We suggest creating a robots.txt file in the web root of the domain to address two issues. First to control the rate at which the website is being crawled which can help prevent a bot/spider from creating a massive number of database connections at the same time. Second to prevent specific bots from crawling the website. We suggest the following defaults, however you might want to add or remove the user agents denied, and adjust the crawl rate but we suggest nothing lower than 3 seconds.
User-agent: * Crawl-delay: 10 User-agent: Baiduspider Disallow: / User-agent: Sosospider Disallow: /