RavenDB version 2.5. Other versions:

Querying

In order to achieve very fast response times, the RavenDB server instance indexes your data in the background whenever it is added or changed. All indexes in RavenDB are Lucene-based, and we take advantage of this to provide a very fast, feature-rich and flexible querying system. RavenDB allows querying using the Lucene syntax; whether they are sent from the Client API via the Linq Provider or through the HTTP RESTful API, they are being translated into a Lucene query and executed against the appropriate index.

For most operations this is done transparently, and all you need to be familiar with is Linq and some RavenDB-specific operations. However, there are many other handy features that only become available once you understand the way RavenDB operates on its indexes.

In this chapter we explore the various querying options available in RavenDB, from the immediately visible and most frequently used ones, to more advanced topics like Includes, Attachments and more.

Before we start, it is important to understand that all queries sent to a RavenDB server use an index to return results. While you can define your own indexes manually (we will see how later), RavenDB does this for you automatically if you haven't done so.

Therefore, there are 2 types of indexes in RavenDB:

  • Static indexes are named indexes which are created explicitly by the user.

  • Dynamic (auto) indexes are created by RavenDB automatically based on queries. The database is able to analyze your queries and generate the required indexes on the fly, if no matching one has been found. RavenDB will manage priorities of these indexes and even can remove them from the system completely if they have not been used for a long time.

Also worth mentioning at this stage is the notion of stale indexes. Because RavenDB follows the "better stale than offline" approach, querying an index may return stale results - for example when a user queries a database that while a mass-update in progress. RavenDB will let the user know if results are stale, and can also be told to wait until non-stale results are available.

We will start by understanding the full range of querying options supported by RavenDB, and showing how they can be used intuitively by the RavenDB Linq provider. Then we will step in and learn about the Map/Reduce indexes that RavenDB uses and how to create them by hand, and after that we will explore more, less-obvious options supported by RavenDB.

Comments add new comment

The comments section is for user feedback or community content. If you seek assistance or have any questions, please post them at our support forums.

Arun
REPLY Posted by Arun on

This might be a very poor question but even though i need a reliable answer so posting it down here... The question is about when i'm having a document with lot information in it. Then if i need any particular information from the document, is querying down the whole document the better idea or just to pull out only the required details will do it for me.Whether there is any difference between both these ways in performance wise? will the selection of the required information alone be performance boost?

SUBMIT COMMENT