The eliminate_duplicates_bucketwise method should work parallel for the buckets

[Ali Hürriyetoglu]

There is a sequential call for every bucket elimination. This should be parallel. Otherwise the process may take hour(s).