Skip to end of metadata
Go to start of metadata

What is OData

The Open Data Protocol (OData) is a Web protocol for querying and updating data that provides a way to unlock your data and free it from silos that exist in applications today. OData does this by applying and building upon Web technologies such as HTTP, Atom Publishing Protocol (AtomPub) and JSON to provide access to information from a variety of applications, services, and stores. The protocol emerged from experiences implementing AtomPub clients and servers in a variety of products over the past several years. OData is used to expose and access information from a variety of sources including, but not limited to, relational databases, file systems, content management systems and traditional Web sites.

OData is consistent with the way the Web works - it makes a deep commitment to URIs for resource identification and commits to an HTTP-based, uniform interface for interacting with those resources (just like the Web). This commitment to core Web principles allows OData to enable a new level of data integration and interoperability across a broad range of clients, servers, services, and tools.

copied from

Teiid Support for OData

When a user successfully deploys a VDB into a Teiid Server, the OData protocol support is implicitly provided by the Teiid server without any further configuration. OData support is currently not available in the Teiid Embedded profile. Teiid makes use of JBoss AS and its already configured RestEasy and OData4J libraries to provide Rest based access to the VDB. The access would be similar to accessing to any web resources deployed on JBoss AS. OData support is implemented and deployed through a single WAR file across the Teiid system. A user can access this WAR file at <jboss-as>/modules/org/jboss/teiid/deployments/odata-9.0.0.Final.war.

How to Access the data?

For example, if you have a vdb by name northwind deployed that has a customers table in a NW model, then you can access that table with an HTTP GET via the URL:

this would be akin to making a JDBC/ODBC connection and issuing the SQL:

Note that you should fully qualify the table name along with the model name. Also, use correct case (upper or lower) as used in the VDB.

the returned results from OData query can be in Atom/AtomPub XML or JSON format. AtomPub XML results are returned by default.

Seeing "EdmEntitySet Not Found" error?

Then, it means that either you supplied the model-name.table-name combination wrong, check spelling and case. Or your table did not have any PRIMARY KEY or UNIQUE KEY(s) on them. Since OData access is more key oriented, it is * MANDATORY* that every table Teiid exposes through OData must have a PK or at least one UNIQUE key. Note that Teiid does support composite PRIMARY KEYs.

Not seeing all the rows?
See the configuration section below for more details. Issue another call with the same URL, but with the $skiptoken query option specified from the previous result:

Users can submit criteria with along their query to filter the results:

this would be similar to making a JDBC/ODBC connection and issuing the SQL

To request the result to be formatted in JSON, add the query option $format=json

Query options can be combined as needed. For example format with a filter:

OData allows for querying navigations from one entity to another. A navigation is similar to the foreign key relationships in relational databases. For example, if the customers table has an association with the orders table on the customers primary key id, then an OData GET could be issued like:

this would be akin to making a JDBC/ODBC connection and issuing the SQL:

More Comprehensive Documentation about ODATA
For detailed protocol access you can read the specification at You can also read this very useful web resource for an example of accessing an OData server.

How to update your data?

Using the OData protocol it is possible to perform CREATE/UPDATE/DELETE operations along with READ operations shown above. These operations use different HTTP methods.

INSERT/CREATE is accomplished through an HTTP method "POST".

Example POST

An UPDATE is performed with an HTTP "PUT".

Example PUT Update of Customer

The DELETE operation uses the HTTP "DELETE" method.

Example Delete


By default OData access is secured using HTTPBasic authentication. The user will be authenticated against Teiid's default security domain "teiid-security". Users are expected to have the odata role. However, if you wish to change the security domain, manually edit the web.xml file WAR file in <modules>/org/jboss/teiid/main/deployments directory. In future versions more WS-Security based configurations will be provided.

SAML Based Security
To provide Single-Sign-On (SSO) using SAML2 for OData access, please take a look in the Security Guide for SAML Based Security For OData section.


The OData WAR file can be configured with following properties in the web.xml file.

