To avoid that (and to boost performance) use placeholders instead:

my $sth = $dbh->prepare("INSERT INTO ques(username, userque, ...) VALU
+ES (?, ?, ...)");
foreach my $t (split '~', $txtque){
     $sth->execute($usernamenow, $t, ...);
}
[download]

when i try to split about 10,000 array of strings and try to insert into mysql - i get server error but ..

OK, I give up -- what's the server error that you're getting? Knowing that will help us understand what the problem might be. And once we have an idea what the problem is, we can start to propose a solution.

Update: It's really quite rude to update a root node without making it clear that you're doing so. If you're going to do that, mark it clearly as an UPDATE (as I've done). Don't just add stuff -- it makes the thread confusing for anyone who comes along later to read it.

...
  $count = $dbh->do("INSERT INTO  ques (...) VALUES (...)",
                    undef,
                    $usernamenow,
                    $txtque,
                    ...
           );
...
[download]

  $count = $dbh->do('INSERT INTO ques (...) VALUES (?,?,?,?,...)',
                    undef,
                    $usernamenow,
                    $txtque,
                    ...
           );
[download]

Vagueness aside, if this is a recurring need also consider reformatting your data into a format which the mysql bulk loader can handle (for instance CSV). That's more than likely going to run faster than any ad hoc DBI you whip up.

2) Please give us an error message. The error you've posted tells us that you are having trouble with a CGI script, and that's about all. Run your script from the command line, not via a web browser, and report the error generated, and maybe someone can help.

3) "here the problem wasn't from mysql command it was from perl it doesn't seems like to split string that is large over 20,000 arrays or about 5MB"

Nope, sorry, you're making a false assumption about the problem. The Perl interpreter can handle split on much bigger strings than that. In fact, I ran the following very simple code to demonstrate that on my laptop PC

use strict;
use warnings;

my $string="";
my %charmap = ( 0 => "a",
                                1 => "~" );

for (1..50000000)
{
        $string .= $charmap{int(rand(2))};
}

open (my $testfile,">","testfile.txt") or die $!;
print $testfile $string;

my @arr = split "~",$string;

for (@arr)
{
        print "$_\n";
}
[download]

it took a while and slowed my laptop down, but it ran without a problem, and that was a string of approximately 50MB - ten times what you're working with.

OK, so by commenting the sql line out you're on the right lines - try to narrow the problem down to the simplest case that still generates an error. But first, find out what the error is - run it locally and see what it does.

Update You could also get useful error messages by putting:     use CGI::Carp qw(fatalsToBrowser); at the beginning of your script.

First of all, are you sure this code works? perltidy reports that it parses this way:

...
  $count = $dbh->do("INSERT INTO  ques (...) VALUES (...)",
                    undef,
                    $usernamenow,
                    $txtque,
                    ...
           );
...
[download]