今天客戶要進行應(yīng)用發(fā)布，首先在測試環(huán)境進行測試，在測試環(huán)境測試索引重建的時候報如下錯誤：

ORA-04030: 在嘗試分配 64544 字節(jié) (sort subheap,sort key) 時進程內(nèi)存不足。

客戶在將這個問題發(fā)給我的時候，首先讓他們檢查一下操作系統(tǒng)ORACLE用戶資源的限制。

因為根據(jù)以往的經(jīng)驗這個問題往往是oracle用戶進程的data seg size限制導(dǎo)致的。

但是通過客戶過來的信息發(fā)現(xiàn) oracle用戶程序數(shù)據(jù)段并沒有任何限制。

$ id oracle

uid=400(oracle) gid=400(oinstall) groups=401(dba)

$ ulimit -a

time(seconds)??????? unlimited

file(blocks)???????? unlimited

data(kbytes)???????? unlimitedstack(kbytes)??????? unlimited

memory(kbytes)?????? unlimited

coredump(blocks)???? unlimited

nofiles(descriptors) 2000

$ ulimit -Ha

time(seconds)??????? unlimited

file(blocks)???????? unlimited

data(kbytes)???????? unlimitedstack(kbytes)??????? unlimited

memory(kbytes)?????? unlimited

coredump(blocks)???? unlimited

nofiles(descriptors) unlimited

然后檢查了一下客戶的SGA和PGA設(shè)置：

SQL> show parameter sga

NAME???????????????????????????????? TYPE??????? VALUE

------------------------------------ ----------- ------------------------------

lock_sga???????????????????????????? boolean???? FALSE

pre_page_sga???????????????????????? boolean???? FALSE

sga_max_size???????????????????????? big integer 15G

sga_target?????????????????????????? big integer 14G

SQL> show parameter pga

NAME???????????????????????????????? TYPE??????? VALUE

------------------------------------ ----------- ------------------------------

pga_aggregate_target???????????????? big integer 8G

$ lsattr -El mem0

goodsize 40960 Amount of usable physical memory in Mbytes False

size???? 40960 Total amount of physical memory in Mbytes? False

客戶的SGA設(shè)置為14G，PGA設(shè)置為8G ，而操作系統(tǒng)的整個內(nèi)存是40G，內(nèi)存遠遠夠用的。

C:\Documents and Settings\shoupeng.yan>sqlplus xxxx/xxxx

SQL*Plus: Release 10.2.0.1.0 - Production on 星期三 3月 21 20:00:37 2012

Copyright (c) 1982, 2005, Oracle.? All rights reserved.

連接到:

Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit Production

With the Partitioning, Oracle Label Security, OLAP, Data Mining Scoring Engine

and Real Application Testing options

SQL> ALTER INDEX PK_ZJ_BDZTB_01 REBUILD;

ALTER INDEX PK_ZJ_BDZTB_01 REBUILD

*

第 1 行出現(xiàn)錯誤:

ORA-04030: 在嘗試分配 64544 字節(jié) (sort subheap,sort key) 時進程內(nèi)存不足

我通過windows下的sqlplus遠程連到數(shù)據(jù)庫中，親自執(zhí)行了一下索引重建，在重建的過程中順監(jiān)控了一下進程的PGA使用量。發(fā)現(xiàn)進程的PGA使用量大約在110M的時候就拋出ORA-04030錯誤了。而系統(tǒng)的整個PGA設(shè)置為8G，不應(yīng)該不夠用。

因此我還是認(rèn)為os limits限制導(dǎo)致的問題原因,既然單個進程的內(nèi)存段有最大限制，我們可以通過開啟索引重建并行度以便每個進程占用的數(shù)據(jù)段減少一半以上。

在我開啟了2個并行度之后，索引確實重建成功了。

SQL> ALTER INDEX PK_ZJ_BDZTB_01 REBUILD PARALLEL 2;

索引已更改。

更奇怪的事情是：我直接登錄到數(shù)據(jù)庫服務(wù)器上進程索引重建，沒有加并行度也可以重建成功。

$ id oracle

uid=400(oracle) gid=400(oinstall) groups=401(dba)

$ sqlplus xxxx/xxxx

SQL*Plus: Release 10.2.0.4.0 - Production on Wed Mar 21 20:07:20 2012

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, Oracle Label Security, OLAP, Data Mining Scoring Engine

and Real Application Testing options

SQL>? ALTER INDEX PK_ZJ_BDZTB_01 REBUILD;

Index altered.

一個是通過Windows客戶端連接到數(shù)據(jù)庫上的，一個是通過本地sqlplus登錄到數(shù)據(jù)庫中的,一個重建索引不成功，一個能成功。唯一的區(qū)別是，客戶端連接時通過監(jiān)聽器，那么問題應(yīng)該出現(xiàn)在監(jiān)聽器上。

為此我做了一個假設(shè)：雖然目前的ulimit -a顯示的結(jié)果表示ORACLE用戶進程的數(shù)據(jù)段沒有限制，但是這個修改很可能是在監(jiān)聽器啟動之后，監(jiān)聽器啟動之后繼承了先前的ulimits的所有設(shè)置，而在此后重新修改的并沒有反饋到監(jiān)聽器中。而通過客戶端sqlplus連到數(shù)據(jù)庫中的時候，

