Data Vault Administrator Guide

6.5
- 6.5.1
- 6.5 HotFix 1
- 6.4.4
- 6.4.3 HotFix 1
- 6.4.3

Back Next

Step 2. Populate the Master XML File

Populate the master XML file with the table attributes and a naming pattern for the source files.

Name the master XML file "

metadata.xml

." All tags in the master XML file must be in UPPER case characters.

The following table describes the XML file tags:

XML Tag	Purpose
IDVSERVICE_DOCUMENT	Root tag for the document. Must be present at the beginning and end of the file. If the tag is not present, Data Vault assumes the document is incomplete.
SERVICE_CALL	Specifies the service parameters JOB_TYPE and SOURCE.
JOB_TYPE	The type of job. Currently only the ARCHIVE job type is supported for this tag.
SOURCE	SOURCE can be either XML or ODBC. If the source is XML, ssaservice expects you to provide the table and column attributes in the master XML file. If the SOURCE is ODBC, ssaservice uses the DSN to connect to the source database and get the table attributes such as column name, data type, precision, scale, and nullability information. Currently, only Rainstor is supported as a source for an ODBC connection.
DSN	The DSN for fetching table attributes from the source database. Valid when SOURCE is ODBC.
DSNUID	User name for the source database. Valid when SOURCE is ODBC.
DSNPWD	Password for the source database. Valid when SOURCE is ODBC.
GLOBAL_PARAM	Parent tag for the global parameters effective for each job. These global parameters can be overridden by table level parameters.
KEEP_DATA	If the value is 0, the bulk file uploader deletes the flat file after the job is complete. For any other value, the source file will be retained after the job is complete. If the job fails due to any reason, the source file will not be deleted.
TIMESTAMP_FORMAT	Specifies the format for timestamp data in the flat file, if the format is different from the Data Vault default format (yyyy-mm-dd-HH.MM.SS.NNNNNNNNNNNN). The timestamp format supports the following special characters as separators: : / . - space For example: yyyy/mm/dd-mm/ss/hh/NNNN or yyyy mm dd-mm ss hh:NNNN
COLUMN_SEPARATOR	String that separates column values in the source flat file.
ROW_SEPARATOR	String that separates rows in the source flat file.
NULL_INDICATOR	String to indicate NULL in the source flat file.
CRYPTO_KEY	Crypto key used for encrypting the SCT file. The crypto key value is visible in the metadata.xml file.
TABLES	Parent tag for a list of tables.
TABLE	Tag for the beginning of the table attribute. The value for NAME is the table name, DATABASE is the database name, and SCHEMA is the schema name.
FILE_PATH	Specifies a directory path with a naming pattern for source files.
SCT_PATH	The directory where SCT files will be created.
COLUMNS	List of columns. This information is used for generating the table DDL and NDL.
COLUMN	Parent tag for column attributes. The NAME value signifies the column name.
TYPE	The source data type. Only Data Vault and Rainstor types are currently supported.
PRECISION	Precision or length of the column.
SCALE	Scale value for appropriate data types.
NULLABLE	1 if the column is nullable, 0 if the column is not nullable.

When you define an attribute at the table level and at the global level, the table level value takes priority.

The XML file must contain all attributes of the columns TYPE, PRECISION, SCALE, and NULLABLE. The bulk file uploader uses the precision and scale depending on the data type. Informatica recommends that you create the columns as nullable. For example, for the DECIMAL data type, the uploader uses both precision and scale. For the CHAR data type, the uploader uses only precision. For the INT data type, the uploader ignores both precision and scale.

When you populate the master XML file, you can use either Data Vault data types or Rainstor data types. If you use Rainstor data types, the uploader converts them to Data Vault data types as shown in the topic "Data Type Mapping."

The maximum length supported by Data Vault of CHAR and VARCHAR columns is 32768. If the precision provided for CHAR or VARCHAR columns exceeds 32768 in

metadata.xml

, the uploader adjust the precision in the DDL to 32768. If the actual data is longer than 32768 characters, a truncation error occurs when the bulk file uploader uploads the data.

Data Vault maps wide characters to the CHAR and VARCHAR types. If the precison of WCHAR or WVARCHAR column is N, the internal precison for CHAR or VARCHAR columns is 2N. The maximum length of WCHAR and WVARCHAR column data that you can load to Data Vault is 16384.

Rename Saved Search

Table of Contents

Data Vault Administrator Guide

Data Vault Administrator Guide

Step 2. Populate the Master XML File

Step 2. Populate the Master XML File