Google's XML-based Scholar harvesting was implemented, as far as I know, as a proof-of-concept, so you'll need to contact them with the URL in order to have your site indexed. The URL to the Google Scholar extract page is something like http://[url-to-journal]/index.php/[journal-path]/gateway/plugin/GoogleScholarPlugin
, where [url-to-journal] and [journal-path] are of course adjusted to suit your own installation.
Since there were a few minor bugs in the 2.1.1 release of the Google Scholar plugin, I'd suggest downloading version 1.10 of GoogleScholarPlugin.inc.php from http://pkp.sfu.ca/cvs/cvsweb.cgi/ojs2/plugins/gateways/googleScholar/GoogleScholarPlugin.inc.php
and replacing the version in your installation with it.
I believe Google is moving away from XML harvesting and towards a meta tag-based approach in the future, which we've accounted for in some of the new features in OJS 2.2, which will be released soon.
Public Knowledge Project Team