Hi! I'm trying to write a script that prints the % of disk space if it's under %80:
Example:
/var/log %50 # it prints
/opt %81 # doesn't print

Script results:
/var/log %50

I'm using the df -Ph command but can't find out how to print it if it's under %80. df's printing everything. Any ideas? Googling's not helping. Thanks!

It’d be easier to help with your script if you posted it. Hard to advise what’s wrong when we don’t know what you’re doing.

Doh! Sorry!

#!/bin/bash

SERVER_LIST=$1

if [ -z "$1" ]; then
echo -e "\nNo arguments supplied!\n"
echo -e "\nPlease call '$0 <argument>' to run this command!\n"
exit 1
fi



echo " "
for HOST in `cat $SERVER_LIST`
do
echo "Servername : $HOST"
echo -e "Disk Space for /opt: "; rosh -l root -n $HOST -t "df /opt/"
echo "------------------------------"
done
echo " "

Please use [code] tags when posting code or output. See the link in my signature.

The code in the script doesn't match what you said in the OP, so we're still not clear on what you're doing.

df won't filter by size, you'd need to apply additional filters to the output of the df command. I'd suggest grep or awk.

As you say, df prints everything, so first you'd need to filter out (exclude) the lines that are not "real", those with Filesystem tmpfs (df -x), or include those that are (df -t) See man df

Then, you'd need to use grep or awk to select the lines you want to see, based on the Use% column. See their man pages.

I am curious why you want to only see partitions where the Use% is below 80%...I'd think the ones that are fuller -- higher than 80%, would be of more concern.

1 members found this post helpful.

What is rosh?

I would use awk.
It takes the output of df, divides column 3 by column 2 and prints the line if the result is less than 0.8.

2 members found this post helpful.

Just learning awk: Couldn't one just select by column 5? Something like

Without the 0+$5 it treats the value as text and not a number I do believe...

1 members found this post helpful.

Could be. I had to use $5<20 to get any filtering at all, but it seemed to work...would text compare as well? (I’d think not)

Sure alphanumeric sorting.

df | awk '$5 > 10 {print}'

And nothing speaks against the -P option that was initially suggested.
Certain versions of df break a long in two even if the output device is not a terminal - the -P prevents that.

Thanks everyone!