https://spreadsheets0.google.com/viewform?formkey=dDZmXzB5cjV2dHBDSmttSnA0UEZyUXc6MA..

Hidden Feature 1 – Ranking Framework

In the past couple of years, we often get requests for the GSA to be improved based on the popularity of certain sites or pages.  With the Ranking Framework, you can do just that.  The Ranking Framework you can specify boosting:

1. Based on a url pattern
2. Based on a specific url

These can been written to a file and via an ATOM post, the GSA will fetch this file and added it to a result policy.  The following is an example file Google’s documentation:

http://www.important.com/ 1
http://very.important.com/ 3
http://not.important.com/ -1
http://www.important.com/personal_stuff –3

You’ll then need to go and save the policy file to crawl the rescoring to occur.  The API include adding and deleting as well.  Good Luck!

The Google Mini along with the MC+A Google SharePoint Sitemap can effectively be configured to crawl Microsoft SharePoint sites.  The MC+A Google SharePoint Sitemap produces a dyanmic site map which the Mini can use to crawl deep within SharePoint.   You can further enhance this by adding a Content Web Part onto the search page which will allow you to have the Google Mini produce search results within SharePoint.

Here’s it working in our environment.

and in edit mode:

How is the integration achieved?

Step 1 – Navigate to the search results page
Step 2 – Switch the search results page to edit mode
Step 3 – Delete the SharePoint search web parts
Step 4 – Add content editor web part
Step 5 Add the following code(modify for your environment)
Content for web part

How does this work?
1) SharePoint search web part reads the k parameter query string and makes an ‘ajax’ type of request to the appliance.
2) The end use broswer communicates directly with the Mini.

If you purchase a Google Mini from MC+A, we offer a free GSS with support matching for the term of the Google Mini. Otherwise, the product lists for $500 and includes a 30 day trial support during the trial. MC+A professional services can be engaged to assist with the integration within SharePoint.

Many companies have spent millions of dollars to integrate enterprise-wide data storage units which include CRM systems, ECM systems, shared file systems, databases, web servers and many others. Each of these systems performs different functions, but they do share something in common: they store tons of information. As a company grows, so does its document numbers. According to studies done by IDC, content increases by over 80% per year. In addition to that, 66% of companies say they have 5 content repositories or more. With all the development of new data and the multiple locations to store the information are workers really finding the most relevant and beneficial information possible? Probably not.

How are workers finding the correct content to do their jobs more effectively? In many cases, with a multitude of companies, they can’t. Even knowledgeable workers who are hired on to perform a specific task due to their experience in it, they may not be able to find the tools necessary to equate their work with their particular company. What does this mean? Well often times it means duplicated efforts, loss in productivity, and wasted time (which more importantly equates to, uh oh, wasted money). Can a company afford, especially in these financial times, to not be fiscally responsible?

What should a company do in order to prevent this waste? Well multiple companies have implemented enterprise-wide solutions that involve extensive hardware and software implementations. This takes its toll on an organization as well. With the high cost of the technology, the system developers to implement, the administrators to ensure the systems are running properly, and the time to go through the entire development lifecycle, a company can spend literally millions of dollars to deploy. This is not even the greatest of all problems. The question remains: after designing, developing, and deploying the system, is it really going to provide knowledge workers with the correct information? If not, then why start in the first place?

The solution that is both economical and user friendly is the Google Search Appliance. End-users and knowledge workers will definitely benefit from using a system that is just like Google.com. They are already familiar with the user interface and will have no trouble searching for documents behind the firewall instead. Wasted time and efforts will be greatly decreased because the Google Search Appliance will crawl the entire enterprise repositories and find the information they are looking for. How will this happen? Google is the leader in search on the web and they have done so through speed and accuracy of searches. Thousands of developers have tuned the search algorithms to ensure the relevance and the speed of the Google Search Appliance. This means that no time will be wasted within an organization developing and tweaking algorithms to receive the most relevant content.

The Google Search Appliance due to its price, lack of labor needed to install, and end-user friendliness makes it the most powerful and best tool for enterprise search. An organization can really benefit right out the box from the Google Appliances. The knowledge workers will be able to find the information they need to make informed and better decisions without wasting time creating new content (especially since more often than not it has already been created). Business owners, executives, and managers can feel confident that employment decisions were accurate. Knowledge workers rejoice, assistance for projects comes in a little yellow box.

Understanding how the Google Search Appliance works is part of what my job entails. Every time I hear something in relation to the Google Search Appliance or the Google Mini its usually their ability to keep secure information secure. From the Google Enterprise Search site, they provide a list of ways in which the Google Appliances can protect important and confidential information.

