You are on page 1of 8

The best practices cover the following major components of Cognos Suite: Cognos Framework Manager Cognos Connection

Cognos Report Studio Cognos Query Studio Cognos Analysis Studio Cognos Event Studio Cognos Metric Studio

COGNOS Framework Manager:

Query reuse can be set on the model or on individual reports. To specify that all reports using a particular model should use cached data, enable the Allow Usage of Local Cache governor on the model in Framework Manager and republish the model. By default, this setting affects all reports that use this model Set Query Processing type as Data Base Only from Limited Locally Resolving the closed paths or Loops between the Query Subjects by creating short cuts or making star schema grouping. Improve query performance by changing the query types.

COGNOS Connection:

Setting the queue time limit refers to the maximum amount of time that a request will wait for a report service connection when all connections are in use. The request will be queued until a connection is available, or up to the configured number of seconds dictated by the queue time limit setting. If no connection is available within the configured time limit, the user will be notified that the request cannot be completed because no resources are available within the given time constraints. In Server Administration configure the category settings to Tuning.

Setting Dispatcher services # of CPUs and memory in relation to other server.

Cognos Report Studio:

Changing the Query Execution Method from default to Concurrent.

By setting Query Local Cache to yes improve the report performance, when query reuse is turned on and you run a report for the first time, the query is stored in the cache of your current session and reused the next time you run the report. Use simple calculations as much as possible. If the report is large splitting the queries into two or more. While working with prompts Use static choices for identical values with Data base. Avoid Cross products and Outer joins.

Cognos Query Studio:

Use Local Cache for the query.

Enable concurrent query execution. When this feature is enabled, the report server determines which queries in the report can be run concurrently. Use governors to reduce system resource requirements and improve performance.

Use aggregations and filters in the reports.

Cognos Analysis Studio: Use custom set allows you to save the filter rules, calculations, and sort that defines a set created in Analysis Studio. Custom sets are listed in the Analysis Items tab and are saved as part of the analysis.

Cognos Event Studio:

Write simple even conditions. Add few tasks to the agent.

Cognos Metric Studio: Create watch point parameters for dash boards Use Parameterized URL methods.

COGNOS Services:

i-flex has successfully executed multiple projects using Cognos. i-flex has experience working with the following in the solutions provided to the customers using Cognos 8 suite of products: Installation, Configuration, Tuning and Administration of Cognos tools. Using ErWin models by importing them into the Cognos Framework Manager. Use of the PowerPlay Transformer to create Cognos PowerPlay Cubes. Report Creation using Cognos PowerPlay. Creation of Cognos ReportNet models using framework manager. Creation of DMRs(Dimensionally modeling relational data) using COGNOS 8 framework manager. Providing OLAP features to end users through Analysis Studio. Reports development using ReportStudio/QueryStudio (ReportNet). User Training. COGNOS Best Practices Some of the i-flex best practices followed in the COGNOS projects are given below: Use Load Balancing servers in a distributed network, it will direct the request to the server that currently has most processing capacity Development, UAT and Production should have same setup

Avoid inclusion of unnecessary table in the Data Models Do not use Cartesian product in a Data Model Use Aggregate tables as applicable Use repositories to store and share Data Models Data returned from data connections should be only as much as required. The number of data connections and queries in a single report should be minimal. Avoid processing of data in Cognos servers, process data in the database / OLAP servers when ever possible. Use sub queries, and co-related queries appropriately to improve performance. Use functions available in database server when ever available. Use multiple tables in a report to display different data instead of including all data in a single table Avoid creating and using custom functions in large result sets. Use Report Templates for identical formatting Sufficient number of test cases need to be designed to test reports and Data Models Ad-hoc queries should be tested determined criteria for performance against pre

The Cognos UAT environment should have the same setup as the testing and production environment

Data made available for UAT should be representative of production data to provide adequate opportunities to the users to test various situations Any new report, adhoc queries or changes in queries / reports should be tested in development / testing environment and UAT before moving to production. The production environment and network utilization should be analyzed to decide the appropriate time for scheduling Reports, scheduling reports at peak utilization time and business hours should be avoided if possible. Use standardized report development approach and standardized user interface. Use materialized views for frequently used complex calculations. Use determinants to restrict double counting Business Objects Best Practices The best practices cover the following major components of Business Objects: Business Objects Universe Business Objects Reports Business Objects Supervisor Business Objects Broadcast Agent Business Objects Universe Standardization of Universe naming conventions to facilitate ease of migration into testing, UAT and Production environment. Ideally the universe name should not change from environment to environment, as it will require

modification of universe name in the reports for each environment Locking of the Universe should be locked while importing, so that another user is not able to simultaneously modify it Setting a large array fetch size, which will prevent multiple fetches via middleware to the database for the same query Allocating weight to each table, which is a way to control the table order in the from clause of the query Business Objects Reports To enhance report performance, following guidelines may be followed Data returned in the data providers should be only as much as required by the query Number of data providers in a single report should be minimal, as this will keep the need for synchronization less and not adversely affect performance Lesser number of local variables used Splitting of a large single report into multiple reports Providing multiple report tabs if possible, instead of keeping all data in one tab Business Objects Broadcast Agent Optimize the various components of BCA. It is worth looking at some of the following important areas: o Implement a Switched hub configuration o Use CPU with L2 Cache o Check and optimize the disk performance Run a pilot Evaluate whether publishing is cheaper than bursting Evaluate the refresh frequency Allow scheduling options only to restricted set of users

You might also like