Monitor whether the master-slave synchronization of MySQL is abnormal. If not, send SMS or email to the administrator. 1) develop a daemons script to detect every 30 seconds.
2) if the following error number (11581159100810071062) occurs during synchronization, the error will be skipped. Stage 3:
3) if the IO and SQL threads are abnormal, notify the administrator by email
Use array technology to implement the above script (get master-slave judgment and error number part)
The master-slave monitoring script is as follows:
#!/bin/bash CODE=( 1158 1159 1008 1007 1062 2003 ) fun_Base(){ #1. Define variables #1.1 capture IP address IP=`ifconfig eth0 |awk 'NR==2'` #1.2 get slave IO and SQL status, Err code My_SQL=`mysql -e "show slave status\G" |egrep "SQL_Running:" |awk ''` My_IO=`mysql -e "show slave status\G" |egrep "IO_Running:" |awk ''` My_CODE=`mysql -e "show slave status\G" |egrep "Last_IO_Errno:" |awk ''` #1.3 timing time variable Time=`date +%F-%H:%M:%S` #1.4 define log directory DIR=/tmp/slave_$ Status_Log=$DIR/slave_status_$.log Check_log=$DIR/slave_check_$.log Erro_log=$DIR/slave_err_$.log #1.5 define email Total="$IP slave status $Time" Mail_Rec="[email protected]" #2. Save the state of slave to the log file [ -d $DIR ] || mkdir $DIR -p mysql -e "show slave status\G" >$Status_Log } #3. Error code to judge the slave status fun_Status(){ RETVAL=0 for ((i=0;i<${#CODE[*]};i++)) do if [ $My_CODE -eq $ ];then mysql -e "stop slave;" && RETVAL=$? [ $RETVAL -eq 0 ] && mysql -e "SET GLOBAL SQL_SLAVE_SKIP_COUNTER = 1;" && RETVAL=$? [ $RETVAL -eq 0 ] && mysql -e "start slave;" && RETVAL=$? [ $RETVAL -eq 0 ] && echo "slave errno code is successful." >$Erro_log [ $RETVAL -eq 0 ] && mail -s "$Total" $Mail_Rec <$Status_Log && mail -s "$Total" $Mail_Rec <$Erro_log fi done } #4. Judge whether IO and SQL threads are normal fun_Check(){ if [ "$My_SQL" == "Yes" -a "$My_IO" == "Yes" ];then echo "slave status is successful." echo "slave status is successful." >$Check_log mail -s "$Total" $Mail_Rec <$Status_Log mail -s "$Total" $Mail_Rec <$Check_log else echo "slave status is failed." echo "slave status is failed." >>$Check_log mail -s "$Total" $Mail_Rec <$Status_Log mail -s "$Total" $Mail_Rec <$Check_log fi } #5. Main function main(){ while true do fun_Base fun_Status fun_Check sleep 20 done } main