Aggregate Data and showing Today's Date in Visualizations

Support - Assistance technique
1

I’m experiencing issues whereby if data is not present in any dataSet for “Today,” then running resource tables + analytics tables will not show that data. However if I run resource tables after data is entered, and then analytics tables, I will see that data (and afterwards, any data that is entered will appear as well in analytics). I saw this note in Jira from @idelcano in reference to continuous analytics:

  • New periods that do not contain data are not created in the resources table (analytics_rs_periodstructure, _periodstructure..).

I’d like to confirm that:

  1. @idelcano does this sound like a spin off of what you are experiencing?
  2. This is “normal” behavior for aggregate data, i.e. that you have to run resource tables after data exists in order for analytics to pick it up for Today’s date?
1 Like
2

Yes, this is exactly the issue we are experiencing.
I believe it’s a bug, and I hope the team can address it. If you enter a value in a new period, continuous analytics will not detect it.
However, if you enter a value in a new period and generate the full analytics and then enter more values for that same period, continuous analytics will detect them as expected.

2 Likes
3

I actually have a lot of interesting findings related to this, for what its worth:

  1. Running analytics tables + resource tables (on the analytics side of data administration or through scheduler) does not result in visibility of analytics for us.
  2. Running resource tables is sufficient for data visibility in analytics if analytics has already been run for that time
    1. I.e. if I run analytics expecting to see the data, and then run resource tables when I realize it does not appear, the resource tables run completion is immediately followed by data visibility (without another analytics table running/completing)
  3. Lastly, a little red herring that we discovered in this process is that mobile pushes in the devices created and lastUpdated time stamp, not the servers—so if the mobile pushes in data, the server could skip that data if it is from a time zone ahead.

What I’m going to try today as a workaround is create a dataSet strictly for periodId creation, i.e. a dataSet and then import data into the next month of daily days. I’m hoping that if I do this and then run resource tables, it will create my periodIds and then let analytics run and show data whenever they are entered.

1 Like
4

@Gassim also curious here if you happen to know if this could be resolved by Data Input Periods?

1 Like
5

Hi @Matthew_Boddie,

Thanks for all these details. It seems like a potential bug that the team is reviewing in the ticket you shared above, Jira

Currently, AFAIK, only entering new data that’s linked to a ‘period’ will create that period in the resource tables.

I think your workaround is a good idea for now, did it work?

I think this issue is addressed here and fixed, is this ticket the same: Jira

6

The workaround did work—Created a data set, put it in with dates up to the end of the month, ran resource tables (specifically). After analytics runs, this data appeared. Now, with the “real” dataSet I have, it is appearing as I have hoped.

1 Like
7

To your point on the Mobile data—that ticket seems quite distinct, and indeed working in v40 with the latest mobile version, this is currently happening. I can build out a ticket for it.

1 Like
8

Last comment I’ll make—Mobile data issue for future dates is a big issue, we got around it by (mis)using Predictors. I have a predictor set up to re-enter the same information that exists to the same data element on the same day (without any manipulation). This “resets” the last updated and brings it back to server relevance (as it has a server accurate timestamp) and allows it to get processed as expected.

1 Like
9

Jira for the mobile timestamp

1 Like