Data source considerations in Data Virtualization

Data source considerations (Data Virtualization)

Data Virtualization supports many relational and non-relational data sources. Connecting Data Virtualization to data sources entails data type mapping.

The following table describes considerations and restrictions for data type mapping in Data Virtualization.

Table 1. Considerations for data type mapping
Data sources	Considerations
Db2® on Cloud	The nchar and nvarchar types are not supported in Data Virtualization as a service.
Google BigQuery	In the Google BigQuery data source,Data Virtualization does not support the use of the SELECT statement for columns with record data types.
Hive	In virtualized tables, you can list only the value of complex data types, such as array, map, struct, and union type. Any other operation on the value of these complex data types is not supported. When you use a SELECT statement or preview LOB data that is greater than 64 kB, Data Virtualization truncates the data to 64 K bytes only.
Informix®	INTERVAL, BIGINT, and BIGSERIAL data types are not supported in Data Virtualization. For more information, see Exceptions occur when using the Connect for JDBC Informix driver. To solve this issue, add an Informix database connection by uploading the Informix JDBC driver. For more information, see Obtain the JDBC driver.
MongoDB	The following MongoDB data types are supported in Data Virtualization as a service: `Int32`, `Int64`, `Double`, `String`, `Boolean`, `Date`, `Binary`. The MongoDB `String` type is mapped to `Varchar` or `Clob` in Data Virtualization as a service based on the sampling of data in the collection. String values that exceed the column size, as determined from the sampling, are truncated. Data Virtualization as a service supports the MongoDB `Date` type range for years 1883 - 9999. An unexpected value might be returned if data is not in this supported range. The MongoDB `Decimal128` type is not supported and is mapped to `Double` in Data Virtualization as a service. After you virtualize content that uses `Decimal128` data, you cannot capture exact precision when you perform arithmetic in Data Virtualization as a service. Other MongoDB data types and datetime string formats are not supported and might result in unexpected result. Fields in a MongoDB document are mapped to columns in a virtual table in Data Virtualization as a service. Columns in the virtual table are unordered and might not match the order of the fields that are defined in the MongoDB document that is returned by MongoDB compass or the MongoDB shell `find()` method.
Netezza®	Blob, XML, and Clob Netezza data types are not supported in Data Virtualization. When you use a SELECT statement or preview LOB data that is greater than 64 kB, Data Virtualization truncates the data to 64 K bytes only.
Oracle	When you use a SELECT statement or preview LOB data that is greater than 64 kB, Data Virtualization truncates the data to 64 K bytes only.
SAP HANA	To configure pushdown of varchar predicates to SAP HANA the following steps must be applied: In the Data Virtualization Run SQL editor, run the following statement to identify the internal Connection ID (`CID`) for your SAP HANA data source: `select * from DVSYS.listrdbc;` Now, run the following two commands to enable pushdown for the CID. In the following example the CID is `SAPHA10005`: `alter server qplex options(add [email protected]_sequence 'Y'); alter server qplex options(add [email protected]_no_trailing_blanks 'Y');` Note: You must ensure that there are no trailing blanks in the varchar data and that the collating sequence in Data Virtualization is the same as on the remote SAP HANA data source.
SAP OData	You cannot preview or query non-readable tables due to the following reasons: The SAP OData data source might have write-only access. In such cases, changing the user permissions does not avoid this issue. The SAP OData data source has read access, but requires filters. This limitation means that you cannot preview data, but you can read it if you specify filters.
SQL server	When you use a SELECT statement or preview LOB data that is greater than 64 kB, Data Virtualization truncates the data to 64 K bytes only.
Teradata	The XML Teradata data type is not supported in Data Virtualization. When you use a SELECT statement or preview LOB data that is greater than 64 kB, Data Virtualization truncates the data to 64 K bytes only.