Associating certain measures with certain dimensions?

[kgl_m]

Hi all, I have an issue that has been flummoxing me.

I'm working on a Helpdesk project, and for this, we have three measures - Count of tickets, Count of tickets with met SLA, and Count of tickets with missed SLA.

Now, the SLA part is coming from a separate table where we can have multiple SLA rows per Ticket.

I need to find a way to handle this in cube so that we don't end up giving the user a wrong count when they ask for Count of tickets.

One way we've thought about doing this is to somehow disassociate the Count measure from the SLA Dimension and disassociate the SLA measures from every dimension other than the SLA Dimension. This way it would display NA if someone dragged and dropped the Count measure along with the SLA Dimension into the analysis.

So... does anyone have any idea on how to do that? I tried the External rollup feature, but didn't get very far since all our measures are coming from the same view (we've joined the Fact table and the SLA table in the view to get a column which says 0 or 1 if the SLA was met).

Any help would be appreciated.

[Lynn]

I would think you'd have just one measure for count of tickets and that the SLA dimension would be how you can show the count by met vs. not met. I think three measures is just more confusing for users. Am I missing something?

What is the business explanation on why a ticket have multiple SLA rows? Your design sounds like the SLA is either met or it isn't. Maybe there are other attributes on that table that cause multiple rows? You don't have to feed all the measures from the same source so maybe you'll have more flexibility if you split things up?