Linux服务器监控

1、下载SendEmail；

2、建立监控脚本：

[oracle@db1 scripts]$ cat check_system_warning.sh
#!/bin/sh
####################################
# 2013.05.07 by zhanghui #
####################################
#定义变量
v_datetime=`date +"%Y-%m-%d %H:%M:%S"`
v_username='oracle'
v_hostname=`hostname`
#获取物理CPU数量
v_physicalcpu=`cat /proc/cpuinfo | grep "physical id" | sort | uniq | wc -l`
#获取逻辑CPU数量
v_logicalcpu=`cat /proc/cpuinfo | grep "processor" | wc -l`
#获取IP地址
v_ip=`/sbin/ifconfig bond0|grep 'inet addr'|awk '{print $2}'|cut -d: -f2`
#发送者邮箱
v_sendmail=''
#发送者邮箱用户名
v_mailname=''
#发送者邮箱密码
v_password=''
#smtp服务器IP或域名：smtp.163.com
v_smtpserver='113.108.212.80'
#接收者邮件，如果是多个用空间分开
v_receivemail=' '
#路径
v_path="/home/${v_username}/scripts"
#设置获取top进程的数量
v_process=15
#系统负载监控，最近5分钟负载及Cpu Idle报警阀值
v_loadvaluesfor5min_warning=20
v_cpuusage_warning=35
#剩余物理内存报警阀值，单位M
v_freememory_warning=512
#free+buffers+cached报警阀值，单位M
v_free_buffers_cached_warning=2000
#使用swap分区报警阀值，单位M
v_useswap_warning=1000
#磁盘使用百分比报警阀值
v_diskusepercent_warning=90
#系统负载监控代码开始
#check sys load and cpuidle
#v_release=`lsb_release -a|head -4|tail -1|awk '{print $2}'|awk -F '.' '{print $1}'`
v_release=6
if [ ${v_release} = 4 ]; then
#RHEL 4.x
v_cpuusage=`mpstat 1 5 | grep Average | awk '{print $9}'|awk -F "." '{print $1}'`
#RHEL 5.x
elif [ ${v_release} = 5 ]; then
v_cpuusage=`mpstat 1 5 | grep Average | awk '{print $10}'|awk -F "." '{print $1}'`
#RHEL 6.x
elif [ ${v_release} = 6 ]; then
v_cpuusage=`mpstat 1 5 | grep Average | awk '{print $11}'|awk -F "." '{print $1}'`
fi;
v_loadvaluesfor5min=`cat /proc/loadavg |awk '{print $1}'|awk -F "." '{print $1}'`
if [ ${v_loadvaluesfor5min} -ge ${v_loadvaluesfor5min_warning} ] || [ ${v_cpuusage} -le ${v_cpuusage_warning} ] ; then
echo "############################" > ${v_path}/top${v_process}_text.log
echo "USER: ${v_username}" >> ${v_path}/top${v_process}_text.log
echo "HOSTNAME: ${v_hostname}" >> ${v_path}/top${v_process}_text.log
echo "IP: ${v_ip}" >> ${v_path}/top${v_process}_text.log
echo "Physical CPU: ${v_physicalcpu}" >> ${v_path}/top${v_process}_text.log
echo "Logical CPU: ${v_logicalcpu}" >> ${v_path}/top${v_process}_text.log
echo "Uptime: "`uptime` >> ${v_path}/top${v_process}_text.log
echo "############################" >> ${v_path}/top${v_process}_text.log
echo "########mpstat 1 5#########" >> ${v_path}/top${v_process}_text.log
mpstat 1 5 >> ${v_path}/top${v_process}_text.log
echo "######## ps -ef|grep "LOCAL=NO"|wc -l #########" >> ${v_path}/top${v_process}_text.log
ps -ef|grep "LOCAL=NO"|wc -l >> ${v_path}/top${v_process}_text.log
echo "######## ps aux | ??? #########" >> ${v_path}/top${v_process}_text.log
ps aux|head -1 >> ${v_path}/top${v_process}_text.log
ps aux|grep -v PID|sort -rn -k +3|head -${v_process} >> ${v_path}/top${v_process}_text.log
#####echo " " | ${v_path}/bsmtp -f ${v_sendmail} -h smtp.api.localdomain -s "${v_hostname} Five Minutes Loadavg: ${v_loadvaluesfor5min} ; CPU idle: ${v_cpuusage} - ${v_datetime}" ${v_receivemail} < ${v_path}/top${v_process}_text.log
${v_path}/sendEmail -f ${v_sendmail} -t ${v_receivemail} -s ${v_smtpserver} -u "${v_hostname} Five Minutes Loadavg: ${v_loadvaluesfor5min} ; CPU idle: ${v_cpuusage} - ${v_datetime}" -xu ${v_mailname} -xp ${v_password} -o message-content-type=text -o message-charset=UTF-8 message-file=${v_path}/top${v_process}_text.log -l ${v_path}/system_warning_sendmail.log
fi;
#系统负载监控代码结束
                                                             
