Google Sites automatically adds rel=canonical for the site’s custom domain.Though since both the Google site in the custom domain are indexed, how long for the canonical to cause GoogleBot to remove the Google site URLs
JOHN MEULLER (Webmaster Analyst from Google):
Probably not that quickly. So what will probably happen is we’ll know about both of these URLs. So if you do as site query for one of these URLs then we’ll show that to you, because we know about it and we think you’re looking for that. But what will happen is we’ll concentrate all of the signals that we have on the designated canonical. So we’ll pass anything we know about links, about anything else around that site, how we want to show it in rankings, the relevance of that site to your canonical– the one that we pick up for indexing. But we may still know about the non-canonical version. So if you do a site query, you’ll probably still see that. But that’s not a sign that anything is broken. So there’s nothing you really need to force, there. Nothing you need to kind of like manually take that out of the index or anything like that. We’re probably already picking that up properly. A good way to kind of check this, which works in a lot of cases, is to look at the cached version of the page. And if the cached version of the page tells you on top the URL is the preferred canonical that you specified, then were already indexing it like that. We’re just showing you the other URL because we think that’s what you’re looking for. But we’re already indexing it primarily with your preferred canonical.