Like BMW with its automotive Mini, Google thought outside the box when it introduced an engine of a completely different stripe. Packaged in a slim, rack-mountable chassis, this one lets you search for data within your own organization. And like that other Mini, Google has attracted devoted followers – particularly, small businesses that yearn for big business search power at a more affordable price.
How Many Minis?
In early 2007, Google released the third generation of its turnkey search box. Building on the Mini’s claim to bundle enterprise functionality for a small business customer, the latest Mini adds several key features previously found in its bigger, pricier sibling, the Google Search Appliance.
As with earlier generations, pricing for the Google Mini begins at $1,995 with indexing up to 50,000 documents and a full year of support and hardware replacement coverage. The Mini does not require an ongoing annual subscription to remain functional, although an optional $995/year extension provides continued access to updates, support and hardware coverage. Note that you cannot extend support on a Mini if its earlier support contract has lapsed without renewal.
Google decided to leave the latest Mini’s physical design well enough alone – it remains a 1U-sized rack-mount box dressed in blue and branded by Google. While Google does not specify the particular hardware components inside the Mini, it runs a customized Linux operating system on what Google describes as “standard” Intel/PC hardware. Despite this, the Mini hardware is not customer-serviceable.
The Mini is designed to run without any display. Unless there is a serious error that needs diagnosing, you interact with the Mini entirely through a Web browser on your network.
Setting Up Mini
Consistent with its turnkey design, the Mini’ setup is very straightforward, and Google provides clear instructions. Most organizations should be able to have the Mini up and indexing their data within minutes of unpacking. The box includes two Ethernet cables, one of which is a crossover cable needed only for initial setup. Using this cable, a secondary network jack built into the Mini and a temporary PC/portable, you configure its primary IP address and other network parameters for compatibility with your existing intranet.
The Mini is a server and, as such, relies on a powerful cooling fan to support operation 24/7. We noted that the previous Mini was surprisingly loud for such a small box, albeit a box that will likely reside in a dedicated server room. As promised, Google has updated the control software, allowing the cooling fan to automatically slow itself down as needed. The Mini is by no means silent, but its decibels have been tamed.
Crawling for Data
The Mini has two primary phases of operation – indexing your data and processing search requests.
To build your index, the Mini will crawl documents available in locations that you configure. You can set the Mini’s crawler loose on any number of network-accessible paths. By defining inclusion and exclusion rules, the Mini can ignore or include specific file types – for example, ignoring all Excel documents and including all PDFs. The Mini is pre-configured with dozens of rules covering most organizations’ typical needs, which are easily changed when you setup the initial crawl.
The Mini can parse over 220 types of files, including HTML, PDF, Microsoft Office, Microsoft PowerPoint, WordPerfect, Lotus Domino, Lotus 1-2-3, compressed archives, all common e-mail formats and so on. Organizations should be careful, though, not to accidentally reveal sensitive information, particularly for Mini installations with public-facing search.
This latest generation Mini introduces two new features that can significantly enhance your index. New document-level security support allows the crawler to index data protected by HTTP Basic or NTLM v1/v2 authentication, as well as LDAP. When secure documents are indexed, they will only be accessible in search results to people who present the necessary credentials originally applied to that document. The Mini does not support any additional authentication systems, however. For that capability, customers should look to Google’s more expensive Search Appliance.
A major addition to the new Mini is the “OneBox for Enterprise,” previously limited to the Search Appliance product. OneBox is an API with included frameworks that allow the Mini to access information from a variety of enterprise data storage systems, such as Oracle, Netsuite, SAS and others.
Administrator-defined “triggers” tell Google’s search engine to activate an appropriate OneBox module, which relays the search query to the business application or database and presents results in the Google search interface.
The Mini crawler can be set to run continuously to catch new or changed data, or full crawls can be run on a periodic schedule. Full crawls ensure that all results are up-to-date but can take more time to process.
Another new feature is sitemap generation, which optimizes your site’s index on Google.com. By exporting your sitemap from the Mini and sending it to Google Webmaster Tools, searches at Google.com will accurately reflect the state of your site to people without access to searching your Mini. Sitemaps can be customized to include or exclude whichever pages you see fit.
It’s All About the Search
Google has a proven search product, and it know it. A major part of the Mini’s appeal is that you can quickly deploy a familiar platform to people, whether within or outside your organization. Many people use Google daily to search the Web – they understand how to input queries and how to interpret the results.
 |
| The Google Mini, a 1U-sized rack-mount search engine. |
The Mini gives administrators a variety of tools to deploy search. The easiest method of all is to simply point people at the Mini’s search page. Out of the box, it looks exactly like the familiar Google search page, with all the usual tabs – “Web,” “Images” and even “Froogle,” plus the addition of a “Local” for queries against the Mini.
If you prefer to provide a local search interface that is more visually integrated with your existing intranet or Internet site, here, too, administrators can choose among simple or more sophisticated solutions. Organizations can deploy a browser- or desktop- based toolbar with integrated Mini searches, providing persistent local search access to internal employees.
Basic Web developers can design their own HTML search forms to query the Mini. Online documentation provides details and examples. Note, though, that customizing the appearance of the input form will not change the appearance of the results output pages.
For that, advanced developers can process Mini search results as XML data. Using XSLT and complimentary server-side languages, Mini search results can be formatted for multi-channel publishing – say, wireless mobile and Internet audiences.
The Mini offers limited control over the order of search results. Google sorts results primarily on relevance, a proprietary algorithm you can’t modify. You can influence the Mini using KeyMatch, which allows you to assign valuable keywords to specific documents; if these documents are delivered on a query using those keywords, they are promoted in the results sequence.
Reportage
Fans of poring over statistics will delight in the Mini’s mountain of reports. All manner of Mini operation can be logged and output, including the current state of the index, any errors encountered while crawling and real-time performance of search results.
A major addition to the Mini’s reports is Google Analytics, typically available as a separate (though free) service. Useful for people who host their Mini on a public Web site, Google Analytics reports on how visitors use your site – where they come from, which pages they visit first, which pages they visit last and which pages they don’t visit at all, among other factors.
Ready to Go
The Mini delivers on its promises, and with this next-generation model, Google continues to add useful features while maintaining the same pricing.
Mini pricing varies only in how many documents it can index, from 50,000 ($1,995) to 100,000 ($2,995) to a maximum of 300,000 ($8,995), all with one year of support. Two-year licenses are also available. An individual Mini can be upgraded to a higher license should it outgrow its original needs.
Aaron Weiss a technology writer, screenwriter and Web development consultant who spends his free time stacking wood for the winter in Upstate New York. His Web site is: bordella.com
Do you have a comment or question about this article or other small business topics in general? Speak out in the SmallBusinessComputing.com Forums. Join the discussion today!
