April 2019, by Paul Hausser, Envisn, Inc.
Most institutional data used for reporting purposes comes from one or more data warehouses which are typically aggregations of internal transaction systems like SAP, Oracle, etc. But more frequently today an equally large portion comes from external sources. This could be vendors/ suppliers, government sources, customers, etc. In fact, this segment of data is typically growing at a much faster rate than internal sources as companies need for this data becomes critical to just in time supply chain management as well as quality control.
While use of external data sources has become essential across many industries it also poses challenges because the data may not have the same level of quality as that of internal data. And if there is a problem with a given data source, whether internal or external, is discovered it’s important to know immediately where and how it is being used. Beyond that there are other questions that may need to be addressed such as:
- What data sources are being used across Cognos Analytics?
- Are these data sources internal or external?
- Where and how are they used within Cognos?
- What data items from these data sources are being used?
- Which objects in the Content Store are using data from a given database?
- With what frequency is this data being viewed? And by whom?
Little, if any, of this needed information is available within Cognos itself beyond which data sources are being used. For this you will likely need a third party solution to answer these questions. In this case here we are using the NetVisn product to get data source and related information on actual usage within the Cognos environment.
If a major problem with a data source is identified it can be important to immediately identify its impact and then take action to correct or isolate it. This requires both speed and precision for it to work successfully.
Figure 1 shows some of the data sources being used in a Cognos environment. The SQLNORTHWIND.NORTHWIND.DBO data source has been identified as having a problem with some of its data. The ORDER DETAILS table within this data source has incorrect data and needs to be replaced.
Our next step (Figure 2) is to do an analysis of SQLNORTHWIND.NORTHWIND.DBO data source and identify all of the objects in the Content Store that are using any of the data items from its ORDER DETAILS table.
The results of our analysis are shown in Figure 3 shows some of the total of 45 items found that are using data from the ORDER DETAILS table within this data source.
The final step would be to simply do a model item mass update replacing all of data items using that particular table with the new, corrected version from that data source. This step will also automatically validate the data lineage insuring that the report object will function correctly.
What have we done here? We have identified a data source in the Cognos environment where there is a problem with a data source. We then did a dependency analysis to identify all objects that are currently using this data. This dependency analysis further identifies each data item within the data source and exactly where it is being used. Then it’s simply a matter of updating the FM model with the new, correct data source.
Data provenance is something every Cognos administrator needs to be concerned about and that requires fully understanding the entire data chain from original source to end point consumption in reports, graphs, etc. Plus, it just makes sense to know your data’s provenance; where it came from and how it was created.
© Envisn, Inc. – 2019 – All Rights Reserved. Cognos Data Management