一,设计背景
由于所在公司ORACLE数据库较多,传统人工监控表空间的方式较耗时,且无法记录历史表空间数据,无法判断每日表空间增长量,在没有gridcontrol/cloudcontrol软件的情况下,笔者设计如下表空间监控方案,大家也可以根据自己的实际情况对下面的方案进行修改。
二,设计思路
通过dblink将来查询到的表空间数据集中汇总到一张表里通过crontab跑定时任务从各台服务器获取表空间使用情况信息。
三,具体实施步骤
1.所在oracle数据库ip地址信息(下面为举例说明具体情况要根据所在环境设置)
hostname | 备注 | ipaddress | tnsname | instancename |
tbsmonitor | 监控数据库 | 10.1.21.1 | tbsmonitor | tbsmonitor |
database1 | 被监控数据库 | 10.1.21.2 | database1 | database1 |
database2 | 被监控数据库 | 10.1.21.3 | database2 | database2 |
database3 | 被监控数据库 | 10.1.21.4 | database3 | database3 |
2.在tbsmonitor主机上创建tbsmonitor表空间
1
|
create tablespace tbsmonitor datafile ‘/opt/u01/app/oradata/tbsmonitor/tsmonitor.dbf’ size 50M autoextend on;
|
3.在tbsmonitor和database1/database2/database3上建立tbsmonitor用户用来做表空间监控。
1
|
create user tsmonitor identified by I11m8cb default tablespace tsmonitor;
|
4.为了tbsmonitor用户赋权用来查找表空间使用情况。
1
2
3
4
5
|
grant resource to tbsmonitor;
grant create session to tbsmonitor;
grant create table to tbsmonitor;
grant select on dba_data_files to tbsmonitor;
grant select on dba_free_space to tbsmonitor;
|
5.在tbsmonitor上建立database1/ database2/ database3的tnsnames.ora连接,在tnsnames.ora文件中加入
1
2
3
4
5
6
7
8
9
10
11
12
|
DATABASE1 =
(DESCRIPTION=
(ADDRESS=(PROTOCOL=TCP)(HOST=10.1.21.1)(PORT=1521))
(CONNECT_DATA=(SID= database1)))
DATABASE2 =
(DESCRIPTION=
(ADDRESS=(PROTOCOL=TCP)(HOST=10.1.21.2)(PORT=1521))
(CONNECT_DATA=(SID= database2)))
DATABASE3 =
(DESCRIPTION=
(ADDRESS=(PROTOCOL=TCP)(HOST=10.1.21.3)(PORT=1521))
(CONNECT_DATA=(SID= database3)))
|
6.修改/etc/hosts文件,如果有dns服务器的话可以略过
1
2
3
|
10.1.21.2 database1
10.1.21.3 database2
10.1.21.4 database3
|
7.在tbsmonitor主机设置dblink,这样就能通过dblink从被监控服务器远程抽取表空间信息。
1
2
3
4
5
6
7
8
9
|
create database link TO_DATABASE1
connect to TSMONITOR identified by I11m08cb
using ‘DATABASE1’;
create database link TO_DATABASE2
connect to TSMONITOR identified by I11m08cb
using ‘DATABASE2’;
create database link TO_DATABASE3
connect to TSMONITOR identified by I11m08cb
using ‘DATABASE3’;
|
8.建立tbsmonitor表,表空间统计数据将插入这张表。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
create table tbsmonitor.tbsmonitor
(
ipaddress VARCHAR2(200),
instancename VARCHAR2(200),
tablespace_name VARCHAR2(200),
datafile_count NUMBER,
size_mb NUMBER,
free_mb NUMBER,
used_mb NUMBER,
maxfree NUMBER,
pct_used NUMBER,
pct_free NUMBER,
time DATE
) tablespace tbsmonitor;
|
9. 在crontab中运行每日0点1分更新数据库表空间信息的脚本tbsmonitor.sh(我根据业务需要每日统计一次,大家也可以通过业务要求修改统计频率)
1
|
1 0 * * * /opt/u01/app/oracle/tbsmonitor.sh
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
|
#!/bin/bash
#FileName: tbsmonitor.sh
#CreateDate:2016-01-1
#version:1.0
