标签：shell查看系统性能瓶颈 shell监控linux系统
linux服务器敲命令反应慢，网站访问慢，到底什么情况？根据本人的经验，主要原因可能是系统资源到达瓶颈，已经无法处理更多请求。在有监控系统情况下，可以直接通过WEB页面可视化看出是CPU瓶颈？硬盘瓶颈？还是网络瓶颈？如果公司服务器较少或者云服务器，就有可能没有一套监控系统，这时就要登陆到服务器，一条一条的敲命令，查找分析性能瓶颈。命令这么多，咋记得住啊！就算记得住，输入也费劲，于是就有了这个脚本，为了以后自己使用，另外也想分享给博友，学shell朋友能从中得到一丢丢启发。写的比较仓促，内容有点粗略，还望君见谅！
脚本目的：分析系统资源性能瓶颈
脚本功能：
1、查看CPU利用率与负载（top、vmstat、sar）
2、查看磁盘、Inode利用率与I/O负载（df、iostat、iotop、sar）
3、查看内存利用率（free、vmstat）
4、查看TCP连接状态（netstat）
5、查看CPU与内存占用最高的10个进程（top、ps）
6、查看网络流量（ifconfig）
脚本说明：通过一些常用的性能分析工具，计算出我们想知道的信息。
脚本如下：
# cat show_sys_info.sh
#!/bin/bash
#
# blog：lizhenliang.blog.51cto.com
which vmstat &>/dev/null
if [ $? -ne 0 ]; then
echo "vmstat command no found, please install procps package."
exit 1
fi
which iostat &>/dev/null
if [ $? -ne 0 ]; then
echo "iostat command no found, please install sysstat package."
exit 1
fi
if [ $LOGNAME != root ]; then
echo "Please use the root account operation."
exit 1
fi
while true; do
select input in cpu_load disk_load disk_use disk_inode mem_use tcp_status cpu_top10 mem_top10 traffic; do
case $input in
cpu_load)
#CPU利用率与负载
echo "---------------------------------------"
i=1
while [[ $i -le 3 ]]; do
echo -e "\033[32m 参考值${i}\033[0m"
UTIL=`vmstat |awk ‘{if(NR==3)print 100-$15"%"}‘`
USER=`vmstat |awk ‘{if(NR==3)print $13"%"}‘`
SYS=`vmstat |awk ‘{if(NR==3)print $14"%"}‘`
IOWAIT=`vmstat |awk ‘{if(NR==3)print $16"%"}‘`
echo "Util: $UTIL"
echo "User use: $USER"
echo "System use: $SYS"
echo "I/O wait: $IOWAIT"
i=$(($i+1))
sleep 1
done
echo "---------------------------------------"
break
;;
disk_load)
#硬盘I/O负载
echo "---------------------------------------"
i=1
while [[ $i -le 3 ]]; do
echo -e "\033[32m 参考值${i}\033[0m"
UTIL=`iostat -x -k |awk ‘/^s/{OFS=": ";print $1,$NF"%"}‘`
READ=`iostat -x -k |awk ‘/^s/{OFS=": ";print $1,$6"KB"}‘`
WRITE=`iostat -x -k |awk ‘/^s/{OFS=": ";print $1,$7"KB"}‘`
IOWAIT=`vmstat |awk ‘{if(NR==3)print $16"%"}‘`
echo -e "Util:"
echo -e "${UTIL}"
echo -e "I/O Wait: $IOWAIT"
echo -e "Read/s:\n$READ"
echo -e "Write/s:\n$WRITE"
i=$(($i+1))
sleep 1
done
echo "---------------------------------------"
break
;;
disk_use)
#硬盘利用率
DISK_LOG=/tmp/disk_use.tmp
DISK_TOTAL=`fdisk -l |awk ‘/^Disk.*bytes$/&&/\/dev\/sd[a-z]/{printf $2" ";printf "%d",$3;print "GB"}‘`
USE_RATE=`df -h |awk ‘/^\/dev\/sd[a-z0-9]/{print int($5)}‘`
for i in $USE_RATE; do
if [ $i -gt 90 ];then
PART=`df -h |awk ‘{if(int($5)==‘‘‘$i‘‘‘) print $6}‘`
echo "$PART = ${i}%" >> $DISK_LOG
fi
done
echo "---------------------------------------"
echo -e "Disk total:\n${DISK_TOTAL}"
if [ -f $DISK_LOG ]; then
echo "---------------------------------------"
cat $DISK_LOG
echo "---------------------------------------"
rm -f $DISK_LOG
else
echo "---------------------------------------"
echo "Disk use rate no than 90% of the partition."
