Headers

You can use request headers to control the following service behaviors:

Some of these behaviors can be controlled with query parameters instead of in headers. See Custom query parameters .

Hybrid Data Pipeline buffers the results of an OData query, allowing clients to page back and forth through the results using the $top and $skip system query parameters. The $top parameter specifies how many results to return in the first response and $skip specifies where to start in the result set to return the next set of results. When the Hybrid Data Pipeline service receives an OData query for which it has a buffered result and the $skip query parameter is either not specified or is set to zero, Hybrid Data Pipeline can page back to the beginning of the buffered result or execute a new query.

By default, Hybrid Data Pipeline treats a query where $skip is missing or set to zero as a request to re-execute the query in the backend data source. You can change default behavior in the data source definition, or in the request with the ddcloud-refresh-result header. The header value overrides the setting in the Refresh Result field of the data source definition.

The credentials for the backend data source can be stored in the data source definition on the General tab. If they are not, you must supply them in requests using the ddcloud-datasource-header header.

The credentials for the backend data source can be stored in the data source definition on the General tab. If they are not, you must supply them in requests using the ddcloud-datasource-password header.

In some cases, the Hybrid Data Pipeline OData service can optimize requests to the backend data store when you use the ddcloud-top-mode to specify how a client will be using the $top system parameter to page through results. A value of 0 indicates that the client will use $top to limit the result set and will rarely request the remaining entities. A value of 1 indicates that the client will often use $top and $skip to page through results.

Hybrid Data Pipeline applies the optimization only to queries that meet the following conditions:

When the conditions are met, Hybrid Data Pipeline will generate only a SELECT statement that includes the data store-specific syntax for limiting the rows returned. If the client queries the same entity collection again but specifies $top and $skip to fetch more entities, the service executes a new query. The results might contain some of the entities already received from the first request.

In the following example, the ddcloud-top-mode is set to 1, directing the Hybrid Data Pipeline service to fetch the complete result set and not to attempt optimization:

The OData 4.0 specification defines a Prefer header, odata.maxpagesize, that can be used to control the page size for server-driven paging. In server-driven paging, the server returns partial results and includes a link the client can use to get the next set of results.

You can set the page size in the data source definition, on the OData tab, in the Page Size field. The request header value for odata.maxpagesize overrides the value specified in the data source definition. In the following example, the maximum page size is set to 4000, resulting in up to 4000 entities per page.