用snmp和mrtg监控linux server

 

上回看了platinum的"mrtg能做些什么"贴子，很受启发，于是也试着做了一个。但当时对snmpd不是很熟，所以没有做成功，后来仔细看了一下 snmd的有关文章和RH中的/etc/snmp/snmpd.conf文件，发现用snmp+mrtg可以很好的实现对局域网内服务器状态的监控。
现在就以用snmp+mrtg监控一台局域网内的redhat机器（IP：192.168.13.103）的网卡、内存、CPU、DISKIO为例子，谈一下如何实现。基本的法办就是用一台redhat监控机器（IP：192.168.13.105），通过snmpwalk命令去抓目标服务器的状态数据，然后用mrtg画出图来。
1、首先我们要把目标snmpd.conf文件的配好。这是用snmpwalk命令一抓取数据的关健。下面是目标机器（IP： 192.168.13.103）上的/etc/snmp/snmpd.conf文件部份内容，红色的部份是我对snmpd.conf所做的改动。
[root@wy1 root]# cat /etc/snmp/snmpd.conf

####
# First, map the community name "public" into a "security name"

#    sec.name   source       community
com2sec notConfigUser   default    public   #定义community名称为 public，映射到安全名 notConfigUser。

####
# Second, map the security name into a group name:

#    groupName    securityModel securityName
group notConfigGroup v1           notConfigUser #定义安全用户名notConfigUser映射到notConfigGroup组。
group notConfigGroup v2c           notConfigUser

####
# Third, create a view for us to let the group have rights to: #定义一个view,来决定notConfigUser可以操作的范围。


# Make at least   snmpwalk -v 1 localhost -c public system fast again. #定义可查看的snmp的范围。
#    name           incl/excl     subtree       mask(optional)
view systemview included .1.3.6.1.2.1.1
view systemview included .1.3.6.1.2.1.25.1.1
view all included   .1
####
# Finally, grant the group read-only access to the systemview view. #给notConfigGroup组所定义view名 all 以只读权限。

#    group       context sec.model sec.level prefix read write   notif
access   notConfigGroup ""    any    noauth exact   all   none none#access   notConfigGroup ""    any    noauth exact   mib2 none none

# -----------------------------------------------------------------------------

# Here is a commented out example configuration that allows less
# restrictive access.

# YOU SHOULD CHANGE THE "COMMUNITY" TOKEN BELOW TO A NEW KEYWORD ONLY
# KNOWN AT YOUR SITE.   YOU *MUST* CHANGE THE NETWORK TOKEN BELOW TO
# SOMETHING REFLECTING YOUR LOCAL NETWORK ADDRESS SPACE.

##    sec.name   source       community
#com2sec local     localhost    COMMUNITY
#com2sec mynetwork NETWORK/24    COMMUNITY

##     group.name sec.model   sec.name
#group MyRWGroup   any        local
#group MyROGroup   any        mynetwork
#
#group MyRWGroup   any        otherv3user
#...

##           incl/excl subtree                          mask
#view all included   .1                            80

## -or just the mib2 tree-

#view mib2 included   .iso.org.dod.internet.mgmt.mib-2 fc
#view mib2 included   .iso.org.dod.internet.mgmt.mib-2 fc


##             context sec.model sec.level prefix read write   notif
#access MyROGroup ""    any    noauth 0    all none none
#access MyRWGroup ""    any    noauth 0    all all all


其实配制一个snmpd.conf文件不算太难，
（1）首选是定义一个共同体名(community)，这里是public，及可以访问这个public的用户名（sec name），这里是notConfigUser。Public相当于用户notConfigUser的密码：）
#    sec.name   source       community
com2sec notConfigUser   default    public
（2）定义一个组名（groupName）这里是notConfigGroup，及组的安全级别，把notConfigGroup这个用户加到这个组中。
         groupName    securityModel securityName
group notConfigGroup v1           notConfigUser
group notConfigGroup v2c           notConfigUser
（3）定义一个可操作的范围(view)名， 这里是all，范围是 .1
#    name           incl/excl     subtree       mask(optional)
   view   all          included     .1
   （4）定义notConfigUser这个组在all这个view范围内可做的操作，这时定义了notConfigUser组的成员可对.1这个范围做只读操作。
    #    group       context sec.model sec.level prefix read write   notif
access   notConfigGroup ""    any    noauth exact   all   none none

