In a Google Groups thread, a forum member wonders about the details and logistics about Google's cache. JohnMu, a Google representative, offers his $0.02 and gives some interesting tidbits.
First and foremost, if you experience any cache issues or errors, or if you run into a "stuffed" cache, you need not worry, says JohnMu.
What if you don't want your site to be cached? Use the "noarchive" meta tag, he says. He explains that the "no-cache" meta tag and similar HTTP header tags (on the other hand) are for browsers only.
Do you suffer any penalties if your page can't be cached or if the cache is not coming up properly? John says no.
Though I realize there are sometimes situations where the contents of a cached URL can help determine what is happening (say when a hacked page is returning different content to the Googlebot), I haven't seen many (or any?) situations where a missing or "broken" cached page has had any impact on crawling, indexing or ranking of a site.
Forum discussion continues at Google Groups.
