ORACLE9i_性能调优基础九(Managing Statistics)

Managing Statistics

gather_table_stats

gather_index_stats

gather_schema_stats

gather_database_stats

gather_stale_stats

Gather Statistics

SQL> insert into ta values(3,'booboo');
1 row created.
SQL> commit;

SQL>  select num_rows, blocks from dba_tables where owner='SYS' and table_name='TA';
no rows selected
SQL> exec dbms_stats.gather_table_stats('sys','ta');
PL/SQL procedure successfully completed.

SQL>  select num_rows, blocks from dba_tables where owner='SYS' and table_name='TA';
NUM_ROWS     BLOCKS
---------- ----------
5         60

Index Statistics

1.Index level(height)

2.Number of leaf blocks and distinct keys

3.Average number of leaf blocks per key

4.Average number of data blocks per key

5.Number of index entries

6.Clustering factor

7.Data dictionary view :dba_indexes

Column Statistics

1.Number of distinct values

2.Lowest value, highest value(stored in RAW[binary] format)

3.Last analyze date and sample size

4.Data dictionary view : user_tab_col_statistics

Segment-Level Statistics

1.Logical reads:Sampled

2.Buffer busy waits:continuous

3.Db block changes :sampled

4.Physical reads:continuous

5.Physical writes:continuous

6.Physical reads direct:Continuous

7.Physical writes direct: Continuous

8.Global cache cr blocks served: continuous

9.Global cache current blocks served:Continuous

10.ITL waits:Continuous

11.Row lock waits: Continuous

setment-level statistics query:

1.v$segstat_name

2.v$segstat

3.v$segment_statistics

Using Dynamic Sampling(动态取样)

Histograms

创建直方图：　dbms_stats.gather_table_stats

dba_tab_histogram

Skewed Data

　不均匀的数据　应该在job_status上建直方图

begin
for i in 1..100
loop
insert into usr1.ta values(i,'boobooke');
end loop;
commit;
end;
/

SQL> select endpoint_number,endpoint_value from dba_histograms where owner='USR1' and table_name='TA' and column_name='ID';

ENDPOINT_NUMBER ENDPOINT_VALUE
--------------- --------------
1              0
3              1
4              2
5              3

SQL> exec dbms_stats.gather_table_stats('usr1','TA',method_opt=>'FOR COLUMNS SIZE 10 ID');

PL/SQL procedure successfully completed

SQL> select endpoint_number,endpoint_value from dba_histograms where owner='USR1' and table_name='TA' and column_name='ID';

ENDPOINT_NUMBER ENDPOINT_VALUE
--------------- --------------
0              1
1             11
2             21
3             31
4             41
5             51
6             61
7             71
8             81
9             91
10            100

11 rows selected.

Optimizer Cost Mode

System Statistics: 　包括IO ,CPU 等

1.gater_system_stats 自动收集

2.set_system_stats

3.get_system_stats

自动收集：



手动收集



Import System Statistics

Copy Statistics Between DBS