ok,这样我们的snmpd.conf文件就基本配成了，用service   snmpd   restart重启snmpd服务。现在我们做一个测试，在监控机上打下面的命令：
   [root@wy2 root]# snmpwalk -v 1 192.168.13.103 -c public system
SNMPv2-MIB::sysDescr.0 = STRING: Linux wy1 2.4.20-8smp #1 SMP Thu Mar 13 17:45:54 EST 2003 i686
SNMPv2-MIB::sysObjectID.0 = OID: NET-SNMP-MIB::netSnmpAgentOIDs.10
SNMPv2-MIB::sysUpTime.0 = Timeticks: (7565377) 21:00:53.77
SNMPv2-MIB::sysContact.0 = STRING: Root <[email protected]>; (configure /etc/snmp/snmp.local.conf)
SNMPv2-MIB::sysName.0 = STRING: wy1
SNMPv2-MIB::sysLocation.0 = STRING: wy1.wuying.com (edit /etc/snmp/snmpd.conf)
SNMPv2-MIB::sysORLastChange.0 = Timeticks: (10) 0:00:00.10
``````````````````````
“Linux wy1 2.4.20-8smp”操作系统的信息已经出来了：）
现在我们在目标机上来写一些脚本来显标MEM、CPU、DiskIO
MEM数据的抓取脚本：
[root@wy1 root]# cat   mfree.sh
#!/bin/sh
/usr/bin/free -m | grep Mem |awk '{print $4}'
/usr/bin/free -m | grep Mem |awk '{print $2}'
[root@wy1 root]# sh   mfree.sh (上面一个数据是内存使用量，下面的是内存总量,M)
442
1006
CPU数据的抓取脚本
[root@wy1 root]# cat   cpustat.sh
#!/bin/sh
idle=`sar   -u 1 3 | grep Average | awk '{print $6}'`
used=`echo "101 - $idle" | bc -l -s`
echo $used
echo $idle
DiskIO数据的抓取脚本
[root@wy1 root]# cat   iostat.sh （显示硬盘IO，k/s）
#!/bin/sh
used1=`sar -d 1 3 | tail -1 | awk '{print $4}'`
used2=`echo "$used1 / 2" | bc -l`
echo $used2
echo $used2

好现在我们已经能得到这数据了，怎么才能让监控主机通过snmpd得到这些数据呢？可以在目标主机的/etc/snmp/snmpd.conf文件下面加个这些行：
exec .1.3.6.1.4.1.2021.53 mfree /bin/sh /root/mfree.sh
exec .1.3.6.1.4.1.2021.54 cpustat /bin/sh /root/cpustat.sh
exec .1.3.6.1.4.1.2021.55 iostat /bin/sh /root/iostat.sh
service   snmpd   restart   重启目标主机上的snmpd服务。
这样在监控主机上运行:
[root@wy2 root]# snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.53
UCD-SNMP-MIB::ucdavis.53.1.1 = INTEGER: 1
UCD-SNMP-MIB::ucdavis.53.2.1 = STRING: "mfree"
UCD-SNMP-MIB::ucdavis.53.3.1 = STRING: "/bin/sh /root/mfree.sh"
UCD-SNMP-MIB::ucdavis.53.100.1 = INTEGER: 0
UCD-SNMP-MIB::ucdavis.53.101.1 = STRING: "442"
UCD-SNMP-MIB::ucdavis.53.101.2 = STRING: "1006"
UCD-SNMP-MIB::ucdavis.53.102.1 = INTEGER: 0
其中UCD-SNMP-MIB::ucdavis.53.101.1 = STRING: "442" 中的442就是mfree.sh输出的第一个数据，1006是mfree.sh输出的第二个数据。OK，再做一下加工：
[root@wy2 root]# snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.53 | grep 53.101 | awk -F\" '{print $2}'
443
1006
好，我们已经通过snmpd从监控主机上得到了目标主机上内存使用的这两个数据了：）,其它的两个脚本也一样：
   CPU使用、空闲。
[root@wy2 root]# snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.54 | grep 54.101 | awk -F\" '{print $2}'
1.17
99.83
DISK IO 状态：
[root@wy2 root]# snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.55 | grep 55.101 | awk -F\" '{print $2}'
43.00000000000000000000
43.00000000000000000000

最后是编辑监控主机上的mrtg.cfg文件，在Target[xxxx]中加入上面的命令，下面是我的mrtg.cfg文件，前面的两个Target是我用/usr/bin/cfgmaker --global 'WorkDir: /home/httpd/mrtg/net' --global 'Options[_]:growright,bits' --ifref=ip [email protected]   命令生成的，后面的是我跟据platinum 的"mrtg能做些什么"的贴子内的mrtg.cfg文件改的。（在精华里，大家可以看一下）。
