loader image

Performance Optimization of Analytics Platform

  • Home
  • Performance Optimization of Analytics Platform

Migration of SAP BO from version 3.1 to 4.1

 

About the Client:

Client is an American manufacturer of medical devices and patient monitoring systems founded in 1915. Headquartered in Skaneateles Falls, New York, it was family-owned until it was acquired in 2015.

Background:

Client was working with a legacy system which was not able to propel them for the next phase of evolution in their business. Client had a vision to merge with a larger entity to properly take on the extremely competitive medical devices business of North America. 

Challenges:

Client has been facing a three-pronged challenge with their Visual Analytics 

  1. Performance: Their legacy Visual Analytics system was not able to withstand huge data volume due to which dashboard and report took 15-20 minutes to load.
  2. Security: Security categorization based on departments was missing along with row level security which means anyone from any department can look at all the reports.
  3. Availability: System crash was a common issue with 15 minutes of downtime every second day, as the system was not able to cope with high number of requests from users.

Duration of the Project

4 months

Tools used

SAP BW, BO, Xcelsius,
Webi Report, Apache Load Balancer, Tomcat Web Server

Approach: 

  • Right Angle Solution (RAS) consultants and Business Analysts spoke to different business groups and understood the business structure
  • RAS business analyst also analyzed user behavior based on the system data
    • This was achieved by taking a look at the audit data on how the system has been used
    • RAS also identified the Peak and Lowest point of usage of the system
  • In this process RAS also looked at the existing application architecture and CPU and memory allocation
  • A report was prepared by RAS with critical issues along with the recommendation

Solution:

  • Identified how many services are required for the business, which helped cull the unnecessary services to free up the system memory 
  • Created services based on the usage of the business instead of what was available 
  • Scaled the system with respect to peak demand 
  • Added an Apache load balancer to the Tomcat web server to improve the performance on the front end
  • Utilized the indexed crawling, by reducing the number of calls to the system
  • Integrated SAP BW with BO and implemented single sign on
  • Created user groups and inherited BW security in BO

 

Outcome:

  • Response time for Dashboard and Reports decreased to 1 minute from 15 minutes earlier.
    • Efficiency of 93.33% achieved
  • Row level and department level security implemented
  • System down time reduced to 15 minutes every quarter from 15 minutes every second day
    • Efficiency of 96.8% achieved