by Rick Ryan - Envisn, Inc.
In our last blog - Demystify Your Cognos Content Store Part 1 - we covered the role of the Cognos Content Store, what’s in it and why its content is important to administrators. Now we are going to focus on:
- How the Cognos Content Store is structured.
- How to access the Cognos Content Store for key information.
- Some of the challenges in doing this.
The metadata in the Content Store is organized in a set of database tables. At last count there were approximately 35 of these. These get created when the Cognos Configuration Manager is created for the first time. As users are added to the environment and more content gets created the Content Store begins to grow in size. Interestingly, this growth often isn’t noticed until it becomes a problem. One of the reasons for this is that there are not any simple or convenient ways to get answers to the basic questions of size, item count, etc. You can do queries into the Cognos Content Store directly to get this information but most administrators find this tedious and time consuming.
While there are other ways that you can directly query these tables for occasional data needed, you may want to think twice before you build any kind of a systemic capability this way. Why? Well, the first reason is that IBM Cognos reserves the right to change these tables and their structure at any time. You could potentially create some query routines or even an application only to find out that a new version release has made it unusable. IBM Cognos recommends that you use their SDK for this if you plan to access the content store in a systemic manner.
But there is a second, and perhaps more important reason that doing queries against the Cognos Content Store database is not the way to go if your plan is to use this approach for a large number of interactive queries. That is that the Cognos Content Store is not optimized to handle large numbers of administrative queries. It is purposed differently since its primary role is to handle high speed reporting for BI users. It is unlikely that it could ever directly handle a high volume of interactive queries without creating contention with user requirements.
IBM Cognos provides a Content Manager Browser Tool on its web site that can be downloaded and used by administrators to get at some of the needed information. This offers the user a tool that can meet some basic, though limited, needs. It can give you a summary of the number and size of the various objects and it then displays these in a simple format. The user can also query folders and types of objects. The user would have to do this on a regular basis and then build a basic data repository to track this over time in order to see any meaningful trends. While this approach will certainly work, it is tedious and provides only limited information. Typically, once you get past size and the number of objects the next questions are why and where. These bring on another level of complexity.
In response to the obvious question; “How large is a typical Content Store?”, the answer is …. It depends. A Cognos environment of 800 to 1,000 users that has been in place for a year or two may have 12,000 to 20,000 objects, and possibly a lot more than that. Since each of these objects can have multiple (20+) properties that relate to security, schedules, jobs, etc., the level of complexity goes up rapidly. You can see where this is headed. It’s okay to provide a basic tool to get some data out of the Content Store but if you need to begin to address issues like dependencies, relationships, permissions and changes then a more systemic approach is needed.
In our next blog we will talk about Content Store best practices based on real world experiences of Cognos administrators.
© Envisn, Inc., All rights reserved.