Here is a list of ways that Google Enterprise Search will protect sensitive information:

1. The Google Appliances work already existing security policies within the organization.
2. Kerberos Support: Provides native support for Kerberos, enabling a silent authentication process for end-users
3. Metadata Biasing: Administrators can bias results based on metadata (in addition to biasing on source, URL or date)
4. Advanced Reporting: View and export daily and hourly result sets, top queries, special feature usage and more. Report for every query, including reports on which queries receive no clicks by a user and how often the user is clicking on sponsored links in comparison to organic search results or OneBox modules.
5. Secure Web Content: Enable secure searching of information protected by basic HTTP authentication, NTLM versions 1 and 2, as well as Windows Integrated Authentication
6. Enterprise Single Sign-on: Integrate with forms-based single sign-on security systems, included Oblix and Netegrity, to enable seamless searching across secure content
7. Secure Content API: Search across secure content using Google’s SAML Authorization SPI to integrate into existing access control systems
8. X509 Client Certificates: Serve secure results to users using mutually authenticated x509 client certificates
9. Lotus Domino Security: Integrate with Lotus Notes environments using fast, efficient crawling of Lotus Domino servers

Here are nine specific examples of how the Google Appliances can protect sensitive information from reaching the end-users. Coming from a non-IT background, I understand that some of these reasons seem a bit difficult to understand. That is why I added the number one protection item (The Google Appliances work with existing security policing all ready existing within the organization). This means that no matter what environment the Google Appliance is placed in, it will not release information to users who do not identify themselves as an individual who has access to see that information. All this right out of the box!

If more proof is needed you can take a look at Windows Networking’s “10 million documents in a single box” post. It is also understood that security is constantly on Google’s mind for their Enterprise Search as well . There are other high marks that are touted for the GSA such as:

“The Google Search Appliance’s unrivalled security functions were the perfect answer to our requirements and allowed for optimal information sharing”- Administrative and Financial Manager, Societe Generale.

For searching secure content, Google enterprise search seems to be the choice to make. It’s plug-and-play simplicity for security is unmatched in the field of enterprise search. I have seen first-hand the impact that Google Enterprise Search has on an organization and its power and speed are only rivaled by its security.

Next month, I will be analyzing workers and how the Google Search Appliance has benefited them. Happy searching!

Single Sign On – The Problem

Most of the difficulties in completing the form come from small amounts of javascript that is called with the web browser normally submits the form. The appliance does not execute any javascript on the page. It instead simply rewrites the form and presents it to you. Evidence of javascript is when you hover over you see javascript:submitLogin() in your browsers status window.

Single Sign On – The Solution

Using a tool like HTTP Analyzer you can examing the posts and responses during the normal login process. By taking note of the form variables submitted you can recreate this if necessary during the form wizard process.

Be careful to note what the javascript is actually doing. In most cases, there is no logic in the javascript. In that case, you may use the next technique to complete the wizard successfully.

Completing the wizard:

1. Turn off Styles

   Sometimes developers will use Javascript to turn on and off styles. By turning off all styles, you’ll see areas of the HTML that is hidden by CSS styles. In some cases, simply complete the form and save.

2. Editing the HTML that the Google Search Appliance returns to you.

   If you submit a ticket to Google Support with javascript as the cause, the first response will be to see about getting the javascript removed. This will involved multiple internal departments within your company or your clients company.

   I find that this can be difficult at times. At best, a company can roll a version of the form that just the GSA sees. Since you may need to reset the form in the future, a simple swap form temporally will need to be repeated in the future and this too can cause internal stagnation.

   You can save yourself the trouble by simplifying editing the HTML that is returned by the appliance. If you remember earlier, I said to note what actual form variables need to be sent. Simply find the form and make appropriate changes to mimic the original submission. For example, your GSA returned form could look like this.

   <form action="/EnterpriseController" method="post" name="frm1" onsubmit="defaultSubmit(); return false;" />
   <input name="cookieUrl" value="http://sso.mysite.edu:80/amserver/SomethingWrong/" type="hidden" />
   <input name="GoogleOriginalFormMethod" value="GET" type="hidden" />
   <input name="cookieLoginSubmit" value="true" type="hidden" />
   <input name="actionType" value="ssoSites" type="hidden" /><input name=”cookiePattern” value="http://portal.mysite.com/" type="hidden" />
   </form>
   

   Simply change it to this:

   <form action="/EnterpriseController" method="post" name="frm1" onsubmit="defaultSubmit(); return false;">
   <input name="cookieUrl" value="http://sso.mysite.edu:80/amserver/CorrectFormLocation" type="hidden"/>
   <input name="GoogleOriginalFormMethod" value="POST" type="hidden"/>
   <input name="cookieLoginSubmit" value="true" type="hidden"/>;
   <input name="actionType" value="ssoSites" type="hidden"/>
   <input name="cookiePattern" value="http://portal.mysite.com/" type="hidden"/>
   </form>
   

   Then, make sure that all of the text field, hidden and otherwise are included in this form.

   Submit and save.

3. Clicking back and save

There have been times where simply clicking the back button and saving and done the trick.

Of course, MC+A can help you with a detail explanation of this during one of our Search Health Checks or other related support services. Feel free to Contact us.