When a Data Connector import rule is set to update a data class or association (standard or reverse) that are multi rowed (one to many), when the import is ran, the old ones are overwritten with the new ones, not appended with the new ones. Or, the same ones are re-added over and over. For example, a computer's Cost Items data class (a foreign key) include five from one day, and the next when additional ones are added, two, only the new two remain.
For data classes, this is working as designed based on the data class's Update Mode value.
For Data Connector 7.1, for associations, this is working as designed. For Data Connector 7.5, the ability to append associations or reverse associations has been added as a new feature. For more information, refer to the following article:
Data Connector Import Rule overwrites imported associations or reverse associations
For multi rowed data classes, change their Update Mode appropriately to append, not overwrite the existing records. It is recommended to use the Append mode for this, which will add new items to the bottom of the existing list. However, if the import rule is ran more than once and the data contains the same items, these will be re-appended a second time, etc., each time the import rule is re-ran. The only workaround for this scenario is to either remove already imported items or set the import rule to not update existing records. More information on how this works can be found here:
Data Connector Import Rule fails to set a foreign key correctly
For multi rowed associations and reverse associations, there is no Update Mode available and therefore subsequent updates to a multi rowed association always completely overwrites all prior associations, not append with the new. Upgrade to Symantec Management Platform 7.5 for this functionality.
For 7.1, a workaround to this is to re-import all prior multi rowed associations, old and new, for any change to the record. This would require the resource to have an ongoing and every-increasing multiple lines in the data. A CMDB rule may also be able to be used instead of an import rule, if the data is already in the database. (This article does not further discuss how to set up a CMDB rule to do this.)
The following is how an example import rule's data would look to enable multi rowed associations to be retained:
Import run 1 data file
Computer Name, Associated Assets, Serial Number
PC1, PC2, 123456
Import run 2 data file
Computer Name, Associated Assets
This then adds PC2 and PC3 to PC1, retaining both.
Note: This must account for any other columns too, which may result in issues itself, such as unique values such as the Serial Number. If issues occur for other columns, a stand-alone import rule may need to be made that only modifies the multi rowed associations, and another that is responsible for updating the other normal fields of the records.