[root@wy1 root]# cat /home/httpd/mrtg/103/mrtg.cfg
# Created by
# /usr/bin/cfgmaker --global 'WorkDir: /home/httpd/mrtg/net' --global 'Options[_]:growright,bits' --ifref=ip [email protected]


### Global Config Options

#   for UNIX
# WorkDir: /home/http/mrtg

#   or for NT
# WorkDir: c:\mrtgdata

### Global Defaults

#   to get bits instead of bytes and graphs growing to the right
# Options[_]: growright, bits

WorkDir: /home/httpd/mrtg/103
Options[_]:growright,bits
Language: chinese
##############################################
# System: wy1
# Description: Linux wy1 2.4.20-8smp #1 SMP Thu Mar 13 17:45:54 EST 2003 i686
# Contact: Root <[email protected]>; (configure /etc/snmp/snmp.local.conf)
# Location: wy1.wuying.com (edit /etc/snmp/snmpd.conf)
##############################################


### Interface 2 >;>; Descr: 'eth0' | Name: '' | Ip: '192.168.13.103' | Eth: '00-06-5b-19-9d-ea' ###

Target[192.168.13.103_192.168.13.103]: /192.168.13.103:public@192.168.13.103:
SetEnv[192.168.13.103_192.168.13.103]: MRTG_INT_IP="192.168.13.103" MRTG_INT_DESCR="eth0"
MaxBytes[192.168.13.103_192.168.13.103]: 1250000
Xsize[192.168.13.103_192.168.13.103]: 300
Ysize[192.168.13.103_192.168.13.103]: 100
#kmg[192.168.13.103_192.168.13.103]: K/s,M/s
#kilo[192.168.13.103_192.168.13.103]: 1024
Title[192.168.13.103_192.168.13.103]: Traffic for eth0 192.168.13.103 -- wy1
PageTop[192.168.13.103_192.168.13.103]: <H1>;Traffic for eth0 192.168.13.103 -- wy1</H1>;
#Options[192.168.13.103_192.168.13.103]: growright,gauge,nopercent

### Interface 3 >;>; Descr: 'eth1' | Name: '' | Ip: '172.16.0.188' | Eth: '00-06-5b-19-9d-e9' ###

Target[192.168.13.103_172.16.0.188]: /172.16.0.188:public@192.168.13.103:
SetEnv[192.168.13.103_172.16.0.188]: MRTG_INT_IP="172.16.0.188" MRTG_INT_DESCR="eth1"
Xsize[192.168.13.103_172.16.0.188]: 300
Ysize[192.168.13.103_172.16.0.188]: 100
MaxBytes[192.168.13.103_172.16.0.188]: 1250000
#ShortLegend[192.168.13.103_172.16.0.188]:
#kmg[192.168.13.103_172.16.0.188]: b/s,Kb/s
#kilo[192.168.13.103_172.16.0.188]: 1024
Title[192.168.13.103_172.16.0.188]: Traffic for eth1 172.16.0.188 -- wy1
PageTop[192.168.13.103_172.16.0.188]: <H1>;Traffic for eth1 172.16.0.188 -- wy1</H1>;
#Options[192.168.13.103_172.16.0.188]: growright,gauge,nopercent

###MEM status

Target[wy1_mem]:`snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.53 | grep 53.101 | awk -F\" '{print $2}'`
#Targey[wy1_mem]: memTotalReal.0&memAvailReal.0:[email protected]
Xsize[wy1_mem]: 300
Ysize[wy1_mem]: 100
Ytics[wy1_mem]: 7
MaxBytes[wy1_mem]: 1006
Title[wy1_mem]:Memory State of WY1 IP 192.168.13.103 Server
PageTop[wy1_mem]:<H1>;Memory State of WY1 IP 192.168.13.103 Server</H1>;
ShortLegend[wy1_mem]: MB
kmg[wy1_mem]: MB
