Space storage shell script issue!

Shell Programming and Scripting
1

Hi All,

Linux 86x64bits Red Hat Linux O/S

Could someone please check and let me know if the shell script has any syntax error as it's not sending mails

Shell script:

cat dataspace.sh

#!/bin/bash

export DBALIST="xyz@abc.com"

export data_capacity

df -k /oradata > dfk.result
data_capacity=`grep -v File dfk.result | awk -F" " '{ print $4 }' | cut -c 1-3`
if [[ $data_capacity > 30% ]]
then
     mailx -s "Filesystem /oradata is ${data_capacity} filled" $DBALIST < dfk.result
fi

Thanks for your time!

Regards,

a1_win

2

If you'd had a syntax error, the shell or tools would have told you. Use set -x as second command after the shebang (#!...) and see if there is some value or condition not met.

3 
 
echo $data_capacity
if  [ "$data_capacity" -gt "30" ]
then
mailx -s "Filesystem /oradata is ${data_capacity} filled" $DBALIST < dfk.result
else
echo "I am in else part"
fi
4

Hi,

The output is coming as:

Filesystem /oradata is :
39% -> in the next line

The output should have been like:

Filesystem /oradata is:39% filled

Thanks
a1_win

5

Can you please post sample df -k output from your system here. ( the one with /oradata)?

Else simulate the below code for your requirement.

 
df -k /stand
 
/stand                 (/dev/vg00/lvol1       ) :   269032 total allocated Kb
                                                     38864 free allocated Kb
                                                    230168 used allocated Kb
                                                        85 % allocation used

TEST SSYTEM> data_capacity=`df -k /stand | grep -v Kb | awk '{ print $1 }'`
 
TEST SSYTEM>[ $data_capacity -ge 80 ] && mailx  -s "Filesystem /oradata is ${data_capacity}%filled" $DBALIST
6

dfk.result output:

cat dfk.result

Filesystem 1K-blocks Used Available Use% Mounted on
abc:/vol/xyz_oradata
471859200 180505120 291354080 39% /oradata

7 
size=$(nawk '{disk_size=$4;}END{print substr(disk_size,1,length(disk_size)-1)}' dfk.result) && [ "${size}" -gt "30" ] && mail -s "Filesystem /oradata is ${size} filled" $DBALIST < dfk.result
8 
 
data_capacity=`tail -1 df.k | cut -d" " -f4 | sed 's/%//'`
 
[ $data_capacity -ge 30 ] && mailx  -s "Filesystem /oradata is  ${data_capacity}%filled" $DBALIST
9

Thanks to all for your valuable time!

I have now the output as expected!

Please let me know if there is a shell script which checks all the mounts in the Oracle Db Tier that alerts once the space usage crosses over 75%

Thanks for your time once again!

Regards,
a1_win

---------- Post updated 08-24-11 at 04:38 AM ---------- Previous update was 08-23-11 at 11:55 AM ----------

Continuing from - Disk Monitoring shell script giving incorrect information thread..

Using the code below:

$ df -h | grep % | sed 's,%,,g' | awk ' { if ($5>=30) {print $0} } ' > disk_chk_out
$ mailx -s "Disk Check > 30%" abc@xyz.com < disk_chk_out

Still getting the output which should not be receiving from e-mail alert as the criteria is below 30%:

/dev/sda1              99M   21M   74M  22 /boot

tmpfs                 9.8G   28M  9.8G   1 /dev/shm

E-mail alert output as below:

Filesystem            Size  Used Avail Use Mounted on
/dev/sda3              20G   13G  6.2G  68 /
/dev/sda1              99M   21M   74M  22 /boot
                      400G  272G  129G  68 /oradata
                      1.2T  526G  691G  44 /orabkup
                      2.3T  1.6T  728G  69 /software
                      2.3T  1.6T  728G  69 /backup
                      2.3T  1.6T  728G  69 /patches
                      450G  170G  281G  38 /xyz_oradata
                      150G   67G   84G  45 /orabin
                      225G  106G  120G  48 /oradata_TEST
tmpfs                 9.8G   28M  9.8G   1 /dev/shm
                       40G   20G   21G  48 /abc_orabin

Now what's wrong here

Thanks,

a1_win

10

Use below awk:

awk 'NF==5&&$4>30||NF==6&&$5>30'

If it does not help. Post df -h output here.

11

Hi Panyam,

Thanks but that didn't help either

df -h output:

Filesystem            Size  Used Avail Use% Mounted on
/dev/sda3              20G   13G  6.2G  68% /
/dev/sda1              99M   21M   74M  22% /boot
fcsan02:/vol/xxx_oradata
                      400G  272G  129G  68% /oradata
fcnas01:/vol/oraclebackups/xxx
                      1.1T  526G  578G  48% /orabkup
fcnas01:/vol/ebs/software
                      2.3T  1.6T  728G  69% /software
fcnas01:/vol/ebs/backup
                      2.3T  1.6T  728G  69% /backup
fcnas01:/vol/ebs/patches
                      2.3T  1.6T  728G  69% /patches
fcsan03:/vol/xxx_xxoradata
                      450G  170G  281G  38% /xyz_oradata
fcsan03:/vol/xxx_scripttest
                       20G   96K   20G   1% /scripttest
fcsan03:/vol/orabin/xxx
                      150G   67G   84G  45% /orabin
fcsan01:/vol/ebs/eptp51
                       10G  5.5M   10G   1% /ebs/eptp51
fcsan02:/vol/SnapManager_20101206155612743_xxx_oradata
                      225G  106G  120G  48% /oradata_xxx
tmpfs                 9.8G   28M  9.8G   1% /dev/shm
fcsan03:/vol/xxx_oraerp
                       40G   20G   21G  48% /orabin
fcsan03:/vol/xxx_archerp
                       40G  1.4G   39G   4% /oradata2

