OracleBlog – 第 17 页 – 笑看数据库江湖的那些事儿……

检查字段中不可见字符

2015-09-12 没有评论

在一个mysql导入到oracle的库中，某字段显示为’a’，但是用where column=’a’ 返回0行，只有where column like ‘a%’ 才有返回结果。那么，在该字段的末尾，应该是有些不可见字符了。我们可 […]

SQL> ALTER SYSTEM SET SERVICE_NAMES='oracleblog.org, jimmy.oracleblog.org, jimmy.he.oracleblog.org, hejianmin.oracleblog.org, he.jian.min.oracleblog.org, jimmy_he_oracleblog.org, jimmy_he_he.jian_min.oracleblog.org, jian_min_he.jimmy.oracleblog.org, jimmy.he.he_jianmin.oracleblog.org, jimmyhe_hejianmin.oracleblog.org, jimmy.he.he.jianmin.oracleblog.org' scope=both;

ALTER SYSTEM SET SERVICE_NAMES='oracleblog.org, jimmy.oracleblog.org, jimmy.he.oracleblog.org, hejianmin.oracleblog.org, he.jian.min.oracleblog.org, jimmy_he_oracleblog.org, jimmy_he_he.jian_min.oracleblog.org, jian_min_he.jimmy.oracleblog.org, jimmy.he.he_jianmin.oracleblog.org, jimmyhe_hejianmin.oracleblog.org, jimmy.he.he.jianmin.oracleblog.org' scope=both

*

ERROR at line 1:

ORA-32021: parameter value longer than 255 characters

SQL>

虽然可以通过dbms_service包来创建和修改service […]

..experience, Working case

停10g的css进程

2015-07-30 没有评论

在10g中，如果使用asm存储，且不使用clusterware，那么在安装asm的时候，就会要求启动css服务，启动css服务是以root运行运行$ORACLE_HOME/bin下的localconfig add。然后，css服务就开始运行了，你也可以建立asm实例，在asm上建单实例的数据库了。 […]

..experience, Working case

另一种找asm disk dir的方法

2015-07-24 一条评论

其实还有另一种找asm disk dir的方法，这种方法利用了官方提供的amdu的工具，找起来更简单。先利用amdu得到map文件，命令为：

amdu -dump <diskgroup_name|ALL> -diskstring '<diskstring_path>'

1	amdu -dump <diskgroup_name\|ALL> -diskstring '<diskstring_path>'

如，查到所有的diskgroup： [crayon-677d790a5fb03318239 […]

..unix shell, Useful script

查询asm disk directory的脚本

2015-07-19 没有评论

简单写了一个查询asm disk directory。我们一般在asm磁盘mount不了的时候，需要检查磁盘头的信息，有时候盘头的信息会和disk directory的信息不一致，需要手工的修改disk dir或者盘头的信息。由于找disk dir有点麻烦，需要先找到file 1 block 1 […]

..experience, Working case

HP平台的asm要求maxfiles_lim参数为32767

2015-07-03 2条评论

最近在客户遇到一个案例，和大家分享一下，大家如果客户那里也有HP平台，那么要注意在11gR2中kernal参数maxfile_lim要设置成32767，而在11gR1以及之前，这个参数是65536的。（在169706.1文档上，11gR1之前是设置65536的，在11gR2上要求参考install […]

..experience, Working case

asm alertlog大量报错Voting file relocation is required in diskgroup

2015-07-02 没有评论

客户将asm的实例配置错了主机，asm1运行在节点2，asm2运行在节点1，在进行de-config和re-config之后，asm的alertlog大量报错： SYMPTOMS:

NOTE: Attempting voting file refresh on diskgroup GRID
NOTE: Voting file relocation is required in diskgroup GRID
NOTE: Attempting voting file relocation on diskgroup GRID
NOTE: Attempting voting file refresh on diskgroup GRID
NOTE: Voting file relocation is required in diskgroup GRID
NOTE: Attempting voting file relocation on diskgroup GRID
NOTE: Attempting voting file refresh on diskgroup GRID

1

2

3

4

5

6

7

NOTE: Attempting voting file refresh on diskgroup GRID

NOTE: Voting file relocation is required in diskgroup GRID

NOTE: Attempting voting file relocation on diskgroup GRID

NOTE: Attempting voting file refresh on diskgroup GRID

NOTE: Voting file relocation is required in diskgroup GRID

NOTE: Attempting voting file relocation on diskgroup GRID

NOTE: Attempting voting file refresh on diskgroup GRID

CAUSE: [crayon-677d790a61c […]

..experience, Working case

高IO下的library cache lock 和cursor pin S wait on X

2015-07-02 没有评论

今天，客户端某数据库遭遇大量library cache lock 和cursor pin S wait on X的问题，经诊断发现是由于他们底层的ASM存储是多个库共享的，故障时间段别的库占用大量的IO，导致paitsm库的IO响应变慢，db file sequential read从平时的20多毫 […]

..experience, Working case

配置odbc透明网关实现oracle访问postgres DB

2015-07-02 没有评论

最近帮客户配置了一下通过odbc透明网关，实现在oracle内通过db link访问postgres DB。简单记录一下：（1）listener.ora和tnsnames.ora的配置：

[wsj81@localhost admin]$ cat listener.ora
# listener.ora Network Configuration File: /wsj/oracle/app/product/11.2.0/dbhome_1/network/admin/listener.ora
# Generated by Oracle configuration tools.