Property Name Description Default Value
batch-size Number of rows to send back each time, -1 returns all rows 256
skiptoken-cache-time Time interval between the results being recycled/expired between $skiptoken requests 300000
local-transport-name Teiid Local transport name for connection odata
invalid-xml10-character-replacement Replacement string if an invalid XML 1.0 character appears in the data - note that this replacement will occur even if JSON is requested.
No value (the default) means that an exception will be thrown with XML results if such a character is encountered.
proxy-base-uri Defines the proxy server's URI to be used in OData responses. n/a
connection.XXX Sets XXX as an execution property on the local connection. Can be used for example to enable result set cache mode. n/a
"Behind Proxy or In Cloud Environments?"
If the Teiid server is configured behind a proxy server or deployed in cloud environment, or using a load-balancer then the URI of the server which is handling the OData request is different from URI of proxy. To generate valid links in the OData responses configure "proxy-base-uri" property in the web.xml. If this value is available as system property then define the property value like below

Teiid OData server, implements cursoring logic when the result rows exceed the configured batch size. On every request, only batch-size number of rows are returned. Each such request is considered an active cursor, with a specified amount of idle time specified by skip-token-cache-time. After the cursor is timed out, the cursor will be closed and remaining results will be cleaned up, and will no longer be available for further queries. Since there is no session based tracking of these cursors, if the request for skiptoken comes after the expired time, the original query will be executed again and tries to reposition the cursor to relative absolute potion, however the results are not guaranteed to be same as the underlying sources may have been updated with new information meanwhile.


The following feature limitations currently apply.

  • Teiid implements the OData V2 specification along with many V3 features, however OData V3 is not completely supported.
  • Blob support for media types are not supported.
  • "$value" construct to retrieve individual column value is not supported.
  • create/update/delete $links is not supported.
  • $expand is not supported.

Client Tools for Access

OData access is really where the user comes in, depending upon your programming model and needs there are various ways you write your access layer into OData. The following are some suggestions:

  • Your Browser: The OData Explorer is an online tool for browsing an OData data service.
  • Microsoft .NET Framework 3.51: the WCF Data Services framework is available as a separate download for .NET 3.x.
  • Microsoft .NET Framework 4.0: the WCF Data Services framework built into .NET 4.0 (in release candidate as of this writing).
  • Silverlight 3: the Data Services client library for Silverlight is available for download.
  • Java: the Restlet 2.0 library for Java (including Java on your Android phone) supports the OData protocol.
  • Java: Use a library like OData4J for Java based access, or any Rest based framework
  • JavaScript: the XMLHttpRequest object is standard in modern browsers or you can use jQuery, which comes out of the box with .NET 4.0 or is available for download.
  • PHP: the Toolkit for PHPprovides OData support for PHP clients.
  • AJAX: if you're using AJAX for ASP.NET, Microsoft provides the ASP.NET Ajax Library for getting to OData.
  • Excel 2010 PowerPivot: PowerPivot comes with OData support built right in.
  • Windows Desktop: LINQPad is a wonderful tool for building OData queries interactively.
  • Shell Scripts: use CURL tool

How Teiid exposes schema for OData

OData defines its schema using Conceptual Schema Definition Language (CSDL). Every VDB, that is deployed in an ACTIVE state in Teiid server exposes its metadata in CSDL format. For example if you want retrieve metadata for your vdb northwind, you need to issue a query like

Since OData schema model is not a relational schema model, Teiid uses the following semantics to map its relational schema model to OData schema model.

Relational Entity Mapped OData Entity
Model Name Schema Namespace, EntityContainer Name
Table/View EntityType, EntitySet
Table Columns EntityType's Properties
Primary Key EntityType's Key Properties
Foreign Key Navigation Property on EntityType, Association, AssosiationSet
Procedure FunctionImport
Procedure's Table Return ComplexType

Teiid by design does not define any "embedded" ComplexType in the EntityType.

Teiid does not define any one EntityContainer that resulted from different vdb models as a default container, so all entities must be accessed using full path to them.
Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.
  1. Jun 26, 2014

    It states: it is * MANDATORY* that every table Teiid exposes through OData interface must have a PK or at least one UNIQUE key

    but a OData entity of Function will be mapped to a Procedure. Confused on trying to access a procedure via OData url call.

    1. Jun 26, 2014

      Can you clarify what is confusing? Procedures are not tables and the results, if they even exist, do not necessarily have the notion of identity.

      1. Jun 26, 2014

        Your right, just confirmed that using a procedure does work. I was thinking something was implied, when it wasn't. Thx.