kilo[wy1_mem]:1024
YLegend[wy1_mem]: Memory Usage
Legend1[wy1_mem]: 可用内存
Legend2[wy1_mem]: 总内存量
Legend3[wy1_mem]: 可用内存
Legend4[wy1_mem]: 总内存量
LegendI[wy1_mem]: 可用内存
LegendO[wy1_mem]: 总内存量
Options[wy1_mem]: growright,gauge,nopercent

###cpu status
Target[wy1_CPU]:`snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.54 | grep 54.101 | awk -F\" '{print $2}'`
#Targey[wy1_CPU]: memTotalReal.0&memAvailReal.0:[email protected]
Xsize[wy1_CPU]: 300
Ysize[wy1_CPU]: 100
Ytics[wy1_CPU]: 7
MaxBytes[wy1_CPU]: 100
Title[wy1_CPU]:CPU State of WY1 IP 192.168.13.103 Server
PageTop[wy1_CPU]:<H1>;CPU State of WY1 IP 192.168.13.103 Server</H1>;
ShortLegend[wy1_CPU]:
kmg[wy1_CPU]: %
#kilo[wy1_CPU]:1024
YLegend[wy1_CPU]: CPU Usage
Legend1[wy1_CPU]: 已用CPU:
Legend2[wy1_CPU]: 可用CPU:
LegendI[wy1_CPU]: 已用CPU:
LegendO[wy1_CPU]: 可用CPU:
Options[wy1_CPU]: growright,gauge,nopercent

###iostat
###cpu status
Target[wy1_IO]:`snmpwalk -v 1   192.168.13.103 -c public .1.3.6.1.4.1.2021.55 | grep 55.101 | awk -F\" '{print $2}'`
#Targey[wy1_IO]: memTotalReal.0&memAvailReal.0:[email protected]
Xsize[wy1_IO]: 300
Ysize[wy1_IO]: 100
Ytics[wy1_IO]: 7
MaxBytes[wy1_IO]: 10000
Title[wy1_IO]: DISK IO State of WY1 IP 192.168.13.103 Server
PageTop[wy1_IO]:<H1>;DISK IO State of WY1 IP 192.168.13.103 Server</H1>;
ShortLegend[wy1_IO]:
kmg[wy1_IO]: K/s,M/s
kilo[wy1_IO]:1024
YLegend[wy1_IO]: DISK IO SPEED
Legend1[wy1_IO]: IO速度:
Legend2[wy1_IO]: IO速度:
LegendI[wy1_IO]: IO速度:
LegendO[wy1_IO]: IO速度:
Options[wy1_IO]: growright,gauge,nopercent


用indexmaker -o   /home/httpd/mrtg/103/index.html   /home/httpd/mrtg/103/mrtg.cfg   生成网页，在crontab 中加入*/5 * * * * mrtg /home/httpd/mrtg/103/mrtg.cfg
OK，这样我们就能通过mrtg生成的图来监控我们linux的机器的状态了。在IE中打入[url]http://192.168.13.105/mrtg/103[/url],mrtg的监控网页就出来啦（/home/httpd/是我appache的主目录）。当然我们也能用snmpwalk命令来得到安装有snmp服务的win2000机器的状态的数据：
如：

[root@wy1 103]# snmpwalk -v 1 192.168.1.5 -c public HOST-RESOURCES-MIB::hrStorage
HOST-RESOURCES-MIB::hrMemorySize.0 = INTEGER: 3800424 KBytes
HOST-RESOURCES-MIB::hrStorageIndex.1 = INTEGER: 1
HOST-RESOURCES-MIB::hrStorageIndex.2 = INTEGER: 2
HOST-RESOURCES-MIB::hrStorageIndex.3 = INTEGER: 3
HOST-RESOURCES-MIB::hrStorageIndex.4 = INTEGER: 4
HOST-RESOURCES-MIB::hrStorageIndex.5 = INTEGER: 5
HOST-RESOURCES-MIB::hrStorageIndex.6 = INTEGER: 6
HOST-RESOURCES-MIB::hrStorageType.1 = OID: HOST-RESOURCES-TYPES::hrStorageRemovableDisk
HOST-RESOURCES-MIB::hrStorageType.2 = OID: HOST-RESOURCES-TYPES::hrStorageFixedDisk
HOST-RESOURCES-MIB::hrStorageType.3 = OID: HOST-RESOURCES-TYPES::hrStorageFixedDisk
HOST-RESOURCES-MIB::hrStorageType.4 = OID: HOST-RESOURCES-TYPES::hrStorageFixedDisk
HOST-RESOURCES-MIB::hrStorageType.5 = OID: HOST-RESOURCES-TYPES::hrStorageCompactDisc
HOST-RESOURCES-MIB::hrStorageType.6 = OID: HOST-RESOURCES-TYPES::hrStorageVirtualMemory
