Delete删除表数据时对性能的影响分析

上一篇文章中提到（http://www.linuxidc.com/Linux/2014-09/107345.htm），当我们用delete删除数据时，因为无法降低高水位，对表做全表扫描时会带来数据库性能问题，而truncate可以大大改善这一状况，下面来看测试 SQL> conn zlm/zlmConnected. --创建表之前先记录下剩余表空间大小SQL> select * from dba_free_space where tablespace_name="ZLM";
TABLESPACE_NAME FILE_ID BLOCK_ID BYTES BLOCKS RELATIVE_FNO------------------------------ ---------- ---------- ---------- ---------- ------------ZLM 6 128 51380224 6272 6 --创建一个100W行的大表SQL> create table test1（int number）; Table created. SQL> analyze table test1 compute statistics; Table analyzed. SQL> select table_name,num_rows,blocks,empty_blocks,avg_row_len,last_analyzed from dba_tables where table_name like "TEST%"; TABLE_NAME NUM_ROWS BLOCKS EMPTY_BLOCKS AVG_ROW_LEN LAST_ANALYZED------------------------------ ---------- ---------- ------------ ----------- ------------------TEST1 0 0 0 0 27-SEP-14 注意，创建完表结构而未插入数据时，表的高水位是0，并没有为该表分配任何数据块 SQL> declare
2 i number := 0; 3 begin 4 for i in 1 .. 1000000 loop 5 insert into test1 values （i）; 6 end loop; 7 commit; 8 end; 9 / PL/SQL procedure successfully completed. SQL> select count（*） from test1;   COUNT（*）---------- 1000000 --查看高水位状况SQL> select header_file,header_block,bytes,blocks,extents from dba_segments where segment_name like "TEST%"; HEADER_FILE HEADER_BLOCK BYTES BLOCKS EXTENTS----------- ------------ ---------- ---------- ---------- 6 130   13631488 1664   28 SQL> select block_id,extent_id,bytes,blocks from dba_extents where segment_name like "TEST%";   BLOCK_ID EXTENT_ID BYTES BLOCKS---------- ---------- ---------- ---------- 128 0 65536 8 136 1 65536 8 144 2 65536 8 152 3 65536 8 160 4 65536 8 168 5 65536 8 176 6 65536 8 184 7 65536 8 192 8 65536 8 200 9 65536 8 208 10 65536 8 216 11 65536 8 224 12 65536 8 232 13 65536 8 240 14 65536 8 248 15 65536 8 256 16 1048576 128 384 17 1048576 128 512 18 1048576 128 640 19 1048576 128 768 20 1048576 128 896 21 1048576 128 1024 22 1048576 128 1152 23 1048576 128 1280 24 1048576 128 1408 25 1048576 128 1536 26 1048576 128 1664 27 1048576 128 SQL> select * from dba_free_space where tablespace_name="ZLM";
TABLESPACE_NAME FILE_ID BLOCK_ID BYTES BLOCKS RELATIVE_FNO------------------------------ ---------- ---------- ---------- ---------- ------------ZLM 6 1792 37748736 4608 6 SQL> set autot traceSQL> select count（*） from test1;  Execution Plan----------------------------------------------------------Plan hash value: 3896847026 --------------------------------------------------------------------| Id | Operation | Name | Rows | Cost （%CPU）| Time |--------------------------------------------------------------------| 0 | SELECT STATEMENT | | 1 | 2 （0）| 00:00:01 || 1 | SORT AGGREGATE | | 1 | | || 2 | TABLE ACCESS FULL| TEST1 | 1 | 2 （0）| 00:00:01 |--------------------------------------------------------------------  Statistics---------------------------------------------------------- 0 recursive calls 1 db block gets 1599 consistent gets 0 physical reads 0 redo size 526 bytes sent via SQL*Net to client 519 bytes received via SQL*Net from client 2 SQL*Net roundtrips to/from client 0 sorts （memory） 0 sorts （disk） 1 rows processed SQL> delete from test1;
1000000 rows deleted.  Execution Plan----------------------------------------------------------Plan hash value: 2642947686 --------------------------------------------------------------------| Id | Operation | Name | Rows | Cost （%CPU）| Time |--------------------------------------------------------------------| 0 | DELETE STATEMENT | | 1 | 2 （0）| 00:00:01 || 1 | DELETE | TEST1 | | | || 2 | TABLE ACCESS FULL| TEST1 | 1 | 2 （0）| 00:00:01 |--------------------------------------------------------------------  Statistics---------------------------------------------------------- 814 recursive calls 1038983 db block gets 1953 consistent gets 8 physical reads 245334988 redo size 847 bytes sent via SQL*Net to client 769 bytes received via SQL*Net from client 3 SQL*Net roundtrips to/from client 1 sorts （memory） 0 sorts （disk） 1000000 rows processed 用delete删除100W行数据的大表TEST1时，产生了大量的redo（2亿多），另外还有很多一致性读（1953），读取了100多W个数据块，814次递归调用，可以看到，用delete删除表记录，对数据库的性能消耗是很大的，尤其是当delete大量行的时候
SQL> set autot offSQL> select count（*） from test1;   COUNT（*）---------- 0 SQL> select header_file,header_block,bytes,blocks,extents from dba_segments where segment_name like "TEST%"; HEADER_FILE HEADER_BLOCK BYTES BLOCKS EXTENTS----------- ------------ ---------- ---------- ---------- 6 130 13631488 1664   28更多详情见请继续阅读下一页的精彩内容： http://www.linuxidc.com/Linux/2014-09/107346p2.htm

用Delete与Truncate清除表数据对高水位的影响ORA-21561、ORA-15055、ORA-25253 导致DG备库无法应用归档相关资讯表数据性能表数据本文评论查看全部评论（0）