awk question

[Igor Sovin]

Hi!
How to get free space left on file system /var/opt/ignite/recovery/archives
from the following output?

root@lev />bdf
Filesystem kbytes used avail %used Mounted on
/dev/vg00/lvol3 1048576 318056 724872 30% /
/dev/vg00/lvol1 1081344 178744 895592 17% /stand
/dev/vg00/lvol10 20480000 9382744 11010664 46% /var
/dev/vg01/ignite 116686848 101012876 14694408 87% /var/opt/ignite/recovery/archives
/dev/vg00/lvol9 3768320 2676568 1083264 71% /usr
/dev/vg00/lvol6 1048576 140544 900976 13% /tmp
/dev/vg00/lvol8 20480000 2682088 17658928 13% /opt
/dev/vg00/lvol7 1048576 23512 1021288 2% /home
/dev/vg00/lvol5 10256384 3250963 6567630 33% /depot
/dev/vg00/lvol4 2097152 1888853 195328 91% /.audit
anny:/ignite 143327232 51600 134320912 0% /ignite_Old

[Arunvijai_4 · Accepted solution]

Hi Igor,

You can try this,

# bdf /dev/vg01/ignite | awk '{a+=$4} END {printf "%d\n",a}'

-Arun

[Chauhan Amit]

Try following command :

bdf |grep -i /var/opt/ignite/recovery/archives |awk '{print $4}'

output will be in KB

-Amit

[Igor Sovin]

Thank you guys!

[inventsekar_1]

hi igor,
this u can use and get the bdf result in mb.
============================================================
#!/usr/bin/ksh

echo "Filesystem Mbytes used avail %used Mounted on"

bdf | grep -v Filesystem | awk '{ printf("%s %10d %10d %10d %4s %s\n",$1,$2/1024,$3/1024,$4/1024,$5,$6)}'
============================================================

[rariasn]

Hi!

# df -b

rgs,

ran

[rariasn]

Sorry,

df -b|grep -i ignite
/var/opt/ignite/recovery/archives (/dev/vgvarignite/lvrecovery) : 7676261 Kbytes free

rgs,
ran

[Kenan Erdey]

hi;

without awk:
bdf | head -5| tail -1 | tr -s ' ' | cut -f4 -d' '

[Bill Hassell]

Wow, I'm surprised that no one suggested the simplest method:

bdf /var/opt/ignite/recovery/archives

No grep, no awk, nothing but bdf. You could pipe the second line from bdf into read and extract just the free space:

bdf /var/opt/ignite/recovery/archives | while read HEADER REPLY
do
if [ "$HEADER" = "Filesystem" ]
then
read DEV CAP USED FREE PCT FS
MEGSFREE=$(echo $FREE | awk '{print int($1/1024+.5)}')
echo "$FS has $FREE Kbytes ($MEGSFREE megs) free space"
fi
done

The above shows both Kb and megs.

[Peter Nikitka]

Hi,

I think Igors problem was the 2line-output of the bdf command for long device names; Bills response will handle that.

To get a bdf-like 1line-output for further processing in all cases, I suggest (untested):

bdf | awk 'NR>1 {if(ok) print line; ok=0;line=$0}
NF==1 {line=line" "$0;ok=1}
NF>5 {ok=1}
END {print line}'

mfG Peter

[H.Merijn Brand (procura]

Why does no-one promote 'di'?
http://www.gentoo.com/di/
Heh, I didn't know the reference to binary packages from there would be to *my* site. Bet I will have to make 4.5 available soon now, won't I ...

a5:/u/usr/merijn 109 > di -?
di: illegal option -- ?
di ver 4.1 Default Format: smbuvpT
Usage: di [-ant] [-d display-size] [-f format] [-x exclude-fstyp-list]
[-I include-fstyp-list] [file [...]]
-a : print all mounted devices
-d x : size to print blocks in (512 - POSIX, k - kbytes,
m - megabytes, g - gigabytes, t - terabytes, h - human readable).
-f x : use format string
-I x : include only file system types in
-x x : exclude file system types in
-l : display local filesystems only
-n : don't print header
-t : print totals
Format string values:
m - mount point M - mount point, full length
b - total kbytes B - kbytes available for use
u - used kbytes c - calculated kbytes in use
f - kbytes free v - kbytes available
p - percentage not avail. for use 1 - percentage used
2 - percentage of user-available space in use.
i - total file slots (i-nodes) U - used file slots
F - free file slots P - percentage file slots used
s - filesystem name S - filesystem name, full length
t - disk partition type T - partition type, full length
See manual page for more options.
Exit 1
a5:/u/usr/merijn 110 > di -ff /tmp
Free
499.2
a5:/u/usr/merijn 111 > di -n -ff /tmp
499.2
a5:/u/usr/merijn 112 > di -n -ff -dm /tmp
499.2
a5:/u/usr/merijn 113 > di -n -ff -dk /tmp
511221
a5:/u/usr/merijn 114 >

