I have a custom list with more than 5000 items. This article: http://msdn.microsoft.com/en-us/library/ff798465.aspx (the last section, "How Does Indexing Affect Throttling?") states:
"For example, suppose you are working with a list that contains 10,000 items. If you were to build a query that returns the first 100 items sorted by the ID field, the query would execute without issue because the ID column is always indexed. However, if you were to build a query that returns the first 100 items sorted by a non-indexed Title field, the query would have to scan all 10,000 rows in the content database in order to determine the sort order by title before returning the first 100 items. Because of this, the query would be throttled, and rightly so—this is a resource-intensive operation.
In this case, you could avoid the issue by indexing the Title field. This would enable SharePoint to determine the top 100 items sorted by title from the index without scanning all 10,000 list items in the database. The same concepts that apply to sort operations also apply to where clauses and join predicates in list queries. Careful use of column indexing can mitigate many large list performance issues and help you to avoid query throttling limits."
I have done exactly what this article says to do...I have indexed the Title field but I am not able to query against the Title field without hitting the throttle limit.
I examined the NameValuePair table in the content DB, and found that the values from the Title field are not there. I have reproduced the same results on another server as well.
Does anyone have an idea of what is going on here? I am not looking for alternative methods to access items from the list. I want to know how column indexes get created and why the example in the article is not working. Thanks!