The speed at which World-Wide-Web (WWW) spreading its division from an insubstantial number of web-pages to a
enormous centre of web information progressively improves web crawling complications in a search engine. A search
engine control a set of queries from a varying part of this world, and the satisfaction of it only depend on the knowledge
that it collects by means of crawling. The most general habit of the society is information distribution, and it is done by
means of publishing prearranged, semi-structured and amorphous reserve on the web (Nandy, Sarkar, and Das, 2012).
This social practice directs to an exponential expansion of web-resource, and hence it became necessary to crawl for
non stop updating of web-knowledge and variations of some presented sources in any conditions. This paper proposes
feature based crawling algorithm for light weighted and efficient crawling. The scaling technique is used to evaluate the
performance of proposed method with the standard crawler. The great speed presentation is observed after scaling,
and the extract of related web-source in such a extreme speed is examined.