Regards,

a1_win

12

Try out the below one, and let us know, is the one you need as a email ?

 
df -h | awk '/%/ {a=substr($(NF-1),1,length($(NF-1))-1); if(a>30){print $0}}'
13

@ a1_win:
Since you ignore PMs and moderator notes I want to make you think about, that if you continue to ignore to use code tags, you collect infraction points which will result in a permanent ban after collecting an amount of them. So maybe consider reading your PMs and use code tags.
See (B) in this link if you do not understand how to use them:
The UNIX and Linux Forums - Forum Rules

14

Hi,

Still I am getting the below output in e-mail alert which is below 30% criteria:

40G  1.4G   39G   4% /oradata2

I used the code tag as above but didn't help.

df -h | awk '/%/ {a=substr($(NF-1),1,length($(NF-1))-1); if(a>30){print $0}}'

Thanks,
a1_win

---------- Post updated at 06:39 AM ---------- Previous update was at 06:35 AM ----------

Sure Zaxxon, will start using the code tags by typing as Code Tag which I juts did in the last post

Thanks for the information

15

code tag means

 
 your data

---------- Post updated at 04:18 PM ---------- Previous update was at 04:16 PM ----------

Strange, why it is taking 4 as greater than 30

still checking...

 
$ nawk '/%/ {a=substr($(NF-1),1,length($(NF-1))-1); if(a > 30){print $0}}' test                                                                   
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda3              20G   13G  6.2G  68% /
                      400G  272G  129G  68% /oradata
                      1.1T  526G  578G  48% /orabkup
                      2.3T  1.6T  728G  69% /software
                      2.3T  1.6T  728G  69% /backup
                      2.3T  1.6T  728G  69% /patches
                      450G  170G  281G  38% /xyz_oradata
                      150G   67G   84G  45% /orabin
                      225G  106G  120G  48% /oradata_xxx
                       40G   20G   21G  48% /orabin
                       40G  1.4G   39G   4% /oradata2

 
bash-3.00$ cat test 
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda3              20G   13G  6.2G  68% /
/dev/sda1              99M   21M   74M  22% /boot
fcsan02:/vol/xxx_oradata
                      400G  272G  129G  68% /oradata
fcnas01:/vol/oraclebackups/xxx
                      1.1T  526G  578G  48% /orabkup
fcnas01:/vol/ebs/software
                      2.3T  1.6T  728G  69% /software
fcnas01:/vol/ebs/backup
                      2.3T  1.6T  728G  69% /backup
fcnas01:/vol/ebs/patches
                      2.3T  1.6T  728G  69% /patches
fcsan03:/vol/xxx_xxoradata
                      450G  170G  281G  38% /xyz_oradata
fcsan03:/vol/xxx_scripttest
                       20G   96K   20G   1% /scripttest
fcsan03:/vol/orabin/xxx
                      150G   67G   84G  45% /orabin
fcsan01:/vol/ebs/eptp51
                       10G  5.5M   10G   1% /ebs/eptp51
fcsan02:/vol/SnapManager_20101206155612743_xxx_oradata
                      225G  106G  120G  48% /oradata_xxx
tmpfs                 9.8G   28M  9.8G   1% /dev/shm
fcsan03:/vol/xxx_oraerp
                       40G   20G   21G  48% /orabin
fcsan03:/vol/xxx_archerp
                       40G  1.4G   39G   4% /oradata2
16

This should solve the problem:

 
TEST_SYSTEM>df -h | awk '/%/ {$0=(a?a" "$0:$0);a="";gsub("%","");print $0;next} {a=$0;next}' | awk '$5>30' > disk_chk_out
TEST_SYSTEM>mailx -s "Disk Check > 30%" abc@xyz.com < disk_chk_out
17

substr is the issue..

try this.

 
$ df -h | nawk '/%/ {if(length($(NF-1))> 2) {a=substr($(NF-1),1,length($(NF-1))-1); if(a > 30){print $0}}}'
18

Many thanks to all of you for your time to help resolve this issue now!

itkamaraj: Thanks so much for your time. Yes now it's working fine as per the expected output

itkamaraj: Could you please explain me the code used as below:

Code Tag:

$ df -h | nawk '/%/ {if(length($(NF-1))> 2) {a=substr($(NF-1),1,length($(NF-1))-1); if(a > 30){print $0}}}'

Thanks,
a1_win

19 
 
nawk '/%/ {if(length($(NF-1))> 2) {a=substr($(NF-1),1,length($(NF-1))-1); if(a > 30){print $0}}}'

/%/  --> Match % and process the lines which is having %
$(NF-1) --> Print the value of (last - 1) column value (disk usage column)
length($(NF-1)) --> Get the length of the disk usage column.
if(length($(NF-1))> 2  --> check the length is greater than 2 ( It will eliminate the usage which is less than 10%)
a=substr($(NF-1),1,length($(NF-1))-1);  --> Strip the % value from the disk usage column ( eg: 85% to 85 )
then check a > 30, then print the line.
More example for understaing :
echo "abc edf ghi" | awk '{ print $NF }'            # prints the last record value
echo "abc edf ghi" | awk '{ print $(NF-1) }'  # prints the last-1 record value
echo "abc edf ghi" | awk '{ print $(NF-2) }'  # prints the last-2 record value
echo "abcd efg hi" | awk '{print length($1)}'  # prints the lenght of first record
echo "abcd efg hi" | awk '{print length($2)}'  # prints the lenght of second record
echo "abcd efg hi" | awk '{print length($3)}'  # prints the lenght of third record
echo "abcdefg" | awk '{ print substr($0,1,2)}'   # prints 2 letters from the 1st position