[SOLVED] Bash script to Monitor RHEL server

r34per · 08-13-2019, 12:27 PM

Hi everyone

I am a newbie to the Linux world so please bear with me...

I need to create a bash script to monitor a RHEL server, specifically to monitor the top 10 CPU process' (with usage and process information), the CPU load, the Memory usage, the Disk space and Network traffic monitoring. I then need to have this script running every second for 5 minutes per day and write the output to a log file in the /tmp directory in a format which is understandable by us regular humans, so something like a table with the date and time the reading was taken.

I have tried searching google and this forum, did not find anything that made much sense.

I have a very basic understanding of how to write scripts.

Does anyone have something like this that can help? or can someone point me in the right direction? Any help will be appreciated

Thanks

################
EDIT
################

Sorry should have posted more info - I have created this script so far:

Code:

#! /bin/bash

printf "Date and Time\t\tMemory\t\tDisk\t\tCPU\n"

end=$((SECONDS+180))

while [ $SECONDS -lt $end ]; do

MEMORY=$(free -m | awk 'NR==2{printf "%.2f%%\t\t", $3*100/$2 }')

DISK=$(df -h | awk '$NF=="/informix_backup"{printf "%s\t\t", $4}')

CPU=$(top -bn1 | grep load | awk '{printf "%.2f%%\t\t\n", $(NF-2)}')

CPU2=$(ps auxf|sort -nr -k3 | head -5)

CPU3=$(ps -e -o pcpu,pid,user,args | sort -k1 -nr | head -5)

NOW=$(date +"%m-%d-%Y-%H:%M:%S")

echo "$NOW$MEMORY$DISK$CPU"

echo
printf "ate and Time\t\ttop 10 CPU usage processes with user\n"
echo
echo "$NOW$CPU2"

echo
printf "Date and Time\t\ttop 10 CPU consuming process\n"
echo
echo "$NOW$CPU3"

sleep 1

done

I have a few issues -

The output format comes in a messy blocks and cannot figure out how to make it a readable table
I need to incorporate some kind of network traffic monitoring command in there (so monitoring network traffic from and to a specific NIC)
And I need to log the output to a log file in the /tmp directory

Any help is appreciated

scasey · 08-13-2019, 01:22 PM

See what you can do with top...

Code:

man top

TB0ne · 08-13-2019, 02:00 PM

Quote:

Originally Posted by r34per

Hi everyone

I am a newbie to the Linux world so please bear with me...

I need to create a bash script to monitor a RHEL server, specifically to monitor the top 10 CPU process' (with usage and process information), the CPU load, the Memory usage, the Disk space and Network traffic monitoring. I then need to have this script running every second for 5 minutes per day and write the output to a log file in the /tmp directory in a format which is understandable by us regular humans, so something like a table with the date and time the reading was taken.

I have tried searching google and this forum, did not find anything that made much sense. I have a very basic understanding of how to write scripts. Does anyone have something like this that can help? or can someone point me in the right direction? Any help will be appreciated

Thanks

Really? There are many, MANY hits for "linux system monitoring scripts" (over 12 million)....which ones didn't make sense? First two hits have scripts with examples. You can easily modify them to write to whatever file/directory you want, in whatever format you'd like. The printf statement in bash can get you customized output, as can just using tabs or commas, for easy import into a spreadsheet. Sleep statements can get the one-second interval, a simple counter can tick up to the five-minute mark, and cron can run it whatever time of day you'd like.

So...what have you actually done/tried so far? Read the "Question Guidelines" link.

BW-userx · 08-13-2019, 06:52 PM

that is defiantly a piece together project if you're going to google about it.

Quote:

need to create a bash script to monitor a RHEL server, specifically to monitor the top 10 CPU process' (with usage and process information), the CPU load, the Memory usage, the Disk space and Network traffic monitoring. I then need to have this script running every second for 5 minutes per day and write the output to a log file in the /tmp directory in a format which is understandable by us regular humans, so something like a table with the date and time the reading was taken.

1. How to monitor / get top 10 CPU process?
2. how to monitor / get memory usage?
3. how to get disk space?
4. how to check network traffic?
5. how to use job to run a script every 5 minutes?
6. how to output from a script into a file?

one step at a time, put it all in a wrapper called a script, with some testing until it works the way you want and you got yourself a finished project.

just for the heck of it

Code:

$ top -b -n 10 | awk 'FNR>6 {print $1" "$2" "$10" "$11" "$12}' > toppers ; cat toppers
PID USERNAME TIME WCPU COMMAND
45652 userx 0:47 3.71% mpv
67686 root 11:56 2.20% Xorg
77560 userx 8:26 1.17% firefox
78279 userx 4:27 0.49% gkrellm
66040 userx 0:44 0.39% qbittorrent
94845 userx 0:02 0.20% firefox
77717 userx 20:30 0.10% firefox
98081 userx 0:54 0.00% firefox
77744 userx 0:42 0.00% python3.6
11934 userx 0:37 0.00% firefox

r34per · 08-14-2019, 10:46 AM

Quote:

Originally Posted by TB0ne

Really? There are many, MANY hits for "linux system monitoring scripts" (over 12 million)....which ones didn't make sense? First two hits have scripts with examples. You can easily modify them to write to whatever file/directory you want, in whatever format you'd like. The printf statement in bash can get you customized output, as can just using tabs or commas, for easy import into a spreadsheet. Sleep statements can get the one-second interval, a simple counter can tick up to the five-minute mark, and cron can run it whatever time of day you'd like.

So...what have you actually done/tried so far? Read the "Question Guidelines" link.

Thanks for the response! I have edited the main post to show how far I have got. Also have listed the issues I have there.

I have searched Google but 99% of the time its asking me to install an external tool, which is something I cannot do. I can only work with what comes as part of the standard OS monitoring tools. There were a few pointers I was able to pick up which is where most of the above commands came from.

r34per · 08-14-2019, 10:48 AM

Quote:

Originally Posted by BW-userx

that is defiantly a piece together project if you're going to google about it.

1. How to monitor / get top 10 CPU process?
2. how to monitor / get memory usage?
3. how to get disk space?
4. how to check network traffic?
5. how to use job to run a script every 5 minutes?
6. how to output from a script into a file?

one step at a time, put it all in a wrapper called a script, with some testing until it works the way you want and you got yourself a finished project.

just for the heck of it

Code:

$ top -b -n 10 | awk 'FNR>6 {print $1" "$2" "$10" "$11" "$12}' > toppers ; cat toppers
PID USERNAME TIME WCPU COMMAND
45652 userx 0:47 3.71% mpv
67686 root 11:56 2.20% Xorg
77560 userx 8:26 1.17% firefox
78279 userx 4:27 0.49% gkrellm
66040 userx 0:44 0.39% qbittorrent
94845 userx 0:02 0.20% firefox
77717 userx 20:30 0.10% firefox
98081 userx 0:54 0.00% firefox
77744 userx 0:42 0.00% python3.6
11934 userx 0:37 0.00% firefox

Thanks

This was most helpful - I did a Google search and like you said tried to take it one step at a time. Got each individual command working then tried to tie it all together. Im still having issues though - I have updated the main post. Any pointers will be appreciated

BW-userx · 08-14-2019, 12:20 PM

you might want to work on your formatting
https://linuxconfig.org/bash-printf-...-with-examples
should give you some good examples ...