那里有监控主机性能（内存、i/o、cpu）使用情况的脚本？谢谢！

novazhao

那里有监控主机性能（内存、i/o、cpu）使用情况的脚本？谢谢！
每个月要生成日志报表！。

WSRF

vmstat, iostat, sar + perl, 你可以自己来写，可以针对你自己的业务和主机使用情况来做出很好的统计和报表。

zly0216

有没有现成的啊？谢谢，

peakdaifeng

#! /bin/sh
Month=`date "+%m"`
Day=`date "+%d"`
Year=`date "+%y"`
YEAR=`date "+%Y"`
Num=5

LogDir=/tmp/peak/TT

CPULog=$LogDir/CPU_MEM_$YEAR$Month$Day.txt
CRITICAL=$LogDir/CRITICAL_CPU_MEM_$YEAR$Month$Day.txt
WARNING=$LogDir/WARNING_CPU_MEM_$YEAR$Month$Day.txt

#####Monitor CPU & Memory#####
vmstat -t 2 $Num >;>; $CPULog

CNT=`tail -$Num $CPULog|awk '$1 ~ /^[0-9]/ {if($14+$15>;94 || $16<6){print $0}}' | wc -l`

if [ $CNT -gt 0 ] ; then
  echo "***********CRITICAL***********" >;>; $CRITICAL
  date >;>; $CRITICAL
  head -3 $CPULog >;>; $CRITICAL
  tail -$Num $CPULog|awk '$1 ~ /^[0-9]/ {if($14+$15>;94 || $16<6){print $0}}' >;>; $CRITICAL
fi

CNT=`tail -$Num $CPULog |awk '$1 ~ /^[0-9]/ {if($6+$7>;0){print $0}}' | wc -l`

if [ $CNT -gt 0 ] ; then
  echo "***********WARNING***********" >;>; $WARNING
  date >;>; $WARNING
  head -3 $CPULog >;>; $WARNING
  tail -$Num $CPULog |awk '$1 ~ /^[0-9]/ {if($6+$7>;0){print $0}}' >;>; $WARNING
fi


#####Monitor HW ERROR#####
CNT=`errpt -d H -s $Month$Day"0000"$Year |wc -l`

if [ $CNT -gt 0 ] ; then
  HWERR=$LogDir/HW_ERROR_$YEAR$Month$Day.txt
  echo "***********HW_ERROR***********" >;>; $HWERR
  date >;>; $HWERR
  errpt -d H -s $Month$Day"0000"$Year >;>; $HWERR
fi


#####Monitor Memory#####
lsps -a >;>; $CPULog

CNT=`tail -1 $CPULog |awk '{if($5>;20){print $0}}'| wc -l`

if [ $CNT -gt 0 ] ; then
  echo "***********CRITICAL***********" >;>; $CRITICAL
  date >;>; $CRITICAL
  tail -2 $CPULog >;>; $CRITICAL
fi

svmon -G >;>; $CPULog

CNT=`tail -7 $CPULog |grep memory |awk '{if($3/$2*100>;80){print $0}}'| wc -l`

if [ $CNT -gt 0 ] ; then
  echo "***********CRITICAL***********" >;>; $CRITICAL
  date >;>; $CRITICAL
  tail -7 $CPULog >;>; $CRITICAL
fi


你自己试一下，有些地方可能要更改

hot2ice

可以使用nmon工具，省去写脚本的麻烦了，并且可以通过excel出图形分析：

For Data-Collect-Mode = spreadsheet format (comma separated values)
        Note: use only one of f,F,z,x or X and make it the first argument
        -f            spreadsheet output format [note: default -s300 -c288]
                         output file is <hostname>;_YYYYMMDD_HHMM.nmon
        -F <filename>; same as -f but user supplied filename
        -r <runname>;  goes into spreadsheet file [default hostname]
        -t            include top processes in the output
        -T            as -t plus saves command line arguments in UARG section
        -s <seconds>;  between snap shots
        -c <number>;   of refreshes
        -l <dpl>;      disks/line default 150 to avoid spreadsheet issues. EMC=64
.
        -g <filename>; User decided Disk Groups (see above -g)
        -D            Skip disk configuration sections
        -E            Skip ESS  configuration sections
        -W            Include WLM sections
        -I <percent>;  Ignore process percent threshold (default 0.1)
                      don't save TOP stats if proc using less CPU than this %
        -A            Include Async I/O Section
        -m <dir>;      nmon changes to this directory before saving data to a fil
e
        example: collect for 1 hour at 30 second intervals with top procs
                 nmon64 -f -t -r Test1 -s30 -c120
        To load into a spreadsheet like Lotus 1-2-3:
        sort -A *nmon >;stats.csv
        transfer the stats.csv file to your PC
        Start 1-2-3 and then Open <char-separated-value ASCII file>;
Capacity planning mode - use cron to run each day
        -x            sensible spreadsheet output for CP =  one day
                      every 15 mins for 1 day ( i.e. -ft -s 900 -c 96)
        -X            sensible spreadsheet output for CP = busy hour
                      every 30 secs for 1 hour ( i.e. -ft -s 30 -c 120)