echo "---------------------------------------"
fi
break
;;
disk_inode)
#硬盘inode利用率
INODE_LOG=/tmp/inode_use.tmp
INODE_USE=`df -i |awk ‘/^\/dev\/sd[a-z][0-9]/{print int($5)}‘`
for i in $INODE_USE; do
if [ $i -gt 90 ]; then
PART=`df -h |awk ‘{if(int($5)==‘‘‘$i‘‘‘) print $6}‘`
echo "$PART = ${i}%" >> $INODE_LOG
fi
done
if [ -f $INODE_LOG ]; then
echo "---------------------------------------"
cat $INODE_LOG
echo "---------------------------------------"
rm -f $INODE_LOG
else
echo "---------------------------------------"
echo "Inode use rate no than 90% of the partition."
echo "---------------------------------------"
fi
break
;;
mem_use)
#内存利用率
echo "---------------------------------------"
MEM_TOTAL=`free -m |awk ‘{if(NR==2)printf "%.1f",$2/1024}END{print "G"}‘`
USE=`free -m |awk ‘{if(NR==3) printf "%.1f",$3/1024}END{print "G"}‘`
FREE=`free -m |awk ‘{if(NR==3) printf "%.1f",$4/1024}END{print "G"}‘`
CACHE=`free -m |awk ‘{if(NR==2) printf "%.1f",($6+$7)/1024}END{print "G"}‘`
echo -e "Total: $MEM_TOTAL"
echo -e "Use: $USE"
echo -e "Free: $FREE"
echo -e "Cache: $CACHE"
echo "---------------------------------------"
break
;;
tcp_status)
#网络连接状态
echo "---------------------------------------"
COUNT=`netstat -antp |awk ‘{status[$6]++}END{for(i in status) print i,status[i]}‘`
echo -e "TCP connection status:\n$COUNT"
echo "---------------------------------------"
;;
cpu_top10)
#占用CPU高的前10个进程
echo "---------------------------------------"
CPU_LOG=/tmp/cpu_top.tmp
i=1
while [[ $i -le 3 ]]; do
echo -e "\033[32m 参考值${i}\033[0m"
ps aux |awk ‘{if($3>0.1)print "CPU: "$3"% -->",$NF|"sort -k2 -nr |head -n 10"}‘ > $CPU_LOG
cat $CPU_LOG
i=$(($i+1))
sleep 1
done
echo "---------------------------------------"
break
;;
mem_top10)
#占用内存高的前10个进程
echo "---------------------------------------"
MEM_LOG=/tmp/mem_top.tmp
i=1
while [[ $i -le 3 ]]; do
echo -e "\033[32m 参考值${i}\033[0m"
ps aux |awk ‘{if($4>0.1)print "CPU: "$4"% -->",$NF|"sort -k2 -nr |head -n 10"}‘ > $MEM_LOG
cat $MEM_LOG
i=$(($i+1))
sleep 1
done
echo "---------------------------------------"
break
;;
traffic)
#查看网络流量
while true; do
read -p "Please enter the network card name(eth[0-9] or em[0-9]): " eth
if [[ $eth =~ ^eth[0-9]$ ]] || [[ $eth =~ ^em[0-9]$ ]]; then
break
else
echo "Input error,please input again."
fi
done
echo "---------------------------------------"
echo -e " In ------ Out"
i=1
while [[ $i -le 3 ]]; do
OLD_IN=`ifconfig $eth |awk ‘/RX bytes/{print $2}‘ |cut -d: -f2`
OLD_OUT=`ifconfig $eth |awk ‘/RX bytes/{print $6}‘ |cut -d: -f2`
sleep 1
NEW_IN=`ifconfig $eth |awk ‘/RX bytes/{print $2}‘ |cut -d: -f2`
NEW_OUT=`ifconfig $eth |awk ‘/RX bytes/{print $6}‘ |cut -d: -f2`
IN=`awk ‘BEGIN{printf "%.1f\n",‘$((${NEW_IN}-${OLD_IN}))‘/1024/128}‘`
OUT=`awk ‘BEGIN{printf "%.1f\n",‘$((${NEW_OUT}-${OLD_OUT}))‘/1024/128}‘`
echo "${IN}MB/s ${OUT}MB/s"
i=$(($i+1))
sleep 1
done
echo "---------------------------------------"
break
;;
*)
echo "---------------------------------------"
echo "Please enter the number."
echo "---------------------------------------"
break
;;
esac
done
done运行效果如下：
# sh show_sys_info.sh
本文出自 “李振良的技术博客” 博客，请务必保留此出处http://lizhenliang.blog.51cto.com/7876557/1687612
标签：shell查看系统性能瓶颈 shell监控linux系统
原文地址：http://lizhenliang.blog.51cto.com/7876557/1687612
