Monitoring system health on linux

I am running Red Hat EL 4, and I need to check the CPU useage, memory useage, disk useage, and network useage and write those values to an XML file as a percentage of total capacity.

The format needs to be something like this:
<cpu>48.0</cpu>
<ram>32.5</ram>

The file needs to be updated with new values every 5 seconds or so. Any ideas?
LVL 2
Thanatos2kAsked:
Who is Participating?
 
pjedmondCommented:
Disc bit is easy:

df | grep /dev/hda2 | awk {'print "<disc>" $5 "</disc>"'}
0
 
pjedmondCommented:
Yes....lots of ideas;)
0
 
pjedmondCommented:
RAM not too difficult:

cat /proc/meminfo | grep "Mem:" | awk '{t=int($2);u=($3); printf "<ram>%02.0f\%</ram>\n", (100*u/t)}'
0
Ultimate Tool Kit for Technology Solution Provider

Broken down into practical pointers and step-by-step instructions, the IT Service Excellence Tool Kit delivers expert advice for technology solution providers. Get your free copy now.

 
pjedmondCommented:
cpu idle is reasonably easy:

top -b -n1 | sed -n '6p' | awk {'print "cpu>" $8+0 "</cpu>"'}

(The +0 forces it to get rid of the %. Note that top -b, seems to add 2 extra lines on my test system, you may need to correct this for your system by changing the 6p to 4p, or whatever by trial and error.

Total cpu is also not to bad:

top -b -n1 | sed -n '6p' | awk {'print "cpu>" $2+$3+$4+$5+$6+$7 "</cpu>"'}

Now to put into a suitable script:)



0
 
pjedmondCommented:
report.sh:

-------8X---------
#!/bin/bash

while true; do
        df | grep /dev/hda2 | awk {'print "<disc>" $5 "</disc>"'}
        cat /proc/meminfo | grep "Mem:" | awk '{t=int($2);u=($3); printf "<ram>%02.0f\%</ram>\n", (100*u/t)}'
        top -b -n1 | sed -n '6p' | awk {'print "<cpu>" $2+$3+$4+$5+$6+$7 "%</cpu>"'}
sleep 300
done
-------8X---------

Obviously it takes time to run the commands, so it isn't exactly synchronised......

./report.sh > logfile.txt

Obviously, you might decide to add so date stamping in some form with the date command?

However...you may wish to look at nagios:

www.nagios.org

or mon:

www.kernel.org/software/mon

before trying to re-invent the wheel? Alternatively, as they are open source, you may pick up a few nice ideas from them:)

With respect to network monitoring, there is no % useage that I am aware of. What you could do is over a period of time record the number of packets sent, (cat /proc/net/dev), and then look at the number of packets sent in a 5 minute period as a percentage of a theorectical maximum, but this would not be realistic, so you'll have to have a bit of a re-think as to exactly what you'd like to record there.

HTH:)

0
 
xDamoxCommented:
Hi,

CPU usage can be captured by doing:

top | awk '/Cpu\(s\)/ { print "<cpu>", $2, "</cpu>" }'

The Memory usage can be captured by doing:

top | awk '/Mem:/ { print "<memory>",$5,"</memory>" }'

0
 
owensleftfootCommented:
0
 
pjedmondCommented:
Worth noting that xDamox's solutions don't work for me as my top output is:
-------------------8X---------------------------------
103 processes: 100 sleeping, 3 running, 0 zombie, 0 stopped
CPU states:  cpu    user    nice  system    irq  softirq  iowait    idle
           total    0.3%    0.0%    0.0%   0.0%     0.0%    0.0%   99.6%
Mem:   513504k av,  497544k used,   15960k free,       0k shrd,  224284k buff
                    377012k actv,   71784k in_d,    6988k in_c
Swap: 1044216k av,  122120k used,  922096k free                   98408k cached
-------------------8X---------------------------------

My tests were carried out on a RHEL3 server, so it is possible that RHEL4 may be slightly different? The same applies to the /proc/

paths, as the names may have changed slightly in RHEL4?
0
 
xDamoxCommented:
For your information pjedmond mine was tested on: Fedora

top - 23:54:39 up 2 days, 10:44,  2 users,  load average: 0.35, 0.25, 0.09
Tasks: 105 total,   1 running, 104 sleeping,   0 stopped,   0 zombie
Cpu(s):  7.8% us,  1.6% sy,  0.3% ni, 89.9% id,  0.4% wa,  0.0% hi,  0.0% si,  0.0% st
Mem:    515764k total,   379384k used,   136380k free,    21396k buffers
Swap:  1052248k total,    75996k used,   976252k free,   228492k cached
0
 
pjedmondCommented:
xDamox - Can I ask which version of FC?
0
 
pjedmondCommented:
Always nice to have a sanity check to look out for mistakes and spurious differences. Morale of the story is that Thanatos2k is going to have to understand the scripts before using them;)
0
 
xDamoxCommented:
pjedmond Fedora Core 5
0
 
gtkfreakCommented:
Try out ksysguard in case you are using KDE Desktop or gnome-system-monitor in case you are using the Gnome Desktop
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.