No Data - Change Layer inconsistencies

#1

Hey there,

I’ve come across the following issue when applying a Change Layer to the No Data action of my Tabular View, so I was seeing if this is a known issue and whether there’s a way to incorporate a delay in order to consistently trigger the No Data event.

Basically I’ll set up a flat table along with 3-4 Filter Values connecting to the Bridge Parameters of the Cube Script. The Cube Script itself is what’s populating the results into this Tabular view.

I’ve set up a simple No-Data layer and applied a Change Layer to the No Data action for this Table object. At this point, I’m seeing inconsistencies where the No Data change layer event will be triggered, and then in other cases, it will just display the Table with no records.

I’m able to replicate this scenario simply by changing out my filter selections tp produce a No-Data scenario. It seems like there’s an issue with the asynchronous request not fully acknowledging the response of the full/empty records being returned from the Cube Script?

Has anyone come across this type of issue before, and if so, are there any known workarounds on how to address this?

#2

I had somewhat similar issue but it turned out that there was data in the metric set just not in the column being displayed. if the row has any data then it is not no data.
are you truly getting NO DATA or are you running into a NULL column in a row result like I was?

#3

Hey James,

As far as I can tell they were returning No Data (0 Records) when testing this scenario through both the BI Visualization & SQL Reporting Scripts directly. I’ll run a trace through SQL Profiler to verify the output results and assure there aren’t any NULL columns being returned when the empty table is being produced within the visualization.

#4

ok, just wanted to make sure it was not that issue before others got in deeper.

#5

Unfortunately the test case scenarios were all valid in that 0 records would still intermittently return No Data, and other cases the change layer wouldn’t trigger properly.

I don’t mean to branch this off into another issue, but I’m seeing some similarities with the standard Loading Overlay where it would constantly spin even the full data-set has been loaded into the object. This issue issue is also intermittent. Based on the behaviors I’ve encountered so far, it seems like there’s a possible bug with how these particular actions are being acknowledged? I’m just wanting to verify if you’ve encountered this type of issue through your instance as well?

#6

Hi Curtis,

Something like this probably requires deeper investigation. I’d recommend getting in touch with support to try to help. If you can find a way to isolate the behaviour in a simple example to send support, it will go a long way too.

#7

Thanks Jeff.

Yeah, the issue seems to be consistent throughout a few of our Dashboards, so I will reach out to Dundas Support to delve further into this issue.