如果你是linux服务器且需要报警可以参考我的
[code="java"]#!/bin/sh

mailAddresses=(13466789569@139.com);
function sendMail() {
for address in ${mailAddresses[*]}; do
echo $1 | mutt -s $1 13466789569@139.com $address;
done;

}

#服务不可用监控

tomcatSSonlinePid=ps aux | grep tomcat-ssonline | grep -v grep | grep -v retomcat | awk '{print $2}' ;
echo "tomcat ssonline pid=" $tomcatSSonlinePid;
if [ -z "$tomcatSSonlinePid" ]
then
sh /usr/program/scripts/startSsonline.sh &
sendMail "ssonline alert-server tomcat-ssonline not start，starting！";
fi

tomcatForumPid=ps aux | grep tomcat-forum | grep -v grep | grep -v tail | grep -v retomcat | awk '{print $2}' ;
echo "tomcat forum pid=" $tomcatForumPid;
if [ -z "$tomcatForumPid" ]
then
sh /usr/program/scripts/startForum.sh &
sendMail "ssonline alert-server tomcat-forum not start，starting！";
fi

tomcatV1Pid=ps aux | grep tomcat-v1 | grep -v grep | grep -v tail | grep -v tail | grep -v retomcat | awk '{print $2}' ;
echo "tomcat v1 pid=" $tomcatV1Pid;
if [ -z "$tomcatV1Pid" ]
then
sh /usr/program/scripts/startV1.sh &
sendMail "ssonline alert-server tomcat-v1 not start，starting！";
fi

jamesPid=ps aux | grep james | grep -v grep | grep -v tail | grep -v retomcat | awk '{print $2}' ;
echo "james pid=" $jamesPid;
if [ -z "$jamesPid" ]
then
sh /usr/program/scripts/startJames.sh &
sendMail "ssonline alert-server tomcat-james not start，starting！";
fi

mysqlPid=ps aux | grep libexec/mysqld | grep -v grep | grep -v tail | grep -v retomcat | awk '{print $2}' ;
echo "mysql pid=" $mysqlPid;
if [ -z "$mysqlPid" ]
then
/usr/local/mysql/bin/mysqld_safe &
sendMail "ssonline alert-mysql not start，starting！";
fi

#服务异常监控

wget http://localhost:8080/monitor/jvm/status.jsp -O status;
ssonlineStatus=cat ./status | tr '\n' ' ' | tr '\r' ' ' | awk 'gsub(/^ *| *$/,"")';
echo "ssonline status";
echo $ssonlineStatus;

if [ -n v1Status -a "200" = "$ssonlineStatus" ]
then
echo "ssonline ok";
else
sendMail "ssonline alert-tomcat-ssonline status warn!";
fi

wget http://localhost:9080/forum/monitor/jvm/status.jsp -O status;

forumStatus=cat ./status | tr '\n' ' ' | tr '\r' ' ' | awk 'gsub(/^ *| *$/,"")';
echo "forum status";
echo $forumStatus;

if [ -n v1Status -a "200" = "$forumStatus" ]
then
echo "forum ok";
else
sendMail "ssonline alert-tomcat-forum status warn!";
fi

wget http://localhost:10080/v1/monitor/jvm/status.jsp -O status;
v1Status=cat ./status | tr '\n' ' ' | tr '\r' ' ' | awk 'gsub(/^ *| *$/,"")';
echo "v1 status";
echo $v1Status;

if [ -n v1Status -a "200" = "$v1Status" ]
then
echo "v1 ok";
else
sendMail "ssonline alert-tomcat-v1 status warn!";
fi

wget http://localhost/monitor/status_db.jsp -O status;
dbStatus=cat ./status | tr '\n' ' ' | tr '\r' ' ' | awk 'gsub(/^ *| *$/,"")';
echo "mysql status";
echo $dbStatus;

if [ -n v1Status -a "200" = "$dbStatus" ]
then
echo "mysql ok";
else
sendMail "ssonline alert-mysql status warn!";
fi