監(jiān)聽器派生出的服務(wù)器進程繼承了監(jiān)聽器中的ulimits相關(guān)的設(shè)置，導(dǎo)致服務(wù)器進程的數(shù)據(jù)段還是有限制的。而通過本地連接的數(shù)據(jù)庫不通過監(jiān)聽器，直接繼承了當(dāng)前的設(shè)置，所以本地連接建立索引不會報ORA-4030錯誤。

然后在OS上通過smit.script 我找了當(dāng)前修改oracle用戶的limits限制的腳本：

#????[Dec 21 2010, 16:36:27]#

x() {

if [ $# -ge 2 ]

then

for i in "$@"

do

spam="$spam \"$i\""

done

eval chuser $spam

fi

}

x data='-1' stack='-1' oracle

這條命令是在2010年12月21日執(zhí)行的。

而監(jiān)聽器是在10-DEC-2010 09:44:38 啟動的，到目前運行了467天，也就是監(jiān)聽器在chuser data='-1' stack='-1' oracle 命令執(zhí)行之前就啟動了。

$ lsnrctl status

LSNRCTL for IBM/AIX RISC System/6000: Version 10.2.0.4.0 - Production on 21-MAR-2012 20:11:18

Copyright (c) 1991, 2007, Oracle.? All rights reserved.

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=SXRYXDB)(PORT=1521)))

STATUS of the LISTENER

------------------------

Alias???????????????????? LISTENER

Version?????????????????? TNSLSNR for IBM/AIX RISC System/6000: Version 10.2.0.4.0 - Production

Start Date??????????????? 10-DEC-2010 09:44:38

Uptime??????????????????? 467 days 10 hr. 26 min. 40 sec

Trace Level?????????????? off

Security????????????????? ON: Local OS Authentication

SNMP????????????????????? OFF

Listener Log File???????? /u01/oracle/product/db10gr2/network/log/listener.log

Listening Endpoints Summary...

(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=SXRYXDB)(PORT=1521)))

Services Summary...

Service "zyxdb" has 1 instance(s).

Instance "zyxdb", status READY, has 1 handler(s) for this service...

Service "zyxdb_XPT" has 1 instance(s).

Instance "zyxdb", status READY, has 1 handler(s) for this service...

The command completed successfully

既然問題是由于監(jiān)聽器導(dǎo)致的，那么重啟一下監(jiān)聽器，重新讀取修改過的ulimits,即可解決問題。

$ lsnrctl stop

LSNRCTL for IBM/AIX RISC System/6000: Version 10.2.0.4.0 - Production on 21-MAR-2012 20:11:32

Copyright (c) 1991, 2007, Oracle.? All rights reserved.

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=SXRYXDB)(PORT=1521)))

The command completed successfully

$ lsnrctl start

LSNRCTL for IBM/AIX RISC System/6000: Version 10.2.0.4.0 - Production on 21-MAR-2012 20:11:43

Copyright (c) 1991, 2007, Oracle.? All rights reserved.

Starting /u01/oracle/product/db10gr2/bin/tnslsnr: please wait...

TNSLSNR for IBM/AIX RISC System/6000: Version 10.2.0.4.0 - Production

System parameter file is /u01/oracle/product/db10gr2/network/admin/listener.ora

Log messages written to /u01/oracle/product/db10gr2/network/log/listener.log

Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=SXRYXDB)(PORT=1521)))

Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=EXTPROC0)))

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=SXRYXDB)(PORT=1521)))

STATUS of the LISTENER

------------------------

Alias???????????????????? LISTENER

Version?????????????????? TNSLSNR for IBM/AIX RISC System/6000: Version 10.2.0.4.0 - Production

Start Date??????????????? 21-MAR-2012 20:11:43

Uptime??????????????????? 0 days 0 hr. 0 min. 0 sec

Trace Level?????????????? off

Security????????????????? ON: Local OS Authentication

SNMP????????????????????? OFF

Listener Parameter File?? /u01/oracle/product/db10gr2/network/admin/listener.ora

Listener Log File???????? /u01/oracle/product/db10gr2/network/log/listener.log

Listening Endpoints Summary...

(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=SXRYXDB)(PORT=1521)))

(DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=EXTPROC0)))

Services Summary...

Service "PL***tProc" has 1 instance(s).

Instance "PL***tProc", status UNKNOWN, has 1 handler(s) for this service...

The command completed successfully

$ sqlplus / as sysdba

SQL*Plus: Release 10.2.0.4.0 - Production on Wed Mar 21 20:11:47 2012

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, Oracle Label Security, OLAP, Data Mining Scoring Engine

and Real Application Testing options

SQL> alter system register;

System altered.

SQL>

再次通過客戶端執(zhí)行索引重建命令問題解決：

C:\Documents and Settings\shoupeng.yan>sqlplus xxxx/xxxx

SQL*Plus: Release 10.2.0.1.0 - Production on 星期三 3月 21 20:12:18 2012

Copyright (c) 1982, 2005, Oracle.? All rights reserved.

連接到:

Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit Production

With the Partitioning, Oracle Label Security, OLAP, Data Mining Scoring Engine

and Real Application Testing options

SQL>?ALTER INDEX PK_ZJ_BDZTB_01 REBUILD;

索引已更改。

SQL>

總結(jié)

以上是生活随笔為你收集整理的oracle10g 04030,一次ORA-04030问题的诊断（一）的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。