Load Average monitor for RHEL AS servers

[skt_skt]

Do you know of any tools coming with Linux which can monitor the load average and we can configure them to alert us when a certain a value is reached?

My LINUX versions include 2.1,3,4,5.1 etc.

what would be your suggetion on this

[Ivan Ferreira]

>>> coming with Linux

You can create a simple script, for example:

ALARM=10
uptime | tr -d "," | tr "." "," | awk -v ALARM=$ALARM '$10 > $ALARM { print $0 }'

You can then add notifications via mail.

Or you can use tools like nagios, zabbix, ganglia. Better if you do this because you will have historical graphs.

[Sreedharamurthy K]

you can use uptime, w (along with the above pipe to get mail alerts).

You can also use 'sar -q' to know the details.

If you have installed either Glance/ Performance Agent, you can configure it to send alert on yoru console, mail, manager console etc. It has better alarming capabilities.. based on conditions/probability..

[Raj_Ranjan]

Better use NAGIOS. we use to monitor all our servers(unix/linux/windows) using this tool.

[Ragu_3]

The tool 'dstat' is very nifty and you can use its output in an alert script. It is a combination of vmstat+iostat+ifstat.

[Jeeshan]

Hi Santosh

check this script, if the load goes to 6 in 15 minutes, it will create a text file. you can set the mail to parameter to send you mail.

#!/bin/bash
# Set up limit below
NOTIFY="6.0"
FTEXT='load average:'

# 15 min
F15M="$(uptime | awk -F "$FTEXT" '{ print $2 }' | cut -d, -f3)"

# compare it with last 15 min load average
RESULT=$(echo "$F15M > $NOTIFY" | bc)

# if load >= 6.0 create a file /tmp/file.txt
if [ "$RESULT" == "1" ]; then
echo 'LOAD ISSUE'>/tmp/file.txt
fi