We have a lot of technical pages in the index (i've written about that before). They don't have similar fragments in the URL, so we can't close them in robots.txt file. We add meta noindex nofollow to every page about 3 weeks ago, but they are still all in the index.