The attribute configuration for Extensible MAs only allows single-value attributes to be selected. I have managed to export multi-value attributes from the metaverse by manually retrieving the data via SQL (I intend to change this approach to use the FIM web services). This work-around is only good for the initial export of data. My question is, how should I go about ensuring changes to these multi-value attributes in the data source are detected by FIM and flow through to my Extensible MA ? Without including the multi-value attributes in the attribute flow, FIM will not synchronize these attributes.

Hi Barry, Actually you can configure a multi-valued attribute for your extensible MA using an Attribute-Value pair text file (select "File format: AVP" in the "Select Template Input File" panel of the create new MA dialog). This depends on the file format you use, you cannot do this with a CSV file, for example. Try with a template input file like EmployeeID:10 DeltaOperation:Add FirstName:Terry LastName:Adams SomeAttribute: one SomeAttribute: two SomeAttribute: three You will see that 'SomeAttribute' will be imported as multi-valued, and you will also be able to specify that an attribute is multi-valued by clicking 'Edit' in the Configure Attributes page. Cheers, PaoloPaolo Tedesco - http://cern.ch/idm

Thanks Paolo. When dealing with call-based extensible MAs, is it correct to say that the file format (ie. AVP versus CSV) only dictates the format that your "confirming import" file needs to be presented in? In other words, if I change the format of my call-based extensible MA from delimitted to AVP then all I need to do is produce a "confirming import" file according to the AVP format? Regards

Hi Barry, yes, you are correct. The call-based MA does not need to deal with files during exports. Cheers, PaoloPaolo Tedesco - http://cern.ch/idm

This is actually not correct, the CSV format does support multi-Valued attributes. You need to formulate your discovery file appropriately to get the option. This is from the Help Reference in FIM: Multiple-value attributes and their header names must be in sequential order. A delimited field with any attribute cannot appear between delimited fields that compose a single multiple-value attribute. The following is an example of a valid delimited file format containing a single multiple-value attribute (Phone) composed of two delimited fields (Phone, Phone). Joe Stepongzi - Identity Management Consultant - ILM MVP - www.microsoftIdM.com,ilmXframework.codeplex.com

Thanks for the correction Joe. For the example given in the Help Reference, does this only cater for multi-value attributes of which the maximum number is known? ie. phone attributes where there are up to two values? Is this expected to cater for other attributes such as roles, where the number of roles attached to a user may be unknown?