When you change driver parameters, you tune driver behavior to align with your network environment. For example, you might find the default publisher polling interval to be shorter than your synchronization requires. Making the interval longer could improve network performance while still maintaining appropriate synchronization.
Driver parameters are divided into the following settings:
To configure the driver parameters:
In iManager, select DirXML Management > Overview.
Using the Find, Browse, or Search feature, locate the DirXML Driver for Delimited Text set.
The following figure illustrates using the Search feature.
View the driver overview by clicking the driver icon (Delimited Text).
Access the Driver Configuration page by clicking the driver icon again.
Scroll to the Driver Parameters section, make changes, then click OK.
To make changes to the Driver Settings section, see Driver Settings.
To make changes to the Subscriber Settings section, see Subscriber Settings.
To make changes to the Publisher Settings section, see Publisher Settings.
The following figure illustrates the driver settings and their default values in the sample configuration.
Field Delimiter indicates the character that is used to delimit field values in the input files. It must be one character.
If the values of any of the input fields contain this character, enclose the entire value in quotes to prevent it from being seen as a delimiter.
NOTE: Changing this delimiter parameter to something other than a comma does not automatically change the delimiter character used in the output files when a Subscriber is used. To change the delimiter character in the output files, edit the Output Transform style sheet. The delimiter character is assigned to a variable near the top of that style sheet.
Field Names is a comma-separated list of attribute names that can be referred to in the Schema Mapping rule. In the input files, the fields of the records must correspond to the order and positioning of the names in this list.
For example, if you list eight field names in this parameter, each record of the input files should have eight fields separated by the field delimiter character. On NetWare® and Windows, see sample.csv in the delimitedtext/samples directory for an example. On Solaris and Linux, sample.csv is located in the /usr/lib/dirxml/rules/delim directory.
The following table lists the default values:
Parameter | Sample Configuration Value |
---|---|
Field Names (Field 1, Field 2, Field 3...) |
LastName,FirstName,Title,Email,WorkPhone,Fax, |
Object Class Name is the Novell® eDirectoryTM class name that should be used when creating new objects to correspond to input files.
This parameter prevents you from inadvertently creating a situation in which the driver writes output files that are immediately read in again as input of the same driver.
The default is No. By default, the driver won't load if all the following conditions occur:
If you want to feed the output of the Subscriber channel into the input of the Subscriber channel as a way to detect eDirectory events to trigger other changes in eDirectory, set this parameter to Yes. For example, if you want the Full Name attribute updated when the Given Name, Surname, or Initials attributes are updated, set this parameter to Yes.
The following figure illustrates the Subscriber settings and their default values in the sample configuration.
Output File Path is the directory on the local file system where output files will be created. An error occurs if this directory doesn't exist.
Platform | Sample Configuration Value |
---|---|
Windows |
c:\csvsample\output |
Solaris or Linux |
/csvsample/output |
NetWare |
Specify the volume (for example, sys:csvsample\output) |
Output files have a unique name that ends with the characters in the Destination File Extension parameter. If the output files from a Subscriber are used as input files for the Publisher of another Delimited Text DirXML Driver, the destination file extension must match the source file extension parameter of the second driver.
When the Destination File Character Encoding parameter contains no value, the default Java character encoding for your locale are used.
To use an encoding other than the default for your locale, enter one of the canonical names from the Supported Encodings table.
NOTE: The Publisher and Subscriber can use different character encodings.
This parameter determines the maximum number of transactions that are written to a single output file. When the file transaction limit is reached, the file closes, and a new file is created for subsequent transactions. If you don't want to limit the number of transactions that can be written to a single file, leave this parameter blank or set it to zero. For more information, refer to Maximum Time in Seconds Before Flushing All Transactions.
If no new transactions have been written to the output file in the amount of time specified in this parameter, the file is closed. When new transactions need to be written, a new output file is created. If you don't want to limit the time that can pass before the output file is closed, leave this parameter blank or set it to zero.
If a value is supplied for this parameter, the current output file is closed at the specified time each day. Subsequent transactions are written to a new file. This parameter does not prevent the Maximum Number of Transactions or the Maximum Time in Seconds parameters from also acting as output file thresholds. If you use this parameter and only want one file per day, set the other two parameters to zero.The format of this parameter can be HH:MM:SS (using the 24-hour clock) or H:MM:SS AM/PM. An hour is required, but the minutes and seconds are optional. Because the parameter assumes local time, any time zone information included in the value is ignored.
NOTE: The previous three parameters (Maximum Number of Transactions Per Output File, Maximum Time in Seconds Before Flushing All Transactions, Time of Day to Flush All Transactions) are all capable of acting as a threshold for the transaction size a file is able to grow to, or for the time that it will remain open to accept new transactions.
As long as an output file is still open for writing by the Delimited Text driver, it shouldn't be considered as finalized, and you should avoid opening the file in any other process until the driver closes it. For this reason, one of the three previous parameters must be set to assure that output files don't remain open indefinitely. To avoid this condition, if the driver detects that all three parameters are blank (or zero) it automatically sets the Maximum Number of Transactions per Output File to the value of 1.
The following table lists the Publisher settings and their default values in the sample configuration.
The Publisher looks for new input files in the Source File Path, which is a directory on the local file system.
The Publisher uses only files that have the extension specified in this parameter. After the files have been processed, the value of the Source File Rename Extension parameter is appended to the filename, so the Publisher won't try to process the same file again. If the value of the Source File Rename Extension parameter is left blank, the source file is deleted after it is processed.
When the Source File Character Encoding parameter contains no value, the default Java character encoding for your locale is used.
To use an encoding other than the default for your locale, enter one of the canonical names from the Supported Encodings table.
If the Source File Extension parameter is .xml, the Source File Character Encoding can be indicated in one of two ways:
The DirXML XML parser handles the following character encodings:
NOTE: The Publisher and Subscriber can use different character encodings.
For information on Source File Rename Extension, see Source File Extension in this section (Publisher Settings).
IMPORTANT: If you change the default, use only characters that are valid in filenames on your platform. Invalid characters cause the rename to fail and the driver to reprocess the same file repeatedly.
When the Publisher has finished processing all source files, it waits the number of seconds specified in this parameter before checking for new source files to process.
If your data synchronization goes only one way, disable the channel that you won't use. To disable one of the channels, clear the filters on the channel you don't need and don't specify a path for the input or output directory, depending on the channel.
For example, if you only need a Publisher channel:
On the Filter editor in iManager, clear the filters on the Subscriber object.
Save the changes by clicking OK.
In the Driver Parameters section, scroll to Subscriber Settings and remove the path specified for the Output File Path.
If you only need a Subscriber channel, clear the filters on the Publisher object and remove the path specified for the Source File Path in the Driver Parameters section.
You can use XML files in XDS format instead of comma-separated value (CSV) files with the driver.
Because you generally will want to use this driver only with a Publisher or Subscriber channel, perform only the steps from the section that you need.
To have the driver accept input in XML format, change the input file extension to .xml.
To have the driver send output in XDS format, remove the Event Transform and Output Transform style sheets from the Subscriber channel.
In iManager, select eDirectory Administration > Delete Object.
Browse to and select the SubscriberEventTransformSS object for the driver.
Click OK.
Click Repeat Task.
Browse to and select the OutputTransformSS object for the driver.
Click OK twice.