方法一、LR + SiteScope/nmon方法二、使用rstatd包1、下載rpc.rstatd-4.0.1.tar.gz2、解壓縮tar -zxvf rpc.rstatd-4.0.1.tar.gz3、配置./configure4、編譯make5、安裝make install6、啟動rpc.rstatd7、在LoadRunner中添加計數(shù)器average load :在過去的1分鐘，的平均負(fù)載cpu utilization: cpu的使用率disk traffic: disk傳輸率paging rate： 每秒從磁盤讀到物理內(nèi)存，或者從物理內(nèi)存寫到頁面文件的內(nèi)存頁數(shù)Swap-in rate: 每秒交換到內(nèi)存的進(jìn)程數(shù)Swap-out rate: 每秒從內(nèi)存交換出來的進(jìn)程8、將服務(wù)設(shè)置為自動啟動（Linux啟動時自動啟動這些服務(wù)）：vi /etc/rc.d/rc.local如：#rpc.rstatd--絕對路徑/usr/local/sbin/rpc.rstatd方法3、使用Shell腳本使用Shell腳本收集Linux資源寫入csv文件，再通過LR的Analysis導(dǎo)入csv文件（Tools - Extenal Monitors - Import Data... ）Shell腳本如下所示：#!/bin/bash # (C) 2006 Mark Boddington, http://www.badpenguin.co.uk # Licensed under the GNU GPL Version 2.# ***** Version 0.2 ***** # TODO -- Create CSV parsing rules for the netstat.# ***** Configuration ***** # set LOG to the directory you want to write the performance data to. # set SLEEP to the number of seconds you want to sleep between samples # set HDD to the number of had disks in your machine.LOG=/home/mark/PerfMon/LIVE SLEEP=10 HDD=2HTYPE=$(uname -s)genStat() {now=$( date +%S )while [ "$now" -ne "30" ]dosleep 1now=$( date +%S )donewhile :;dodat=$(date +%Y%m%d,%H:%M:%S)day=$(date +%Y%m%d )iostat -x 1 2 | sed -e"s/^/(.*/)/$dat /1/" | grep "[0-9]/." | tail -${HDD} >> ${LOG}/io.${day}.log &vmstat 1 2 | awk "{ print /"$dat/", /$0 }" | tail -1 >> ${LOG}/vm.${day}.log &netstat -i | grep -v Iface | awk "{ print /"$dat/", /$0 }" >> ${LOG}/netstat.${day}.log &uptime >> ${LOG}/uptime.${day}.log &sleep $SLEEPdone }mkcsv() {dat=$1if [ "$HTYPE" == "SunOS" ]then#IO CSVecho date,time,device,r/s,w/s,kr/s,kw/s,wait,actv,wsvc_t,asvc_t,%w,%b > ${LOG}/io.${dat}.csvcat ${LOG}/io.${dat}.log | egrep -v "extended|device" | awk '{ OFS=","; print $1,$12,$2,$3,$4,$5,$6,$7,$8,$9,$10,$11 }' >> ${LOG}/io.${dat}.csv#VM csvecho date,time,k[r],k[b],k[w],swap,free,pg[re],pg[mf],pg[pi],pg[po],pg[fr],pg[de],pg[sr],m0,m1,m2,m1,interupt,syscall,ctxswt,cpu[us],cpu[sys],cpu[idl] > ${LOG}/vm.${dat}.csvcat ${LOG}/vm.${dat}.log | awk '{for(l=1;l<23;l++) { printf("%s,", $l) }; print $23 }' >> ${LOG}/vm.${dat}.csv#uptime csvecho time,users,5min,10min,15min > ${LOG}/uptime.${dat}.csvcat ${LOG}/uptime.${dat}.log | awk '{ OFS=","; if ( $6 ~ /^[hm][ri]/) { print $1,$7,$11$12$13 } else if ( $6 ~/^user/) { print $1,$5,$9$10$11} else { print $1,$6,$10$11$12} }' >> ${LOG}/uptime.${dat}.csvelif [ "$HTYPE" == "Linux" ]then#IO CSVecho "date,time,device,rrqm/s,wrqm/s,r/s,w/s,rsec/s,wsec/s,rkB/s,wkB/s,avgrq-sz,avgqu-sz,await,svctm,%util" > ${LOG}/io.${dat}.csvcat ${LOG}/io.${dat}.log | egrep -v "extended|device" | awk '{for(l=1;l<15;l++) { printf("%s,", $l) }; print $15}' >> ${LOG}/io.${dat}.csv#VM csvecho date,time,r,b,swp,free,buff,cache,si,so,bi,bo,in,cs,us,sy,id,wa > ${LOG}/vm.${dat}.csvcat ${LOG}/vm.${dat}.log | awk '{for(l=1;l<17;l++) { printf("%s,", $l) }; print $17 }' >> ${LOG}/vm.${dat}.csv#uptime csvecho time,users,5min,10min,15min > ${LOG}/uptime.${dat}.csvcat ${LOG}/uptime.${dat}.log | awk '{ OFS=","; if ( $4 ~ /^min/) { print $1,$5,$9$10$11 } else { print $1,$4,$8$9$10} }' >> ${LOG}/uptime.${dat}.csvelseecho "Hmmm - An unexpected error occured. Have you change the host type?"fi}if [ "$HTYPE" != "SunOS" -a "$HTYPE" != "Linux" ] thenecho "Error - This script has no knowlege of the System $HTYPE"echo " You will need to do some tweaking."exit ficase $1 in run)genStat;;csv)if [ $# -lt 2 ]thenecho "Error - You must supply a date in the form YYYYMMDD"exitfimkcsv $2;;*)echo -e ":::: Usage ::::"echo -e "$0 run : Collect stats"echo -e "$0 csv YYYYMMDD : Generate CSV from stats"echo "";; esac

?

總結(jié)

以上是生活随笔為你收集整理的LoadRunner监控Linux的三种方法的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯，歡迎將生活随笔推薦給好友。