C2090-632 Sample Questions & Answers

Determinants are used to describe the levels of granularity within a dimension. By defining a determinant for the Month level (uniquely identified by Month_Key) and another for the Day level (uniquely identified by Day_Key), you inform the Cognos query engine about the data's structure. When a query involves the Month-level forecast fact, Cognos will aggregate correctly at the month level and will not incorrectly multiply the forecast values across all days.

The Cognos query engine generates a cross-product when it cannot find a valid join path between the tables included in a query. This is most commonly caused by a missing or incorrectly defined relationship between the relevant query subjects. Without a defined join on a common key (e.g., Customer_ID), the engine has no choice but to join every row from the first table to every row in the second.

Parameter maps are key-value pairs stored within the model. They are ideal for managing static lists of values, such as translations for report titles or mapping codes to descriptive names. The session parameter for language can be used as the key to look up the correct title.

Session parameters hold values that are specific to a user's session, such as user name, account information, or preferred language. They are perfect for driving dynamic behavior like security filters, where the user's region can be passed from the authentication source and used in a macro.

This is the standard and most efficient method for handling role-playing dimensions. A single query subject for 'Ports' is maintained. Shortcuts (which act as aliases) are then created for each 'role' (Origin and Destination). This allows for unambiguous join paths to the fact table while maintaining a single point of definition for the port dimension, simplifying maintenance.

Stand-alone calculations can reference query items from any query subject that has a relationship path to the others used in the expression. This allows for the creation of complex business logic (like profit margin, which needs revenue and cost from different subjects) that spans multiple tables. Embedded calculations are limited to items within the same query subject.

This is the direct method for implementing object-level security in Framework Manager. By modifying the security properties of the object itself (the query subject), you can define which groups or roles have access. Denying access to a broad group like Everyone and then granting specific access to HumanResources ensures that only authorized users can see or use the query subject.

This is the recommended IBM practice for managing large, complex enterprise models. Segmenting the model into linked projects addresses all the requirements. It drastically improves performance as modelers only open the segment they are working on. It allows for independent development and publishing cycles. The shared conformed dimensions in the core project ensure that cross-functional reporting remains possible by providing the common link between the business area projects.

This is true. The 'Analyze Publish Impact' feature is a powerful tool that analyzes the Content Store to find dependencies. It can identify reports, queries, and other objects that use the model items you have modified, added, or deleted. This allows a modeler to assess the potential downstream impact of their changes before publishing the package.

Understanding the business context is paramount. Knowing what decisions the data will drive helps the modeler understand the required measures, dimensions, and levels of granularity.

Security requirements must be understood from the beginning, as they can significantly influence the model's design, including join strategies and the need for specific security tables.

The model must contain accurate and agreed-upon business logic. Capturing the precise formulas for KPIs like 'Profit Margin' or 'Year-over-Year Growth' is a core responsibility of the metadata modeler.

For ratios or percentages, the 'Regular Aggregate' property should be set to 'Calculated'. This instructs the Cognos engine to first sum the components of the calculation (sum of all Revenue, sum of all Cost) at a given aggregation level, and only then perform the division. Setting it to 'Sum' or 'Average' would incorrectly sum or average the pre-calculated row-level ratios, leading to mathematical errors.