delete大批量数据引起空间爆满处理

前一天晚上做大表删除操作测试，分段删除，没删除100000条commit一次，由于存储过程打印了执行时间中途断网，无法看到执行时间
于是直接kill了任务
结果早上登录测试数据库发现空间爆满
然后一时糊涂去弄了表的shrink跟dbms_stats.gather_table_stats想着收缩表释放空间
弄到一半突然想起，表的收缩最多是释放表空间的空闲空间，并非系统空间
后来想想，大量删除反复操作过程引发了undo的扩展，后来检查表空间后确实如此
最后在线更换undo tablespace后删除原tablespace后恢复

附上存储过程

 create or replace procedure delete_table as
 i number(15);
 l_start number default dbms_utility.get_time;
 begin
    dbms_output.put_line('begin time:'||to_char(SYSTIMESTAMP,'HH24:MI:SS:FF2'));
    i:=0;
    for x in (select log_id from user_log where user_log_id<46380000)
    loop
    delete from user_log where user_log_id = x.user_log_id;
    i:=i+1;
    if (i>10000) then
        begin
            --dbms_output.put_line('delete ok.'||i);
            commit;
            i:=0;
        end;
    end if;
    end loop;
    commit;
    dbms_output.put_line('ok.end time:'||to_char(SYSTIMESTAMP,'HH24:MI:SS:FF2'));
    exception when others then
    rollback;
end;
 /

delete_table

另外调试时记得 set serveroutput on

遇到的报错提示

[oracle@centos5 osa]$ dba

SQL*Plus: Release 10.2.0.4.0 - Production on Fri Jun 21 09:23:18 2013

Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.

ERROR:
ORA-09817: Write to audit file failed.
Linux-x86_64 Error: 28: No space left on device
ORA-01075: you are currently logged on

Enter user-name:
ERROR:
ORA-01017: invalid username/password; logon denied

Enter user-name:
ERROR:
ORA-01017: invalid username/password; logon denied

SP2-0157: unable to CONNECT to ORACLE after 3 attempts, exiting SQL*Plus

检查磁盘空间，爆满

[oracle@centos5 osa]$ df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda3              93G   89G     0 100% /
/dev/sda1              99M   12M   82M  13% /boot
tmpfs                1006M     0 1006M   0% /dev/shm

想着收缩下大表就有空间了,,,糊里糊涂开始shrink

[oracle@centos5 osa]$ sqlplus test/test

SQL*Plus: Release 10.2.0.4.0 - Production on Fri Jun 21 09:23:26 2013

Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.

Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options

SQL> select table_name,BLOCKS,EMPTY_BLOCKS,NUM_ROWS  from user_tables where table_name = upper('user_log');

TABLE_NAME                         BLOCKS EMPTY_BLOCKS   NUM_ROWS
------------------------------ ---------- ------------ ----------
user_log                      2606643            0  176544060

SQL> alter table user_log enable row movement;

Table altered.

SQL> alter table user_log shrink space cascade;  --索引也能缩小

Table altered.

SQL> select table_name,BLOCKS,EMPTY_BLOCKS,NUM_ROWS  from user_tables where table_name = upper('user_log');

TABLE_NAME                         BLOCKS EMPTY_BLOCKS   NUM_ROWS
------------------------------ ---------- ------------ ----------
user_log                      2606643            0  176544060

SQL>

shrink后还需要重新收集统计信息，但是当空间爆满时同样无法进行统计信息收集

SQL> exec dbms_stats.gather_table_stats('TEST','user_log');
BEGIN dbms_stats.gather_table_stats('TEST','user_log'); END;

*
ERROR at line 1:
ORA-01114: IO error writing block to file %s (block # %s)

想想其实收缩、整理，只是释放表空间的已用空间到空闲空间，并不会回收实际的磁盘

先删除了一些本身在在根目录下的oracle安装文件才可以做undo表空间的重建工作，否则没空间还是无法工作

由于大量操作导致了undo的自动扩展，占用了大量空间，准备收缩undo

[root@centos5 ~]# su - oracle
cd [oracle@centos5 ~]$ sqlplus test/test

SQL*Plus: Release 10.2.0.4.0 - Production on Fri Jun 21 11:39:40 2013

Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.

Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options

SQL> @space.sql

TABLESPACE_NAME      TOTAL       USED RATIO            FREE  MAX_BYTES
--------------- ---------- ---------- ---------- ---------- ----------
SYSTEM                 490     486.69   99.32          3.31       2.94
SYSAUX                 380        377   99.21             3        .81
TEST                 40960   20072.19   49.00      20887.81       3968
TS_OA                   50       5.06   10.12         44.94      39.94
USERS                    5        .44    8.80          4.56       4.56
TS_URP                 200       5.37    2.69        194.63     190.81
UNDOTBS1              9195     134.31    1.46       9060.69       2965
OSA_TEST               500       7.31    1.46        492.69     485.94
TS_IMPTEST              50        .06    0.12         49.94      49.94

9 rows selected.

SQL> select file_name,bytes/1024/1024 from dba_data_files where tablespace_name like 'UNDOTBS1';

FILE_NAME                                         BYTES/1024/1024
-----------------------------------　　　　　　　　----------------
/home/oracle/oradata/osa/undotbs01.dbf            9195

undo表空间的重建工作

--1、建立一个全新的undostb2
SQL>  create undo tablespace undotbs2 datafile '/home/oracle/oradata/osa/undotbs2.dbf' size 200M;

Tablespace created.

--2、修改系统的undo_tablespace为undotbs2
SQL> alter system set undo_tablespace=undotbs2 scope=both;

System altered.
--3、删除原有undo tablespace
SQL> drop tablespace undotbs1 including contents;

Tablespace dropped.
--4、删除数据文件释放空间
rm -rf /home/oracle/oradata/osa/undotbs01.dbf

再来看看空间

SQL> @space.sql              

TABLESPACE_NAME      TOTAL       USED RATIO            FREE  MAX_BYTES
--------------- ---------- ---------- ---------- ---------- ----------
SYSAUX                 380     377.94   99.46          2.06        .63
SYSTEM                 490     486.69   99.32          3.31       2.94
TEST                 40960   20072.19   49.00      20887.81       3968
TS_OA                   50       5.06   10.12         44.94      39.94
USERS                    5        .44    8.80          4.56       4.56
TS_URP                 200       5.37    2.69        194.63     190.81
OSA_TEST               500       7.31    1.46        492.69     485.94
UNDOTBS2               200       1.31    0.66        198.69     198.69
TS_IMPTEST              50        .06    0.12         49.94      49.94

再来收集统计信息查看下最高水位

SQL> select table_name,BLOCKS,EMPTY_BLOCKS,NUM_ROWS  from user_tables where table_name = upper('user_log');

TABLE_NAME                         BLOCKS EMPTY_BLOCKS   NUM_ROWS
------------------------------ ---------- ------------ ----------
user_log                      2606643            0  176544060

SQL>  exec dbms_stats.gather_table_stats('TEST','user_log');

PL/SQL procedure successfully completed.

SQL> select table_name,BLOCKS,EMPTY_BLOCKS,NUM_ROWS  from user_tables where table_name = upper('user_log');

TABLE_NAME                         BLOCKS EMPTY_BLOCKS   NUM_ROWS
------------------------------ ---------- ------------ ----------
user_log                      1719906            0  117470482    --shrink后统计信息更新了

如果是生产系统那会很惨，日常监控要非常注意数据库空间、系统空间