#端口异常监控
ports=(80 8080 9080 10080 25 110 11211 3306);

for port in ${ports[*]}; do
status=netstat -ntlp | grep :$port;
echo "check port " $port " " $status;

    if [ -z "$status" ]
then
    echo "ssonline alert-port$port warn!";
    sendMail "ssonline alert-port$port warn!";   
fi

done;

#cpu 磁盘内存异常

eth="eth0 eth1 "
function get_eth_info()
{
name=$1
cat /proc/net/dev | awk -F ':' '{if(NR>2)print $1" "$2" "$9" "$10}' | awk '{print$1" "$2" "$3" "$5}' | while \
read line
do

echo $line

data=echo $name" "$line | awk '{if($1==$2)print $3" "$4" "$5}'
if ! [ -z "$data" ]
then
echo $data
fi
done
}
function sum_eth()
{
rm -fr 12321.txt
for name in $eth
do
ret=get_eth_info $name
#echo $ret | awk '{bps+=$0;pps+$1;dps+=$2;print bps" "pps" "dps}'
echo $ret >> 12321.txt
done
cat 12321.txt | awk 'BEGIN{bps=0;pps=0;dps=0}{bps+=$1;pps+=$2;dps+=$3}END{print bps" "pps" "dps}'
}
flow=sum_eth
disk=df | awk 'BEGIN{total=0;avl=0;used=0;}NR > 1{total+=$2;used+=$3;avl+=$4;}END{printf"%d", avl/total*100}'
#top -b -n 1 | grep -w Mem | awk '{print"tot_mem: "$2" used_mem: " $4 " free_mem: "$6" left:"$6/$2"%"}'
mem=top -b -n 1 | grep -w Mem | awk '{printf"%d",$6/$2*100}'
#top -b -n 1 | grep -w Cpu | awk '{print"cpu: "$5}' | awk -F '%' {print$1}
cpu=top -b -n 1 | grep -w Cpu | awk '{print$5}' | awk -F '%' '{printf"%d",$1}'
tm=date +%s
if ! [ -f "flow.txt" ]
then
echo $tm" "$flow >flow.txt
sleep 1
tm=date +%s
flow=sum_eth
fi
old_flow=cat flow.txt
new_flow=echo $tm $flow
echo $new_flow >flow.txt
#echo $old_flow
#echo $new_flow
final_flow=echo $old_flow $new_flow | awk '{dif_tm=$5-$1;dif_byte=$6-$2;dif_pkt=$7-$3;dif_dpkt=$8-$4;printf"%d %d %d",dif_byte*8/dif_tm, dif_pkt/dif_tm, dif_dpkt/dif_tm}'
#final_flow=echo $old_flow $new_flow | awk '{print$0}'

echo "cpu "$cpu
echo "disk "$disk
echo "mem "$mem
echo "nic "$final_flow
if [ $cpu -le 30 ]
then
echo "ssonline alert-cpu unused $cpu warn!";
sendMail "ssonline alert-cpu unused $cpu warn!";

fi

if [ $disk -le 30 ]
then
echo "ssonline alert-disk unused $disk warn!";
sendMail "ssonline alert-disk unused $disk warn!";

fi

#if [ $mem -le 0 ]
#then

echo "ssonline alert-mem unused $mem warn!";

sendMail "ssonline alert-mem unused $mem warn!";

#fi[/code]

如果你要纯走java 单机可以考虑
1、jmx获取服务器的一些信息如堆内存等可以使用sigar收集单机信息
2、数据库可以查看通过数据库api获取连接数等
3、发email可以通过javamail
4、发短信需要第三方提供这个之前公司有自己的一套东西

如果监控多服务器
snmp等

第三方现成的
munin（我目前有这个只监控）
nagios(这个比较强大能报警)

lsioui

ssh如何开发服务监控程序

4条回答

为你推荐