Shell脚本中执行sql语句操作mysql的5种方法

对于自动化运维,诸如备份恢复之类的,DBA经常需要将SQL语句封装到shell脚本。本文描述了在Linux环境下mysql数据库中,shell脚本下调用sql语句的几种方法,供大家参考。对于脚本输出的结果美化,需要进一步完善和调整。以下为具体的示例及其方法。

1、将SQL语句直接嵌入到shell脚本文件中


--演示环境  

[root@SZDB ~]# more /etc/issue  

CentOS release 5.9 (Final)  

Kernel \r on an \m  

  

root@localhost[(none)]> show variables like 'version';  

+---------------+------------+  

| Variable_name | Value      |  

+---------------+------------+  

| version       | 5.6.12-log |  

+---------------+------------+  

  

[root@SZDB ~]# more shell_call_sql1.sh   

#!/bin/bash  

# Define log  

TIMESTAMP=`date +%Y%m%d%H%M%S`  

LOG=call_sql_${TIMESTAMP}.log  

echo "Start execute sql statement at `date`." >>${LOG}  

  

# execute sql stat  

mysql -uroot -p123456 -e "  

tee /tmp/temp.log  

drop database if exists tempdb;  

create database tempdb;  

use tempdb  

create table if not exists tb_tmp(id smallint,val varchar(20));  

insert into tb_tmp values (1,'jack'),(2,'robin'),(3,'mark');  

select * from tb_tmp;  

notee  

quit"  

  

echo -e "\n">>${LOG}  

echo "below is output result.">>${LOG}  

cat /tmp/temp.log>>${LOG}  

echo "script executed successful.">>${LOG}  

exit;  

  

[root@SZDB ~]# ./shell_call_sql1.sh   

Logging to file '/tmp/temp.log'  

+------+-------+  

| id   | val   |  

+------+-------+  

|    1 | jack  |  

|    2 | robin |  

|    3 | mark  |  

+------+-------+  

Outfile disabled.  

2、命令行调用单独的SQL文件


[root@SZDB ~]# more temp.sql   

tee /tmp/temp.log  

drop database if exists tempdb;  

create database tempdb;  

use tempdb  

create table if not exists tb_tmp(id smallint,val varchar(20));  

insert into tb_tmp values (1,'jack'),(2,'robin'),(3,'mark');  

select * from tb_tmp;  

notee  

  

[root@SZDB ~]# mysql -uroot -p123456 -e "source /root/temp.sql"  

Logging to file '/tmp/temp.log'  

+------+-------+  

| id   | val   |  

+------+-------+  

|    1 | jack  |  

|    2 | robin |  

|    3 | mark  |  

+------+-------+  

Outfile disabled.  

3、使用管道符调用SQL文件


[root@SZDB ~]# mysql -uroot -p123456 </root/temp.sql  

Logging to file '/tmp/temp.log'  

id      val  

1       jack  

2       robin  

3       mark  

Outfile disabled.  

  

#使用管道符调用SQL文件以及输出日志  

[root@SZDB ~]# mysql -uroot -p123456 </root/temp.sql >/tmp/temp.log  

[root@SZDB ~]# more /tmp/temp.log  

Logging to file '/tmp/temp.log'  

id      val  

1       jack  

2       robin  

3       mark  

Outfile disabled.  

4、shell脚本中MySQL提示符下调用SQL


[root@SZDB ~]# more shell_call_sql2.sh  

#!/bin/bash  

mysql -uroot -p123456 <<EOF  

source /root/temp.sql;  

select current_date();  

delete from tempdb.tb_tmp where id=3;  

select * from tempdb.tb_tmp where id=2;  

EOF  

exit;  

[root@SZDB ~]# ./shell_call_sql2.sh  

Logging to file '/tmp/temp.log'  

id      val  

1       jack  

2       robin  

3       mark  

Outfile disabled.  

current_date()  

2014-10-14  

id      val  

2       robin  

5、shell脚本中变量输入与输出


[root@SZDB ~]# more shell_call_sql3.sh  

#!/bin/bash  

cmd="select count(*) from tempdb.tb_tmp"  

cnt=$(mysql -uroot -p123456 -s -e "${cmd}")  

echo "Current count is : ${cnt}"  

exit   

[root@SZDB ~]# ./shell_call_sql3.sh   

Warning: Using a password on the command line interface can be insecure.  

Current count is : 3  

  

[root@SZDB ~]# echo "select count(*) from tempdb.tb_tmp"|mysql -uroot -p123456 -s  

3  

  

[root@SZDB ~]# more shell_call_sql4.sh  

#!/bin/bash  

id=1  

cmd="select count(*) from tempdb.tb_tmp where id=${id}"  

cnt=$(mysql -uroot -p123456 -s -e "${cmd}")  

echo "Current count is : ${cnt}"  

exit   

  

[root@SZDB ~]# ./shell_call_sql4.sh   

Current count is : 1  

  

#以上脚本演示中,作抛砖引玉只用,对于输出的结果不是很规整友好,需要进一步改善和提高。