现象描述
访问 Zabbix Web,出现如下错误提示:
• Error in query [SELECT * FROM history_uint h WHERE h.itemid='25067' ORDER BY h.clock DESC LIMIT 1 OFFSET 0] [Table './zabbix/history_uint' is marked as crashed and should be repaired] • Error in query [SELECT * FROM history_uint h WHERE h.itemid='26280' ORDER BY h.clock DESC LIMIT 1 OFFSET 0] [Table './zabbix/history_uint' is marked as crashed and should be repaired] • Error in query [SELECT * FROM history_uint h WHERE h.itemid='26286' ORDER BY h.clock DESC LIMIT 1 OFFSET 0] [Table './zabbix/history_uint' is marked as crashed and should be repaired]
解决办法
1、首先进入mysql命令台:
mysql -u root -p
回车,然后输入密码
2、查询所有的库
mysql> show databases;
3、进入数据库“zabbix”是库名
mysql> use zabbix;
4、检查表
check table history_uint;
(history_uint :出现错误的表)用来检查出现问题的表的状态,出现错误就正常。
mysql> check table history_uint; +---------------------+-------+----------+-------------------------------------------------------+ | Table | Op | Msg_type | Msg_text | +---------------------+-------+----------+-------------------------------------------------------+ | zabbix.history_uint | check | warning | Table is marked as crashed | | zabbix.history_uint | check | warning | 1 client is using or hasn't closed the table properly | | zabbix.history_uint | check | error | record delete-link-chain corrupted | | zabbix.history_uint | check | error | Corrupt | +---------------------+-------+----------+-------------------------------------------------------+ 4 rows in set (2 min 48.75 sec)
5、修复表
repair table history_uint;
mysql> repair table history_uint; +---------------------+--------+----------+--------------------------------------------------+ | Table | Op | Msg_type | Msg_text | +---------------------+--------+----------+--------------------------------------------------+ | zabbix.history_uint | repair | warning | Number of rows changed from 28813609 to 28843608 | | zabbix.history_uint | repair | status | OK | +---------------------+--------+----------+--------------------------------------------------+ 2 rows in set (9 min 12.42 sec)
6、再次检查表。
check table history_uint;
mysql> check table history_uint; +---------------------+-------+----------+----------+ | Table | Op | Msg_type | Msg_text | +---------------------+-------+----------+----------+ | zabbix.history_uint | check | status | OK | +---------------------+-------+----------+----------+ 1 row in set (13.57 sec)
7、ok 搞定。
预防措施
1、一定要备份一次数据库,起码保留了表结构,有些可有可无的数据,可以直接覆盖。
2、重要的数据要经常注意备份,一般一个月左右备份一次。
3、出现此类错误,一般能够解决,经上面的修复方法是比较可行的。但偶尔会有数据丢失的情况,强烈建议先备份。
Windows下使用MyISAM存储引擎碰到此问题的解决方法
首先需要定位到你的mysql的bin目录,里面包含myisamchk.exe文件的目录
./myisamchk -c -r 数据库表MYI文件的路径 (例如:D:/mysql/data/hangban/user.MYI)
如果还不行,就-f 强制修复
D:mysqlin> cd .. 返回上级目录
D:mysql> cd data 进入数据库所在目录(mysql的数据库文件都是放在data这个目录里面的)
D:mysqldata> cd hangban 进入数据库,windows服务器中数据库就是一个文件夹(这里以hangban为例子)
D:mysqldatahangban> myisamchk -r user
- recovering (with sort) MyISAM-table 'user'
Data records: 7216
- Fixing index 1
- Fixing index 2
- Fixing index 3
D:mysqldatahangban>
user为数据库的表名,提示哪个表错误,就修复哪个表。