Release Notes for iPlanet Meta-Directory




Version 5.0, SP1 Patch 1

---

These release notes contain important information available at the time of the version 5.0, SP1 Patch 1 release of iPlanet Meta-Directory. Read this document before you begin using iPlanet Meta-Directory.

These release notes contain the following sections:

• Installation Notes

• Problems Solved
  • Multi-value / single-value to no-value problems fixed

  • Other Multi-Valued Attribute problems fixed

• For More Information

Installation Notes

---

For information on installing the version 5.0, SP1 Patch 1 release of Meta-Directory 5.0, see the iPlanet Meta-Directory Installation Guide listed in the reference below:

http://docs.sun.com/db?p=coll/S1_ipMetaDir_50sp1

Problems Solved

---

This section lists and describes the problems reported for this version 5.0, SP1 Patch 1 release of iPlanet Meta-Directory.

Multi-value / single-value to no-value problems fixed

In previous release, the attribute value fails to update in the destination database when value goes from multi-value or single-value to no-value in the source database (i.e. all the value(s) for the attributes are deleted), and the destination database continues to have the value(s) for that attribute. This problem appears for both the type of data flow, i.e. from the external data source to connector view (read cycle) as well as from connector view to the external data source (write cycle). This problem was logged as 4538869 "Attribute value fails to update when value goes from 1 or many to zero or null" and was subsequently split into three as follows:

4684959 - Active Directory Connector

4684962 - NT Connector

4684965 - Universal Text Parser

The above problems (4538869, 4684959, 4684962, 4684965) have been corrected in this release. Some configuration changes are needed for the fix to take effect.

Problem (# 4684959) Active Directory Connector
The following two new parameters have been added to the [general] section in adc.ini file.

1. usermultitonovalattr : this is the comma separated list of the AD User attributes which can go from multi-value or some value to no-value.

2. groupmultitonovalattr : this is the comma separated list of the AD Group attributes which can go from multi-value or some value to no-value.

For example,

usermultitonovalattr=telephoneNumber,mail,facsimileTelephoneNumber

groupmultitonovalattr=uniquemember,description

Problem (#4684962) NT Connector
The following two new configuration parameters have been added within ntdc.conf file :

1. NT User No Value Attributes : this is the comma separated list of the NT User attributes which can go from multi-value or some value to no value.

2. NT Group No Value Attributes : this is the comma separated list of the NT Group attributes which can go from multi-value or some value to no value.

   For example

   NT User No Value Attributes=mail,description,ntUserScriptPath

   NT Group No Value Attributes=uniqueMember,description

Problem (#4684965) Universal Text Parser (UTP)
The following new configuration parameter has been added within task.cfg file :

MultiValToNoValAttr : this is the comma separated list of the attributes from external data source which can go from multi-value or some value to no value.

We now give examples of how to use the configuration parameter - MultiValToNoValAttr for the 3 different types of data format.

1. CSV data

   MultiValToNoValAttr=EMAIL,DESCRIPTION

2. LDIf data

   MultiValToNoValAttr=mail,description

3. NVP data

   MultiValToNoValAttr=mail,description

The configuration parameter "MultiValToNoValAttr" has no effect in the write cycle when a configuration parameter "lineformat" or "importlineformat" is specified for the write direction. If the configuration parameter "IgnoreEmptyFields" is set to TRUE, then the attributes with no value does not appear in the .out file.

NOTE: The following note applies to all of the above three problems (# 4684959, #4684962, #4684965). The attribute names listed against the above parameters should be the attribute names used in the external data source and one should not specify the attribute names used at the connector view end.The respective configuration files (ntdc.conf,adc.ini,task.cfg) would contain the new configuration parameters mentioned with no values listed against it. Its the responsibility of the user to configure these parameters for the given connector.

Other Multi-Valued Attribute problems fixed


Problem (# 4706047) Multi-value to Single-value problem in Universal Text Connector during read cycle
In previous release, when any multi-valued attribute in the external data source goes from multi-value to single-value then the deletion of the corresponding values in the connector view for that attribute does not succeed in some cases.

The problem has been corrected in this release. Now all the multi-value to single-value updates for any multi-valued attribute in external data source are propagated properly to the connector view.

Problem (# 4540428) Support for multivalued constructed attributes.
Join engine now supports flow of multivalued constructed attributes. All the values in the multivalued constructed attribute now flow from CV to MV and from MV to CV based on the rules created.

Problem (# 4539213) Multivalued attribute problem in UTP during write cycle
In previous release, in UTP when data flows from connector view to external data source (which is csv.out, nvp.out or ldif.out text files in case of UTP), if any attribute in the connector view has multiple values, UTP sends only the last value of the multiple values of that attribute to the external data source.

In this release, UTP writes all the values for a multivalued attribute to the external data source (the .out file). The format in which the values of the multivalued attribute appears in the .out file depends on the certain configuration parameters in the task.cfg file. Some of the important configuration parameters are discussed below for each type of data:csv data

1. multivalued attributes can be written to the .out file in 2 different formats
   1. JoinList : Multiple values of the multivalued attributes can be joined together and written to the out file as a single value. This can be achieved through a configurable parameter called JoinList in task.cfg. For example,

      "Pat's Description1;Pat's Description2"

   2. Multivalue List : If multivalued attribute are not configured to be written to the out file as a JoinList then multiple values for that attribute is written to the file as a Multivalue List prefixed with "[L]" tag. For example,

      "[L]test1@test.com\,test2@test.com"

   Note that "," (comma) is a field separator for csv (comma separated values) records and a configuration parameter "EvaluateEscapes" should be set to TRUE in task.cfg file.

2. For nvp data multiple values for the multivalued attributes appear in the out file as multiple attribute-value pairs on separate lines. For example,

   mail=abc12@test.com

   mail=abc123@test.com

3. For ldif data multiple values for the multivalued attribute appear in the out file as multivalued list with the "[L]" tag.The configuration parameter "EvaluateEscapes" should be set to FALSE in task.cfg in order to differentiate between the comma used as a multi value separator and the comma in the value itself.

   mail: [L]abc12@test.com,abc123@test.com

   description: [L]first\, desc,second\, desc

For More Information

---

Useful iPlanet information can be found at the following Internet locations:

• iPlanet Meta Directory Documentation --- http://docs.sun.com/db?p=coll/s1_meta

• iPlanet Product status --- http://www.iplanet.com/support/technical_resources/

• iPlanet Professional Services information --- http://www.iplanet.com/services/professional_services_3_3.html

• iPlanet Developer information --- http://developer.iplanet.com/

• iPlanet Learning solutions --- http://www.iplanet.com/learning/index.html

• iPlanet Product data sheets --- http://www.iplanet.com/products/index.html