一、错误提示：
EXP-00008: Oracle error 4031 encountered
ORA-04031: unable to allocate 4096 bytes of shared memory
("shared pool","BEGIN:EXEC_STR := SYS.DBMS...","PL/SQL MPCODE","BAMIMA: Bam Buffe
二、错误原因：
共享内存太小，存在一定碎片，没有有效的利用保留区，造成无法分配合适的共享区。
三、解决步骤：
1.查看当前环境
SQL show sga
Total System Global Area566812832 bytes
Fixed Size73888 bytes
Variable Size28811264 bytes
Database Buffers536870912 bytes
Redo Buffers1056768 bytes
show parameter shared_pool
NAME TYPEVALUE
------------------------------------ ------- -----
shared_pool_reserved_sizestring1048576
shared_pool_size string20971520
SQL select sum(free_space) from v$shared_pool_reserved;
SUM(FREE_SPACE)
---------------
1048576
我们可以看到没有合理利用保留区
SQL SELECT SUM(RELOADS)/SUM(PINS) FROM V$LIBRARYCACHE;
SUM(RELOADS)/SUM(PINS)
----------------------
.008098188
不算太严重
SQL SELECT round((B.Value/A.Value)*100,1) hardpaseperc
FROM V$SYSSTAT A,
V$SYSSTAT B
WHERE A.Statistic# = 171
AND B.Statistic# = 172
AND ROWNUM = 1;
hardpaseperc
------------------
26.5
2.查看保留区使用情况
SQL SELECT FREE_SPACE,
FREE_COUNT,
REQUEST_FAILURES,
REQUEST_MISSES,
LAST_FAILURE_SIZE
FROM V$SHARED_POOL_RESERVED;
FREE_SPACE FREE_COUNT REQUEST_FAILURES REQUEST_MISSES LAST_FAILURE_SIZE
---------- ---------- ---------------- -------------- -----------------
1048576114604132
最近一次申请共享区失败时该对象需要的共享区大小4132
select name from v$db_object_cache where sharable_mem = 4132;
name
----------------
dbms_lob
-- dbms_lob正是exp时申请保留区的对象
3.查看导致换页的应用
SQL select * from x$ksmlru where ksmlrsiz0;
ADDR INDXINST_ID KSMLRCOMKSMLRSIZKSMLRNUM KSMLRHON KSMLROHV KSMLRSES
50001A88 01 BAMIMA: Bam Buffer4100 64 DBMS_DDL 402745060 730DEB9C
50001ACC 11 BAMIMA: Bam Buffer4108736 DBMS_SYS_SQL 1909768749 730D0838
50001B10 21 BAMIMA: Bam Buffer4112 1576 STANDARD 2679492315 730D7E20
50001B54 31 BAMIMA: Bam Buffer4124 1536 DBMS_LOB 853346312 730DA83C
50001B98 41 BAMIMA: Bam Buffer4128 3456 DBMS_UTILITY 4041615653 730C5FC8
50001BDC 51 BAMIMA: Bam Buffer4132 3760 begin :1 := dbms_lob.getLeng...2942875191 730CFFCC
50001C20 61 state objects 4184 1088 0 00
50001C64 71 library cache 4192488 EXU8VEW 2469165743 730C1C68
50001CA8 81 state objects 4196 16 0 730C0B90
50001CEC 91 state objects 4216 3608 0 730D0838
3.分析各共享池的使用情况
SQL select KSPPINM,KSPPSTVL
from x$ksppi,
x$ksppcv
where x$ksppi.indx = x$ksppcv.indx
and KSPPINM = '_shared_pool_reserved_min_alloc';
KSPPINM KSPPSTVL
------------------------------- --------
_shared_pool_reserved_min_alloc 4400--(门值)
我们看到INDX=5,DBMS_LOB造成换页(就是做exp涉及到lob对象处理造成的换页情况),换出
最近未使用的内存,但是换出内存并合并碎片后在共享区仍然没有合适区来存放数据,说明共享
区小和碎片过多，然后根据_shared_pool_reserved_min_alloc的门值来申请保留区,而门值为4400,
所以不符合申请保留区的条件,造成4031错误。
我们前面看到保留区全部为空闲状态,所以我们可以
减低门值，使更多申请共享内存比4400小的的对象能申请到保留区，而不造成4031错误。
4.解决办法：
1).增大shared_pool （在不DOWN机的情况下不合适）
2).打patch （在不DOWN机的情况下不合适）
3).减小门值 （在不DOWN机的情况下不合适）
因为LAST_FAILURE_SIZE<_shared_pool_reserved_min_alloc所以表明没有有效的使用保留区
SQL alter system set "_shared_pool_reserved_min_alloc" = 4000;
alter system set "_shared_pool_reserved_min_alloc"=4000
*
ERROR at line 1:
ORA-02095: specified initialization parameter cannot be modified
-- 9i的使用方法alter system set "_shared_pool_reserved_min_alloc"=4000 scope=spfile;
4).使用alter system flush shared_pool; (不能根本性的解决问题)
5).使用dbms_shared_pool.keep
5.由于数据库不能DOWN机，所以只能选择3)和4)
运行dbmspool.sql
SQL @/home/oracle/prodUCts/8.1.7/rdbms/admin/dbmspool.sql
找出需要keep到共享内存的对象
SQL select a.OWNER,
a.name,
a.sharable_mem,
a.kept,
a.EXECUTIONS ,
b.address,
b.hash_value
from v$db_object_cache a,
v$sqlarea b
where a.kept = 'NO' and
(( a.EXECUTIONS 1000
and a.SHARABLE_MEM 50000)
ora.EXECUTIONS 10000)
and SUBSTR(b.sql_text,1,50) = SUBSTR(a.name,1,50);
OWNERNAMESHARABLE_MEM KEP EXECUTIONS ADDRESSHASH_VALUE
-----------------------------—--------------- --- ---------- -------- ----------
SELECT COUNT(OBJECT_ID) 98292NO 10320774814BF8 1893309624
FROM ALL_OBJECTS
WHERE OBJECT_NAME = :b1
AND OWNER = :b2
STANDARD286632 NO 13501
DBMS_LOB98292 NO 103750
DBMS_LOB 47536NO 2886542
DBMS_LOB 11452NO 2864757
DBMS_PICKLER10684NO 2681194
DBMS_PICKLER5224 NO 2663860
SQL execute dbms_shared_pool.keep('STANDARD');
SQL execute dbms_shared_pool.keep('74814BF8,1893309624','C');
SQL execute dbms_shared_pool.keep('DBMS_LOB');
SQL execute dbms_shared_pool.keep('DBMS_PICKLER');
SQL select OWNER, name, sharable_mem,kept,EXECUTIONS from v$db_object_cache where kept = 'YES' ORDER BY sharable_mem;
SQL alter system flush shared_pool;
System altered.
SQL SELECT POOL,BYTES FROM V$SGASTAT WHERE NAME ='free memory';
POOL BYTES
----------- ----------
shared pool7742756
large pool614400
Java pool32768
[oracle@ali-solution oracle]$ sh /home/oracle/admin/dbexp.sh
[oracle@ali-solution oracle]$ grep ORA- /tmp/exp.tmp
未发现错误，导出数据成功
四、建议：
由于以上解决的方法是在不能DOWN机的情况下，所以没能动态修改初始化参数，
但问题的本质是共享区内存过小，需要增加shared pool，使用绑定变量，才能根本
的解决问题，所以需要在适当的时候留出DOWN机时间，对内存进行合理的配置。