HOST-RESOURCES-MIB::hrStorageDescr.1 = STRING: A:\
HOST-RESOURCES-MIB::hrStorageDescr.2 = STRING: C:\ Label:   Serial Number 581e89fe
HOST-RESOURCES-MIB::hrStorageDescr.3 = STRING: D:\ Label:New Volume   Serial Number 1cde6e55
HOST-RESOURCES-MIB::hrStorageDescr.4 = STRING: E:\ Label:DATA_BAK   Serial Number 30d29147
HOST-RESOURCES-MIB::hrStorageDescr.5 = STRING: F:\
HOST-RESOURCES-MIB::hrStorageDescr.6 = STRING: Virtual Memory
HOST-RESOURCES-MIB::hrStorageAllocationUnits.1 = INTEGER: 0 Bytes
HOST-RESOURCES-MIB::hrStorageAllocationUnits.2 = INTEGER: 4096 Bytes
HOST-RESOURCES-MIB::hrStorageAllocationUnits.3 = INTEGER: 4096 Bytes
HOST-RESOURCES-MIB::hrStorageAllocationUnits.4 = INTEGER: 4096 Bytes
HOST-RESOURCES-MIB::hrStorageAllocationUnits.5 = INTEGER: 0 Bytes
HOST-RESOURCES-MIB::hrStorageAllocationUnits.6 = INTEGER: 65536 Bytes
HOST-RESOURCES-MIB::hrStorageSize.1 = INTEGER: 0
HOST-RESOURCES-MIB::hrStorageSize.2 = INTEGER: 2050287
HOST-RESOURCES-MIB::hrStorageSize.3 = INTEGER: 15703529
HOST-RESOURCES-MIB::hrStorageSize.4 = INTEGER: 53263499
HOST-RESOURCES-MIB::hrStorageSize.5 = INTEGER: 0
HOST-RESOURCES-MIB::hrStorageSize.6 = INTEGER: 89444
HOST-RESOURCES-MIB::hrStorageUsed.1 = INTEGER: 0
HOST-RESOURCES-MIB::hrStorageUsed.2 = INTEGER: 1233681
HOST-RESOURCES-MIB::hrStorageUsed.3 = INTEGER: 591593
HOST-RESOURCES-MIB::hrStorageUsed.4 = INTEGER: 23583930
HOST-RESOURCES-MIB::hrStorageUsed.5 = INTEGER: 0
HOST-RESOURCES-MIB::hrStorageUsed.6 = INTEGER: 50611
HOST-RESOURCES-MIB::hrStorageAllocationFailures.1 = Counter32: 0
HOST-RESOURCES-MIB::hrStorageAllocationFailures.2 = Counter32: 0
HOST-RESOURCES-MIB::hrStorageAllocationFailures.3 = Counter32: 0
HOST-RESOURCES-MIB::hrStorageAllocationFailures.4 = Counter32: 0
HOST-RESOURCES-MIB::hrStorageAllocationFailures.5 = Counter32: 0
HOST-RESOURCES-MIB::hrStorageAllocationFailures.6 = Counter32: 0
上面显示的是192.168.1.5这台win2000SERVER的磁盘信息，可以看到盘符、每个分区的空间总量、使用量等。其它的相应建值还有：
Diskused:                               HOST-RESOURCES-MIB::hrStorageUsed
CPU:                                  HOST-RESOURCES-MIB::hrProcessorLoad
RUN Proc:                            HOST-RESOURCES-MIB::hrSWRunName
SYS Uptime:                             HOST-RESOURCES-MIB::hrSystemUptime SNMPv2-MIB::sysUpTime
SYS Date:                            HOST-RESOURCES-MIB::hrSystemDate
SYS Device:                             HOST-RESOURCES-MIB::hrDeviceDescr
SYS Descr:                            SNMPv2-MIB::sysDescr
SYS Name:                            SNMPv2-MIB::sysName
netcard speed:                          IF-MIB::ifSpeed
netcard physcal address:             IF-MIB::ifPhysAddress
这样我们就可以用一台linux主机通过snmp+mrtg来监控多台局域网内的linux/win2000SERVER了：）。
现在我最关心的是如何让snmp包穿过防火墙，（我用snmpwalk试了一下，抓不到公司在IDC防火墙后的SERVER的snmp数据）。不然，我就可以用snmp去监控公司在IDC防火墙后的SERVER的状态了。听说perl能实现，但不知用什么样的方法来实现？