#系统内存监控代码开始
#check memory space
v_freememory=`free -m|sed -n '\ : p'|gawk '{print $4}'|head -1`
v_freebuffers=`free -m|sed -n '\ : p'|gawk '{print $6}'|head -1`
v_freecached=`free -m|sed -n '\ : p'|gawk '{print $7}'|head -1`
v_free_buffers_cached_total=`echo "($v_freememory+$v_freebuffers+$v_freecached)" |bc`
v_useswap=`free -m|sed -n '\ : p'|tail -1|cut -d ":" -f 2|gawk '{print $2}'`
if [ ${v_freememory} -le ${v_freememory_warning} ] || [ ${v_useswap} -ge ${v_useswap_warning} ] || [ ${v_free_buffers_cached_total} -le ${v_free_buffers_cached_warning} ] ; then
echo "############################" > ${v_path}/memory_text.log
echo "USER: ${v_username}" >> ${v_path}/memory_text.log
echo "HOSTNAME: ${v_hostname}" >> ${v_path}/memory_text.log
echo "IP: ${v_ip}" >> ${v_path}/memory_text.log
echo "############################" >> ${v_path}/memory_text.log
free -m >> ${v_path}/memory_text.log
echo "########vmstat 1 10#########" >> ${v_path}/memory_text.log
vmstat 1 10 >> ${v_path}/memory_text.log
#####${v_path}/bsmtp -f ${v_sendmail} -h smtp.api.localdomain -s "${v_hostname} MEMORY SPACE WARNING! - ${v_datetime}" ${v_receivemail} < ${v_path}/memory_text.log
${v_path}/sendEmail -f ${v_sendmail} -t ${v_receivemail} -s ${v_smtpserver} -u "${v_hostname} MEMORY SPACE WARNING - ${v_datetime}" -xu ${v_mailname} -xp ${v_password} -o message-content-type=text -o message-charset=UTF-8 message-file=${v_path}/memory_text.log -l ${v_path}/system_warning_sendmail.log
fi;
#系统内存监控代码结束
                                                             
#系统磁盘监控代码开始
#check disk space
v_diskusepercent=`df -P|column -t| sed -n '\ / p'|gawk '{print $5}'|sed s/%//|sort -nr|head -1`
if [ ${v_diskusepercent} -ge ${v_diskusepercent_warning} ] ; then
echo "############################" > ${v_path}/disk_text.log
echo "USER: ${v_username}" >> ${v_path}/disk_text.log
echo "HOSTNAME: ${v_hostname}" >> ${v_path}/disk_text.log
echo "IP: ${v_ip}" >> ${v_path}/disk_text.log
echo "############################" >> ${v_path}/disk_text.log
df -h >> ${v_path}/disk_text.log
#####${v_path}/bsmtp -f ${v_sendmail} -h smtp.api.localdomain -s "${v_hostname} DISK SPACE WARNING!!! - ${v_datetime}" ${v_receivemail} < ${v_path}/disk_text.log
${v_path}/sendEmail -f ${v_sendmail} -t ${v_receivemail} -s ${v_smtpserver} -u "${v_hostname} DISK SPACE WARNING - ${v_datetime}" -xu ${v_mailname} -xp ${v_password} -o message-content-type=text -o message-charset=UTF-8 message-file=${v_path}/disk_text.log -l ${v_path}/system_warning_sendmail.log
fi;
#系统磁盘监控代码结束

3、配置crontab，每两分钟运行一次

[oracle@db1 scripts]$ crontab -l
#os check
*/2 * * * * /home/oracle/scripts/check_system_warning.sh >/dev/null 2>&1