An upgraded version of the Google Search Appliance can handle up to 10 million documents, more than triple the previous maximum of 3 million documents.
Google on Wednesday plans to introduce a new version of the Google Search Appliance that can handle up to 10 million documents.
The previous GSA, Google's search hardware for businesses, topped out at 3 million documents.
"We've made some drastic changes in our server architecture which allow us to scale to 10 million documents," said Nitin Mangtani, lead product manager for Google enterprise search.
Mangtani declined to provide specifics but said that the changes had to do with indexing documents more efficiently and with hardware-based improvements.
GSAs in standard configurations can support up to 30 million documents and can support more in custom configurations.
Most large enterprises have six or more data repositories, said Mangtani, citing a Gartner study that also shows enterprise content growing at a rate of 80% per year. So the need for enterprise search isn't diminishing.
Google's goal in improving the GSA is to provide a simple, secure, single-box search system as an alternative to complex search installations that tie together various front-end, index, and database servers.
Users don't want to search different repositories separately, said Mangtani. They want a single point of entry.
In keeping with the universal search paradigm implemented by Google last year, the GSA presents information from multiple sources -- Documentum, employee directories, Filenet, LiveLink, Sharepoint, wikis, and unstructured data, for example -- on a single search results page.
The revamped GSA also knows some new biasing tricks. Search biasing is a way to skew results based on predetermined criteria. Source biasing, for example, allows an administrator to give more weight the source of a document in determining its relevance for a given search.
In addition to administrator-driven search biasing policies, data biasing, and source biasing, the GSA now allows metadata biasing and front-end biasing. It also supports search personalization so that different company groups, such as engineering and marketing, can be steered toward group-appropriate documents.
A recent InformationWeek report took a look at Google, Microsoft, Yahoo, and dozens of other search specialists and their race to develop next-generation technologies that do a better job of getting people the information they seek. You can download the report here (registration required).
Other new GSA features include keyword-based e-mail alerts (as a way to track document changes), search query performance reporting, and Kerberos authentication support.
The administration interface supports four new languages -- English (as written in the United Kingdom), Portuguese (as written in Portugal), Turkish, and Vietnamese -- bringing the total number supported to 27.
Finally, contextual spell checking is now available in Dutch, French, German, Italian, Portuguese, and Spanish, along with contextual query expansion in Dutch.