SID_LIST_ORCL =
  (SID_LIST =
    (SID_DESC =
      (SID_NAME = pgdb)
      (ORACLE_HOME = /wsj/oracle/app/product/11.2.0/dbhome_1)
      (ENV="LD_LIBRARY_PATH=/usr/lib64:/wsj/oracle/app/product/11.2.0/dbhome_1/lib:usr/local/lib")
      (PROGRAM=dg4odbc)
    )
  )

ORCL =
  (DESCRIPTION_LIST =
    (DESCRIPTION =
      (ADDRESS_LIST =
      (ADDRESS = (PROTOCOL = TCP)(HOST = 19.89.64.111)(PORT = 1531))
      )
    )
  ) 


[wsj81@localhost admin]$ 
[wsj81@localhost admin]$ cat  tnsnames.ora
pgdb =
  (DESCRIPTION=
    (ADDRESS=(PROTOCOL=tcp)(HOST=19.89.64.111)(PORT=1521))
    (CONNECT_DATA=(SID=pgdb))
    (HS=OK)
  )

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

[wsj81@localhost admin]$ cat listener.ora

# listener.ora Network Configuration File: /wsj/oracle/app/product/11.2.0/dbhome_1/network/admin/listener.ora

# Generated by Oracle configuration tools.

SID_LIST_ORCL =

(SID_LIST =

(SID_DESC =

(SID_NAME = pgdb)

(ORACLE_HOME = /wsj/oracle/app/product/11.2.0/dbhome_1)

(ENV="LD_LIBRARY_PATH=/usr/lib64:/wsj/oracle/app/product/11.2.0/dbhome_1/lib:usr/local/lib")

(PROGRAM=dg4odbc)

)

ORCL =

(DESCRIPTION_LIST =

(DESCRIPTION =

(ADDRESS_LIST =

(ADDRESS = (PROTOCOL = TCP)(HOST = 19.89.64.111)(PORT = 1531))

)

[wsj81@localhost admin]$

[wsj81@localhost admin]$ cat tnsnames.ora

pgdb =

(DESCRIPTION=

(ADDRESS=(PROTOCOL=tcp)(HOST=19.89.64.111)(PORT=1521))

(CONNECT_DATA=(SID=pgdb))

(HS=OK)

)

（2）透明网关的配置，这2个文件在$ORA […]

..experience, Working case

instance不随机启动

2015-01-01 没有评论

客户有个机器上的instance，在机器重启后，总是不随机启动，每次都要手工的srvctl的去start一次。这其实是crs的配置缘故：

[root@rac1 ~]# crsctl stat res ora.ora11g.db -p 
NAME=ora.ora11g.db 
TYPE=ora.database.type 
ACL=owner:oracle:rwx,pgrp:oinstall:r--,other::r--,group:dba:r-x,group:oinstall:r-x,user:oracle:r-x 
ACTION_FAILURE_TEMPLATE= 
ACTION_SCRIPT= 
ACTIVE_PLACEMENT=1 
AGENT_FILENAME=%CRS_HOME%/bin/oraagent%CRS_EXE_SUFFIX% 
AUTO_START=restore <<======这里配置的是restore，即上次是手工关闭过，等server重启后，instance也是启动到关机前关闭状态。

1

2

3

4

5

6

7

8

9

[root@rac1 ~]# crsctl stat res ora.ora11g.db -p

NAME=ora.ora11g.db

TYPE=ora.database.type

ACL=owner:oracle:rwx,pgrp:oinstall:r--,other::r--,group:dba:r-x,group:oinstall:r-x,user:oracle:r-x

ACTION_FAILURE_TEMPLATE=

ACTION_SCRIPT=

ACTIVE_PLACEMENT=1

AGENT_FILENAME=%CRS_HOME%/bin/oraagent%CRS_EXE_SUFFIX%

AUTO_START=restore <<======这里配置的是restore，即上次是手工关闭过，等server重启后，instance也是启动到关机前关闭状态。

将auto_start修改成always就可以了。修改方式： [crayon-677d790a […]

..unix shell, Useful script

批量取statspack的脚本

2015-01-01 一条评论

做了一个脚本sprpt_batch.sh:

read line
snap_i_id=$1
end_snap=$2

sqlplus -s /nolog<<EOF
conn /as sysdba;
define begin_snap=${snap_i_id};
define end_snap=${end_snap};
define report_name=sprpt_batch_${snap_i_id}_${end_snap}.txt
set echo off
set feedback off
@myspreport
exit
EOF

1

2

3

4

5

6

7

8

9

10

11

12

13

14

read line

snap_i_id=$1

end_snap=$2

sqlplus -s /nolog<<EOF

conn /as sysdba;

define begin_snap=${snap_i_id};

define end_snap=${end_snap};

define report_name=sprpt_batch_${snap_i_id}_${end_snap}.txt

set echo off

set feedback off

@myspreport

exit

EOF

将$ORACLE_HOME/rdbms/admin/spreport.sql和$ORACLE_HOME/rdbms/admin/sprepins.sql拷贝到工作目录下，重命名成myspre […]

Study note

切出的归档日志小于设置的redo大小

2015-01-01 没有评论

数据库产生大量的小日志，原因和log_buffer的大小和redo log file size有关。为了说明这个问题，我们先来看看从log buffer开始写redo log file的过程：（1）写redo log buffer 在pga中产生redo entry（即change record， […]