grep bdf for aavail > number

[Warren Cook_2]

Hello all,

Would like to know how to perform bdf, grep on 'avail' column for all filesystems greater than a given size. Also want to display all columns from bdf.
So far below cmd prints all fields (no spaces) but picks only one record from bdf (the matching one, I want to test for all > 4000000):
bdf|tr -s " " ||awk '{print $1 $2 $3 $4}'cut -f 2 -d " " |grep 4000000

thx

[Pete Randall]

Warren,

I'm reasonable certain I've seen something like this before and I'm willing to bet that I saw it in one of Bill McNamara's "Favorite Scripts" threads. Here's the first one, which also provides links to the other two. Have a look through them. Even if you don't find exactly what you're looking for, they're an invaluable resource.

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=51050


Pete

[Peter Day_2]

Warren,
Try this -

for a in $(bdf |awk '{print $4}')
do
if [[ "$a" != "avail" ]]
then
if [[ $a -gt 4000000 ]]
then
echo Bigger than 4000000
fi
fi
done

Hope this works, it did on my system.
Regards,
Pete


.

[RAC_1]

bdf|awk '{if($4>=106000) print $4}'

[Steve Steel]

Hi

Since for long file system names bdf uses 2 lines this is difficult to realise easily

df -n|cut -f2 -d"("|cut -f1 -d ")"|while read line
do
xx=$(bdf $line|tail -n +2)
echo $xx|while read a b rest
do
let val=$b
if [ "$val" -gt "4000000" ]
then
echo $xx
fi
done
done|sort -n


Steve Steel

[Mark Greene_1]

Try this:

IZE=512000
OLDIFS=$IFS
IFS="
"
for LINE in `bdf`; do
FIELD2=`echo $LINE|tr -s " "|awk '{print $2 }'`
FS_SIZE=`echo $FIELD2|tr -d "[:alpha:]" `
if [[ "$FS_SIZE" -ge $SIZE ]]; then
echo $LINE
fi
done

[Warren Cook_2]

Thank-you all for great and quick feedback ! I've tried all suggestions and will assign points shortly.

Mark, I'm using your suggestion, I can't get the IF to work. It appears to be comparing strings and always gives a true answer, so all bdf records are printed. I've tried several variations with no luck ... if you or anyone has a suggestion.

I mean this if:
if [[ "$FS_SIZE" -ge $SIZE ]]; then

thx

[Kent Ostby]

bdf | awk '$2>4000000'

[Sridhar Bhaskarla]

Hi,

bdf is not my favorite for checking the utilizations due to below warning from the man page of bdf.

WARNINGS
If file system names are too long, the output for a given entry is
displayed on two lines.

This can confuse your script quite a bit. So, I would write the script using 'df' something like this

LIMIT=100000

cleanup()
{
rm -f /tmp/fs$$ /tmp/fsreport$$
}

trap cleanup 0 1 2

for FS in $(df -l|awk '{print $1}')
do
AVAIL=$(df -k $FS|awk '/free allocated Kb/ {print $1}')
if [ $AVAIL -ge $LIMIT ]
then
bdf $FS |sed '1d'
fi
done


Adjust your if's and limits accordingly.

-Sri

[Kent Ostby]

oops .. missed the "no spaces part"

bdf | tr -s " " | awk '$2>4000000'

[Hein van den Heuvel]

In AWK mormally you count fields from the beginning of the line in.
The one liner below will handle the split lines from bdf transparantly by using NF (number of fields) to address fields relative to the end. It test for AVAIL > 4000000 and prints avail and mouintpoint.

bdf | awk '(NF<4){next} ($(NF-2) > 4000000) {print $(NF-2), $NF}'

Now if you want the filesystem printed too then it becomes more tricky. You need to remember it from the last line read with just 1 field and only pick up the current one if the re are 6 fields. Example

bdf | awk '(NF<5){f=$1; next} (NF>5){f=$1} ($(NF-2) > 4000000) {print f, $(NF-2), $NF}'

fwiw,
Hein.

[Arnold Hausmann]

Here's a nice little script that handles bdf's split lines nicely and formats bdf output nicely as well. We use it a lot to find out how much space is left within a certain percentage limit (the -p option).

Anyway, "dfspace | awk '{if ($4 > 40000) {print$0}}'" should produce your results and avoid the bdf 2-line limitation.

[Bill Hassell]

Here is a script I call bdfmegs. It always produces a single line for every filesystem. There is a -v option for additional filesystem details (like large-files). Best of all, all the numbers are in megs, not kbytes.