BI Data Analytics

Below issues were identified for MotioCI 3.0 upgrade   Content Store corruption --Engage MotioCI Regression Test cases going in Pending State - recycle MotioCI process Not able to sign in URL - IE Cache clearance

Steps for GC logs are as follows: 1) Navigate to /bin folder and locate a file named bootstrap_OSversion.xml file. 2) Make a backup copy of this file. 3) Edit the bootstrap_OSversion.xml file to put the following entry highlighted in bold. This settinng means that Cognos will create upto 10 files with 10MB size. ........ -Xms${dispatcherMaxMemory}m -Xverbosegclog:${install_path}/logs/gc.log,10,10000 -Xgcpolicy:gencon ........ 4) Implement the above setting in all components of the environment(CM, dispatcher) 5) Restart Cognos" Usage: Memory utilization by Cognos to identify patterns and detect rouge reports. Impact: Performance Overhead

Problem:   When Publisher task fails in between for large table due to any issue, it corrupts QVD file. When we try to rerun task after issue resolution, data is lost since QVD was actually incremental load and need to be restored completely.   Solution Best Practices: Set up QVD backups and restoration as Enteprise wide solution Identify QVDs having different types of load Document procedure for support and maintenance Steps: Restore QVD Restart failed task

Error: DPR-ERR-2002 Unable to execute the request because there were no connections to the process available within the configured time limit.   Troubleshooting:   Identify which server error is coming from Identify how frequent it is coming and which service throwing it. Identify type of reports (DQM Vs CQM)   Cause and Solution:   DQM report waiting  for a connection with report service longer than the specified timeout. The successful request flow is as follows   Run the interactive DQM report -> Get a connection with report service -> DQM report runs on Query service -> Query service gives the output to report service for rendering -> user received the output   In this case, the errors are generated at the second step i.e. getting a connection with the report service. This shouldn't be limited only to Query service i.e. your environment seems to be busy interactively which has little to do with the Query servic...

In my view, below are the steps involved in any type of BI project: Planning Prioritization Performing Tasks Iterate based on learning Better we are able to perform a step, subsequent steps become easier to achieve project deliverable and goals.  

  Problem: Even when using Auto generate test cases feature, it may take a while to generate test cases and it may fail for time out. Secondly, this would be frustrating if even after long time, no prompt parameters are imported.   Cause: It's possible its taking so long because there may be a large amount of prompts its going through.   Resolution: Deselect the "Attempt to Import Parameters from Reports, Views and Outputs" option Auto generate and do dry run to see which reports failed for prompts. Update those reports manually or cloning from other test cases to speed up process.      

MotioCI can be helpful to do impact analysis for any Backend DB changes. Steps: 1. Create queries for packages for each query subject. Please remember to put filters to limit data and have good performance. 2. Create test cases for each of them. 3. Run Test Scripts via MotioCI 4. Schedule them to run with comfortable frequency and failure notification. Once a test case fails, you can verify reason for failure (  table lock vs, DB changes) and get them corrected.