ORACLE 9i 统计表

 这几天做项目遇到了千万级表的处理，相关优化的心得特记录下来，以前日后查看。

收集统计表信息有2种方法：

1: ANALYZE TABLE employees COMPUTE STATISTICS;
 

2: exec dbms_stats.gather_table_stats(ownname => 'owner_name',tabname => 'table_name' ,estimate_percent => null ,method_opt => 'for all indexed columns' ,cascade => true);

第二种方法更好而且更加适合分区表，故如果在用了第1种方法统计表信息后反正SQL查询变慢了，可以删除相关统计信息后改用第二种方式，原因如下：

自从Oracle8.1.5引入dbms_stats包，Experts们便推荐使用dbms_stats取代analyze。 理由如下

dbms_stats可以并行分析
dbms_stats有自动分析的功能(alter table monitor )
analyze 分析统计信息的不准确some times

1,2好理解，且第2点实际上在VLDB中是最吸引人的；3以前比较模糊，看了metalink236935.1 解释，analyze在分析Partition表的时候，有时候会计算出不准确的Global statistics .

原因是，dbms_stats会实在的去分析表全局统计信息（当指定参数）；而analyze是将表分区（局部）的statistics 汇总计算成表全局statistics ,可能导致误差。

如果想分析整个用户或数据库，还可以采用工具包，可以并行分析
Dbms_utility(8i以前的工具包)
Dbms_stats(8i以后提供的工具包)
如
dbms_stats.gather_schema_stats(User,estimate_percent=>100,cascade=> TRUE);
dbms_stats.gather_table_stats(User,TableName,degree => 4,cascade => true);

这是对命令与工具包的一些总结

1、对于分区表，建议使用DBMS_STATS，而不是使用Analyze语句。
a) 可以并行进行，对多个用户，多个Table
b) 可以得到整个分区表的数据和单个分区的数据。
c) 可以在不同级别上Compute Statistics：单个分区，子分区，全表，所有分区
d) 可以倒出统计信息
e) 可以用户自动收集统计信息

2、DBMS_STATS的缺点
a) 不能Validate Structure
b) 不能收集CHAINED ROWS, 不能收集CLUSTER TABLE的信息，这两个仍旧需要使用Analyze语句。
c) DBMS_STATS 默认不对索引进行Analyze，因为默认Cascade是False，需要手工指定为True

3、对于oracle 9里面的External Table，Analyze不能使用，只能使用DBMS_STATS来收集信息。

-----------------------------------------------------------------
10G的文档是这么说的：
Do not use the COMPUTE and ESTIMATE clauses of ANALYZE to collect optimizer statistics. These clauses are supported for backward compatibility. Instead, use the DBMS_STATS package, which lets you collect statistics in parallel, collect global statistics for partitioned objects, and fine tune your statistics collection in other ways. The cost-based optimizer, which depends upon statistics, will eventually use only statistics that have been collected by DBMS_STATS

analyze的功能已经明确：
Use the ANALYZE statement (rather than DBMS_STATS) forstatistics collection not related to the cost-based optimizer:

To use the VALIDATE or LIST CHAINED ROWS clauses

To collect information on freelist blocks

在收集与CBO优化器不相关的统计信息的时候ANALYZE语句要优于DBMS_STATS包

-----------------------------------
EX:
begin
for owner in (select username from dba_users where username not in ('SYS','SYSTEM'))
loop
dbms_output.disable;
dbms_output.enable(1000000);
dbms_output.put_line('Schema: '||owner.username);
select sysdate into start_time from dual;
dbms_output.put_line('Analyze start from : '||start_time);
dbms_stats.gather_schema_stats(ownname => owner.username, estimate_percent => 20, block_sample=> true, cascade=>true);
select sysdate into end_time from dual;
dbms_output.put_line('Analyze complete at : '||end_time);
dbms_output.put_line('---------------------------');
end loop;

dbms_stats.gather_table_stats(ownname =>,
tabname =>,
partname =>,
estimate_percent =>,
block_sample =>,
method_opt =>,
degree =>,parallel degree(并行收集维度)　看CPU个数
granularity =>,
cascade =>,true is also gather columns and index’s statistics;
no_invalidate =>);

PS:　

参数说明:

ownname:要分析表的拥有者

tabname:要分析的表名.

