The current product design will cause the data for the custom data class to be overwritten for that computer.
When you get new data from a computer, it will put in the most current data and delete the old entry just like it would for other data classes.
An NSE created for Custom Inventory does not update the data class incrementally (NSE always has attribute partialUpdate="false" for custom data class). They always do a full update.
For example if:
- 1st task populated data class for some computer with row1 = AAA, row2=BBB
- 2nd task populated the same data class for the same computer with row1 = XXX
- only row1 = XXX will be stored in DB after execution of 1st and 2nd tasks
A multirow flag means that each computer may populate several rows for this data class.
Usually, the approach that others have taken are:
- Use different custom data classes, so each contains a specific set of results from each location that the script is trying to obtain information.
or
- Make your scripts robust enough to run simultaneously so that when the event (NSE) for that custom data class is sent, it will contain the results from those scripts simultaneously without one overwriting the other because those were received at different times