#Discription:take the basic information to insert into the table tbs_usage
# Author:FUZHOU HOT
#Email:15980219172@139.com
ORACLE_SID= tbsmonitor
ORACLE_BASE=/opt/u01/app
ORACLE_HOME=/opt/u01/app/oracle
PATH=$ORACLE_HOME/bin:$PATH;export PATH
export ORACLE_SID ORACLE_BASE ORACLE_HOME
date>>/opt/u01/app/oracle/tbsmonitor.sh
sqlplus sys/I11m08cb as sysdba <<EOF >> /opt/u01/app/oracle/tbsmonitor.log 2>&1
@/opt/u01/app/oracle/tbsmonitor/ tbsmonitor.sql;
@/opt/u01/app/oracle/tbsmonitor/database1.sql;
@/opt/u01/app/oracle/tbsmonitor/database2.sql;
@/opt/u01/app/oracle/tbsmonitor/database3.sql;
EOF
echo >> /opt/u01/app/oracle/ tbsmonitor.log
|
11.创建插入脚本(拿database1举例,以此类推)
1
2
3
|
/opt/u01/app/oracle/tbsmonitor/database1.sql; /opt/u01/app/oracle/tbsmonitor/database2.sql;
/opt/u01/app/oracle/tbsmonitor/database3.sql;
/opt/u01/app/oracle/tbsmonitor/ tbsmonitor.sql;
|
Sql脚本如下
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
insert into tsmonitor.tbsmonitor SELECT utl_inaddr.get_host_address(‘DATABASE1’) ipaddress,
(select instance_name from v$instance) instancename,
df.tablespace_name,
COUNT(*) datafile_count,
ROUND(SUM(df.BYTES) / 1048576) size_mb,
ROUND(SUM(free.BYTES) / 1048576, 2) free_mb,
ROUND(SUM(df.BYTES) / 1048576 – SUM(free.BYTES) / 1048576, 2) used_mb,
ROUND(MAX(free.maxbytes) / 1048576, 2) maxfree,
100 – ROUND(100.0 * SUM(free.BYTES) / SUM(df.BYTES), 2) pct_used,
ROUND(100.0 * SUM(free.BYTES) / SUM(df.BYTES), 2) pct_free,sysdate time
FROM dba_data_files@TO_DATABASE1 df,
(SELECT tablespace_name,
file_id,
SUM(BYTES) BYTES,
MAX(BYTES) maxbytes
FROM dba_free_space@TO_DATABASE1
GROUP BY tablespace_name, file_id) free
WHERE df.tablespace_name = free.tablespace_name(+)
AND df.file_id = free.file_id(+)
GROUP BY df.tablespace_name
ORDER BY 6;
|
12.查看表空间使用占比可以使用如下语句(如果要查看某台机器可以带上条件where ipaddress=’xxxx’ and instance=’xxxxx’ and to_char(time,’yyyy-mm-dd’)=’xxxx-xx-xx’)
1
2
3
4
5
6
7
8
9
10
11
|
SELECT IPADDRESS ,
Instancename,
tablespace_name,
datafile_count,
size_mb “表空间大小(M)”,
used_mb “已使用空间(M)”,
TO_CHAR(ROUND((used_mb) / size_mb * 100,
2),
‘990.99’) “使用比”,
free_mb “空闲空间(M)”
FROM tbsmonitor. tbsmonitor order by “使用比” desc
|
13.查看每日增量可以使用如下脚本。(下面显示的是4-8日10.1.21.2表空间增长的情况)
1
2
3
4
5
|
select a.tablespace_name,(b.used_mb–a.used_mb) increase,a.ipaddress from
(select * from tsmonitor.tbs_usage where to_char(time,‘yyyy-mm-dd’)=‘2016-01-04’) a,
(select * from tsmonitor.tbs_usage where to_char(time,‘yyyy-mm-dd’)=‘2016-01-08’) b
where a.tablespace_name=b.tablespace_name and a.IPADDRESS=b.IPADDRESS order by increase desc
select * from tbsmonitor. tbsmonitor where ipaddress=‘10.1.21.2’ and to_char(time,‘yyyy-mm-dd’)=‘2016-01-08’
|
文章转载来自:ttlsa.com