Appropriate usage of IO::CaptureOutput for a mysqldump and gzip?

[loothi]

Hiya.

Messing about with this module IO::CaptureOutput.

I want to do a mysqldump and redirect the raw SQL through gzip. I used to do this with the system command

system("mysqldump --add-drop-table -h $global_host -u $user -p'".$pass."' $db_name  | gzip > $backup_file")

...but if the mysqldump fails, I was still getting a success exit code so I thought I'd split it into the two commands so it's more robust for debugging. Is this method below a decent way to do this? I haven't written any code in ages and am wondering if anyone can check this for me (just the concept, as it's just a snippet and I'd use strict etc in the real script).

#!/usr/bin/perl

use IO::CaptureOutput qw/capture_exec capture/;

sub mysqldump {
    system('mysqldump --add-drop-table -h localhost -u myuser -pmypass
+ mydb ');
}

  capture \&mysqldump, undef, \$stderr, 'outfile.sql';

    if ($stderr) {
        print "FAILED: $stderr\n";
    }else{
        @args = ('gzip','outfile.sql');
        ($stdout, $stderr, $success, $exit_code) = capture_exec(@args)
+;

        if (!$success) {
            print "Failed gzip: $stderr\n";
        } else {
            print "Success gzip\n";
        }   
    }

ta. Loothi.

[remiah]

Hello loothi.

I am afraid for the database size. Dumps becomes very large and the buffer also becomes large. How about using --result-file and --log-error option of mysqldump?

For capturing output of other programes, I see Capture::Tiny examples here at monk many times. I would like to give it a try, but, not yet.

regards