New Member - Array of Hashes between perl scripts via commandline argument

Hello @,
I am trying to pass "Array of Hash" as command line argument from one perl file to another perl file using reference. But the address becomes invalid once it reached another file . Is it because of different process address space between the perl files ? How to pass parameters like array, hashes and array of hashes from one perl file to another perl file using system("perl abc.pl <Array of Hashes>) command.
Code which call another perl command is:
#!/usr/bin/perl
use strict;
my (@AoH, $myRef, $CMD, $NewProcess );
@AoH = (
{
Instance => "orcl",
Metric => "6",
Filter => "Tablespace_name NOT LIKE 'TEMP'",
},
{
Instance => "orcl",
Metric => "16",
Filter => "Segment_name NOT LIKE 'TEMP'",
},

{
Instance => "orcl",
Metric => "10",
Filter => "Any filter statement",
},
{
Instance => "orcl",
Metric => "100",
Filter => "Any filter statement 1",
},
);
$NewProcess = "TCAOH.pl";
$myRef = \@AoH;
print "Callee Address $myRef \n";
$CMD = "perl $NewProcess -f $myRef" ;
print "To execute : $CMD";
system("$CMD > Mytestoutput.txt");


My Code which will be called is :
#!/usr/bin/perl
use strict;
use Getopt::Long;
my (@AoH, $href, $role, $myRef, $input,$ret) ;
print "Inside TCAOH.pl file \n";
$ret = GetOptions ( "f" => \$input);
print "Ret value : $ret \n";
print "Address : $input \n";
for (my $i=0; $i < @$input; $i++) {
print $input->[$i]{Instance}."\n";
print $input->[$i]{Metric}."\n";
print $input->[$i]{Filter}."\n";
}
1;

Output Error is:
Callee Address ARRAY(0x32e6fc)
To execute : perl TCAOH.pl -f ARRAY(0x32e6fc)Can't use string ("1") as an ARRAY
ref while "strict refs" in use at TCAOH.pl line 38.

Here line 38 is:
for (my $i=0; $i < @$input; $i++) {

Please suggest.
End of the day, I want to store set of data into memory and pass to different perl scripts as and when required to avoid I/O activities by each set of perl scripts.
What would be the best way to achieve this objective ?

Use the Storable module to exchange the data structure between the scripts.

http://search.cpan.org/~ams/Storable-2.30/Storable.pm

Thanks a lot for your reply.
Yes, it worked. But again its writing into a flat file and reading from the flat file which means I/O activity.
Is there any way, we can avoid such I/O overheads ?

How big of a data structure are you taking about? The example you've given would only take a few milliseconds of I/O overhead.

You haven't provided enough info on what the real goal is, but my first suggestion would be to redesign/refactor your scripts and turn one of then into a module.

Another option would be to use a database instead of a flat file.