partname:分区的名字,只对分区表或分区索引有用.

estimate_percent:采样行的百分比,取值范围[0.000001,100],null为全部分析,不采样. 常量:DBMS_STATS.AUTO_SAMPLE_SIZE是默认值,由oracle决定最佳取采样值.

block_sapmple:是否用块采样代替行采样.

method_opt:决定histograms信息是怎样被统计的.method_opt的取值如下:

for all columns:统计所有列的histograms.

for all indexed columns:统计所有indexed列的histograms.

for all hidden columns:统计你看不到列的histograms

============================================================================================================

这几天和我待在时间最久的5500万级128个字段的分区表大哥差点把我玩死了，刚开始对指定的索引字段进行查询一看执行计划怎么都用不到索引，一查原来走FTS比index scan的COST更低。但是老大为了执行时间的问题特意和我强调了这个事，不搞定它不行。

GOOGLE了一下，原来alter table这种方式收集的统计信息完全不准确，对索引列的NUM_ROWS无法在柱状图中查找到，故二话不说采用dbms_stats.gather_table_stats方式解决问题。原来不是ORACLE玩我，是我心甘情愿被它玩了2天。。。

柱状图信息查询：

select *
  from all_histograms
 where table_name = upper('table_name');

相关统计命令参考：

Analyze Statement
The ANALYZE statement can be used to gather statistics for a specific table, index or cluster. The statistics can be computed exactly, or estimated based on a specific number of rows, or a percentage of rows:

ANALYZE TABLE employees COMPUTE STATISTICS;
ANALYZE INDEX employees_pk COMPUTE STATISTICS;

ANALYZE TABLE employees ESTIMATE STATISTICS SAMPLE 100 ROWS;
ANALYZE TABLE employees ESTIMATE STATISTICS SAMPLE 15 PERCENT;
DBMS_UTILITY
The DBMS_UTILITY package can be used to gather statistics for a whole schema or database. Both methods follow the same format as the analyze statement:

EXEC DBMS_UTILITY.analyze_schema('SCOTT','COMPUTE');
EXEC DBMS_UTILITY.analyze_schema('SCOTT','ESTIMATE', estimate_rows => 100);
EXEC DBMS_UTILITY.analyze_schema('SCOTT','ESTIMATE', estimate_percent => 15);

EXEC DBMS_UTILITY.analyze_database('COMPUTE');
EXEC DBMS_UTILITY.analyze_database('ESTIMATE', estimate_rows => 100);
EXEC DBMS_UTILITY.analyze_database('ESTIMATE', estimate_percent => 15);
DBMS_STATS
The DBMS_STATS package was introduced in Oracle 8i and is Oracles preferred method of gathering object statistics. Oracle list a number of benefits to using it including parallel execution, long term storage of statistics and transfer of statistics between servers. Once again, it follows a similar format to the other methods:

EXEC DBMS_STATS.gather_database_stats;
EXEC DBMS_STATS.gather_database_stats(estimate_percent => 15);

EXEC DBMS_STATS.gather_schema_stats('SCOTT');
EXEC DBMS_STATS.gather_schema_stats('SCOTT', estimate_percent => 15);

EXEC DBMS_STATS.gather_table_stats('SCOTT', 'EMPLOYEES');
EXEC DBMS_STATS.gather_table_stats('SCOTT', 'EMPLOYEES', estimate_percent => 15);

EXEC DBMS_STATS.gather_index_stats('SCOTT', 'EMPLOYEES_PK');
EXEC DBMS_STATS.gather_index_stats('SCOTT', 'EMPLOYEES_PK', estimate_percent => 15);
This package also gives you the ability to delete statistics:

EXEC DBMS_STATS.delete_database_stats;
EXEC DBMS_STATS.delete_schema_stats('SCOTT');
EXEC DBMS_STATS.delete_table_stats('SCOTT', 'EMPLOYEES');
EXEC DBMS_STATS.delete_index_stats('SCOTT', 'EMPLOYEES_PK');

以及参考LINK：

http://rdc.taobao.com/blog/dba/html/250_how_to_gather_and_delete_column_stats.html

http://www.oracle.com.cn/viewthread.php?tid=105168&extra=&page=2

http://www.blogjava.net/envoydada/archive/2009/02/07/253698.html