People are likely to surf the web using search engines. Crawler (a part of search engine) continuously crawls the web and keeps its collection as fresh as possible. An efficient crawler should address issues like unnecessary burden on the web crawler, parallelism of the crawling process, freshness of the web contents discovered and revisiting frequency of web pages. When pages are changing very fast then these crawlers need to visit the pages as frequently as possible. Today when web size has become very large, these revisits not only engage the network traffic for a longer time but the crawler will also not be able to crawl the complete web in feasible time. In this work, an alternate approach for optimizing the frequency of visits to sites, and a mechanism for computing the dynamic priority for any site has been developed. It adjusts the frequency of visit by dynamically assigning a priority to a site. It employs an ecology of crawl workers to crawl the web sites. The architecture designed is not only incremental but also scalable that can be parallelized at URL queues and crawl workers level, responsible for downloading documents from the web.