while(<MYINPUTFILE>)
    {
            cat /mnt/scripts/lagtime/tmp/input.txt |awk '{printf "%1.5
+0s\t%0.60s\t%0.45s\t %0.45s\n", $1, $2, $3, $4}' | sh /mnt/scripts/la
+gtime/calculate_lag.sh > /tmp/snapvault_status.out.out;
    }
[download]

You're calling the awk command (or trying to... you forgot system) for every line in the input file (input.txt) without using that line anywhere in the command ...  What was the idea?

Awk is needed to put the columns in the right place and then pipe the output to the calculate script which basically looks at a value higher then 24 and then is needs to be placed in the /tmp/snapvault_status.out.out file It looks like I should not use the while statement?

It looks like I should not use the while statement?

Yes, I can't see much use in it — except to slow down the processing ;)

As it looks, calling it just once (using system !) should be sufficient.

I first began using Perl when an old AWK script became unmanagable. a2p (already cited by swampyankee above) converts AWK scripts to Perl and was tremendously helpful for learning how do things in a Perlish way. I strongly recommend experimenting with it if you are familiar with AWK and are learning Perl.

Unfortunately, it just dumps the Perl code and doesn't really explain how constructs in AWK and Perl compare, so I've added an annotated example here. Note the difference in order. On the bash command line, you use the order: (a) specify input (b) specify processing (c) specify output. Here we switch (b) and (c), and open both the input and output streams before processing.

use strict;
use warnings;

# open an input stream
# same effect as:
# cat /mnt/scripts/lagtime/tmp/input.txt|

# if fail, print reason ($!) and abort
open IN, '<', '/mnt/scripts/lagtime/tmp/input.txt'
  or die "Couldn't open input file: $!";

# open an output stream
# same effect as:
# > /tmp/snapvault_status.out.out;

# if fail, print reason ($!) and abort
open OUT, '>', '/tmp/snapvault_status.out.out'
  or die "Couldn't open output file: $!";


# equivalent to your awk program
# { printf "%1.50s\t%0.60s\t%0.45s\t %0.45s\n", $1, $2, $3, $4 }

while (my $line=<DATA>) {
  # chomp: remove record separator
  # split: break into fields using default AWK field separator
  # (a single space which is equivalent to the Perl regex m{ }

  my @aFields = split(m{ }, $line);
  chomp $line;
  printf OUT "%1.50s\t%0.60s\t%0.45s\t %0.45s\n", @aFields
}
[download]

Also, when you open an input or output stream, always check for errors. There are ever so many reasons why a file doesn't get opened when you think it should. If you use open ... or die ..., you will save yourself hours scratching your head wondering why your program doesn't print anything.

To learn more about the commands in this script, see open, chomp, and split. You may also find these in depth tutorials on opening files (perlopentut) and regular expressions (perlretut) helpful.

Best, beth

Update: Added links to relevant Perl documentation.

Nor doesn't make any sense. Why would you want to pass your entire log file through awk for each line of the file? And since you aren't doing anything interesting in Perl, why not write the entire thing in a shell script?

As for your actual question, you're calling system, but you aren't checking its result. Perhaps an error occurred.

First suggestion, make sure that your open worked.

thank you very much I will re do the script now with your suggestions. Still learning sorry if the script I wrote made little sense.

cat /mnt/scripts/lagtime/tmp/input.txt |awk '{printf "%1.50s\t%0.60s\t
+%0.45s\t %0.45s\n", $1, $2, $3, $4}' | sh /mnt/scripts/lagtime/calcul
+ate_lag.sh > /tmp/snapvault_status.out.out;
[download]

awk '{printf "%1.50s\t%0.60s\t%0.45s\t %0.45s\n", $1, $2, $3, $4}' /mn
+t/scripts/lagtime/tmp/input.txt | sh /mnt/scripts/lagtime/calculate_l
+ag.sh > /tmp/snapvault_status.out.out;
[download]

The problem is that the last output file is empty - this shouldn't come as a huge surprise if the last of the input lines causes an empty line to be written to the output - via the /mnt/scripts/lagtime/calculate_lag.sh script - since, within the loop, the output file is rewritten, c/w being appended to, each time.

In the light of this, I think I'd be tempted to

1. Modify /mnt/scripts/lagtime/calculate_lag.sh to take a single line as an argument
2. Coalesce the ssh invocations with the awk command
3. Finally, to re-write it [the snippet] something along the following lines i.e. untested:

use warnings;
use strict;

use autodie;

my @HOSTS = qw/datab01 datab02 datab03/;
my $IN_FILE = q(/mnt/scripts/lagtime/tmp/input.txt);
my $OUT_FILE = q(/tmp/snapvault_status.out.out);

open IN_FILE, qq/>$IN_FILE/;

foreach my $host (@HOSTS) {
    foreach (`ssh $host snapvault status | tail -n +3`) {
        local @_ = split; # Avoid warnings
        printf IN_FILE "%1.50s\t%0.60s\t%0.45s\t %0.45s\n", $_[0], $_[
+1], $_[2], $_[3];
    }
}

close IN_FILE;

open IN_FILE, qq/<$IN_FILE/;
open OUT_FILE, qq/>$OUT_FILE/;

while (<IN_FILE>) { print OUT_FILE `sh /mnt/scripts/lagtime/calculate_
+lag.sh $_` }

close IN_FILE;
close OUT_FILE;
[download]

use strict;
use warnings;
[download]

Thank you all very much for your advice this is very helpfull !!