
  16          19        3 21                  STAT...workarea executions - optimal 
  6           9         3 22                  LATCH.library cache lock 
  4           0        -4 23                  STAT...consistent changes
  17          21        4 24                  STAT...user I/O wait time
  8           3        -5 25                  STAT...physical reads cache prefetch 
  0           5         5 26                  STAT...physical reads prefetch warmup
  0           5         5 27                  STAT...sorts (memory)
  1           6         5 28                  LATCH.redo writing 
  9           3        -6 29                  LATCH.redo allocation 
  9           3        -6 30                  LATCH.simulator lru latch 
  64          56       -8 31                  LATCH.messages 
  15          6        -9 32                  STAT...calls to get snapshot scn: kcmgss 
  2          16        14 33                  STAT...consistent gets from cache  
  20          35       15 34                  STAT...consistent gets
  20          35       15 35                  STAT...recursive cpu usage                                  
  19           3      -16 36                  STAT...CPU used by this session                             
  24           3      -21 37                  LATCH.simulator hash latch                                  
  79          56      -23 38                  STAT...Elapsed Time
  30           3      -27 39                  STAT...sorts (rows)
  2,341    2,394       53 40                  LATCH.cache buffers lru chain                               
  81          27         -54 41                  STAT...free buffer requested                                
  71          12         -59 42                  STAT...physical reads cache                                 
  71          11         -60 43                  LATCH.flashback allocation                                  
  77          17         -60 44                  STAT...physical reads                                       
  71          11         -60 45                  STAT...redo ordering marks                                 
  67           2         -65 46                  STAT...physical read IO requests                            
  71           6         -65 47                  STAT...physical reads for flashback new                     
  67           2         -65 48                  STAT...calls to kcmgas                                      
  68           3         -65 49                  STAT...physical read total IO requests                      
  71           6         -65 50                  LATCH.object queue header operation                        
  147          28        -119 51                  STAT...redo entries                                      
  6,046          86      -5,960 52                  STAT...recursive calls                                   
  6,003           4      -5,999 53                  STAT...execute count                                     
  6,001           2      -5,999 54                  STAT...session logical reads                             
  6,299         178      -6,121 55                  STAT...db block gets                                     
  6,279         143      -6,136 56                  STAT...db block gets from cache                          
  6,279         143      -6,136 57                  STAT...db block changes                                 
  12,131         147     -11,984 58                  LATCH.shared pool                                       
  12,002           5     -11,997 59                  LATCH.library cache pin                                 
  24,012           7     -24,005 60                  LATCH.library cache                                     
  24,017           8     -24,009 61                  LATCH.cache buffers chains                              
  30,998         588     -30,410 62                  STAT...physical read total bytes                       
  581,632      90,112    -491,520 63                  STAT...physical read bytes                             
  581,632      90,112    -491,520 64                  STAT...undo change vector size                         
  530,232      17,572    -512,660 65                  STAT...redo size                                     
  1,617,712     122,516  -1,495,196

  Run1 latches total versus runs -- difference and pct linenumber                 
    Run1        Run2        Diff       Pct 0                        
 91,584         862      -90,722      10,624.59%
测试结果
原文地址:https://www.cnblogs.com/HondaHsu/p/4607130.html