One way:

c:\@Work\Perl\monks>perl -wMstrict -MData::Dump -le
"my @recs = qw(
   D1:111:92111
   D2:112:92111
   D3:111:92111
   D4:133:92111
   D5:111:90222
   D6:112:90222
   D7:111:90222
   D8:112:90222
   );
 ;;
 my %uniques;
 for my $rec (@recs) {
   my (undef, $field1, $field2) = split /:/, $rec;
   ++$uniques{$field2}{$field1};
   }
 dd \%uniques;
 ;;
 for my $field2 (reverse sort keys %uniques) {
   my $field1s = join ',', sort keys %{ $uniques{$field2} };
   print qq{'$field2' has unique values '$field1s'};
   }
"
{
  90222 => { 111 => 2, 112 => 2 },
  92111 => { 111 => 2, 112 => 1, 133 => 1 },
}

'92111' has unique values '111,112,133'
'90222' has unique values '111,112'
[download]