http://mirrors.develooper.com/hpux/di-4.1-10.20.sd.bz
http://mirrors.develooper.com/hpux/di-4.1-11.00.sd.bz
http://mirrors.develooper.com/hpux/di-4.1-11.11.sd.bz
http://mirrors.develooper.com/hpux/di-4.1-11.23.sd.bz
http://mirrors.develooper.com/hpux/di-4.4-11.23.sd.bz

Enjoy, Have FUN! H.Merijn

[H.Merijn Brand (procura]

OK, all 4.5 ports are now up

a5:/u/usr/merijn 101 > di -?
di: illegal option -- ?
di ver 4.5 Default Format: smbuvpT
Usage: di [-ant] [-d display-size] [-f format] [-x exclude-fstyp-list]
[-I include-fstyp-list] [file [...]]
-a : print all mounted devices
-d x : size to print blocks in (512 - POSIX, k - kbytes,
m - megabytes, g - gigabytes, t - terabytes, h - human readable).
-f x : use format string
-I x : include only file system types in
-x x : exclude file system types in
-l : display local filesystems only
-n : don't print header
-t : print totals
Format string values:
m - mount point M - mount point, full length
b - total kbytes B - kbytes available for use
u - used kbytes c - calculated kbytes in use
f - kbytes free v - kbytes available
p - percentage not avail. for use 1 - percentage used
2 - percentage of user-available space in use.
i - total file slots (i-nodes) U - used file slots
F - free file slots P - percentage file slots used
s - filesystem name S - filesystem name, full length
t - disk partition type T - partition type, full length
See manual page for more options.

http://www.cmve.net/~mbrand/di-4.5-10.20.sd.bz
http://www.cmve.net/~mbrand/di-4.5-11.00.sd.bz
http://www.cmve.net/~mbrand/di-4.5-11.11.sd.bz
http://www.cmve.net/~mbrand/di-4.5-11.23.sd.bz

Enjoy, Have FUN! H.Merijn

[Arturo Galbiati]

Hi Igor,
bdf var/opt/ignite/recovery/archives|\
awk 'BEGIN {printf "%-16s %14s %14s %14s %8s %s\n","Filesystem","kbytes","used","avail","%used","Mounted on"}
!/Filesystem/ {KBYTES=KBYTES+$2; USED=USED+$3; AVAIL=AVAIL+$4; NBR=++NBR}
END {printf "%-16s %14s %14s %14s %7s \n","================","===========","==========","==========","===="
printf "%-16s %14d %s %11d %s %11d %s %4d% \n","total",KBYTES,"Kb",USED,"Kb",AVAIL,"Kb",USED*100/KBYTES
printf "%-16s %14d %s %11d %s %11d %s \n"," ",KBYTES/1024,"Mb",USED/1024,"Mb",AVAIL/1024,"Mb"
printf "%-16s %14d %s %11d %s %11d %s \n"," ",KBYTES/1048576,"Gb",USED/1048576,"Gb",AVAIL/1048576,"Gb"}

HTH,
Art

[Bill Hassell]

Peter pointed out that I did not test for a long line in a bdf listing. There is an inserted test for a null size and a second read to get the second line. This is common a problem with NFS filesystems (long source names):

bdf /house | while read HEADER REPLY
do
if [ "$HEADER" = "Filesystem" ]
then
read DEV CAP USED FREE PCT FS
[[ "$CAP" = "" ]] && read CAP USED FREE PCT FS
MEGSFREE=$(echo $FREE | awk '{print int($1/1024+.5)}')
echo "$FS has $FREE Kbytes ($MEGSFREE megs) free space"
fi
done