By Paul Hausser, Envisn, Inc.
Cognos audit data tables remain essentially unchanged since the release of Cognos 8.4 a few years ago so any extension of value from the audit data has to come from using what’s already there more creatively. At Envisn we’ve found some ways to do this by using the Cognos Content Store as a reference data source for the audit data. In this blog we will review three use cases where this can be done to provide greater insight into your Cognos environment:
- Package Usage by Report
- Dispatcher Usage
- Report Run Type
In an earlier blog we mentioned that when used effectively the Cognos audit data can have cumulative benefits; if you’ve created a new dimension or measure you can link it to others to show new information that is the result of one or more of these existing data items used in combination.Doing this over time has positive results. Our products already allow users to see usage views of reports along multiple dimensions including time, department, status (success/failure), etc. that go beyond the standard audit table data.
A logical next step was to begin to look how to find reports using a specific package. Where do we get that? Our NetVisn product provides all of the objects in the Cognos Content Store organized by type and in real time. This includes packages and all the reports dependent on them. So by linking this identifier to each report we can identify the package used. Package usage for a given period would be the sum of all reports/queries used by a given package for that period. See figure 1
Package usage can be valuable to track in most environments. Usage of new packages is an important indicator of its adoption rate and whether data are actually being used. And since we know all the data items in a given package,it’s possible to create a report with a Pareto analysis of actual data usage. We haven’t done this yet as a standard analysis but it can be done with relative ease. This is yet another example of Cognos audit data extensibility when you begin to connect the different elements.
Dispatcher data is something that you virtually never hear about but it’s there. Dispatcher utilization is helpful in understanding the relative load each is carrying across the environments. In a Cognos environment with multiple dispatchers it can be important to maintain some relative balance acrossthem to insure optimal response time for users. This data is captured from COGIPF_HOST_IPADDR in the Cognos audit tables and with little work you can get it organized for use.
Run Type - Batch versus Interactive Reports
Batch reports are virtually all scheduled reports and interactive are ad hocreports. This data is available fromCOGIPF_TARGET_TYPE and the data also shows which studio was used when running ad hoc or batch reports.The distribution of the differences in both run count and run time minutes can be helpful in terms of knowing the distribution of load by ad hoc versus scheduled reports.
Figure 2 is an example of where we have combined audit data for both dispatcher along with run type. The measure is run time minutes. If you look at the data it shows that one of the three dispatcher is handling virtually all the jobs executed during the first half of 2014. Another interesting fact is that over half of all batch or scheduled reports are run between the hours of 6:00 and 8:00 AM while ad hoc or interactive reports are mostly run during the middle hours of the work day from 9:00 AM to 5:00 PM.
The value of Cognos audit data goes far beyond what is available from the audit tables. Somethings to focus on include:
- Using audit data items in creative combinations with one another can leverage its value
- Use the Cognos Content Store where appropriate as a source for reference data
- Create new measures where it makes sense (e.g., session count)
We’ve shown a couple of examples here where this can be done but there a many more that can be created.
© 2014 Envisn, Inc. – All rights reserved.
Comments on this article