The customer is a global leading food & beverage company, which provides software & services to corporates for managing their distributor’s network and implementing downstream data management systems.
The customer had conceptualized a business model towards building an analytics system that could bring about 1100 – 1200 distributors for their products. Both sales representatives and distributors used a software through which they could report day – to -day work to the area sales manager. Daily sales figures got synced and the service officer could view the summary at a glance.
The users (distributors and sales representatives) updated the software daily, which was done only at the end of the day. This gave rise to some serious performance issues at the server level, leading to delays in payments to be made by the cash flow distributor.
Blazeclan conducted extensive consulting workshops and brainstorming sessions to study and evaluate the customer’s current infrastructure. Below are the findings from the assessment.
- Peak traffic during data upload resulted in performance issues at the server level of the software.
- Based on the analysis of usage parameters, it was observed that the infrastructure provisioned was built using compute-intensive instances than the memory intensive ones.
Blazeclan’scertified SAs recommended to decouple the reporting and application servers for overcoming the performance issues. They workedclosely with the customer and formulated an approach involving the restructuring of their infrastructure for bringing high availability and better performance.
Following changes werere commended and executed by the team for the same:
1) Application Level Changes:
- The application and reporting server were decoupled and placed onto separate servers, to ensure high availability and to remove performance bottlenecks.
- Database performance was fine-tuned by setting up the striping EBS volumes and the changes were at the block level storage for windows server to achieve right throughput for DB queries
2) Infrastructural Redesign:
- The compute-intens iveinstances with replaced with the memory-intensive instances. The recommendation of dedicated hosts was also implemented to ensure that the owned MS SQL licenses were not redundant.
3) Migration Strategy:
- A cutover strategy was designed to migrate approximately 1 TB of data to keep the transition time to its minimum. This also helped in overcoming the challenge of low internet connectivity,
- Data backup was taken using AWS Import/Export and then the data sync for delta was done over the internet. This also helped to ensure minimum downtime in the entire process of migration.
- Trend Micro deep security feature was implemented to secure the AWS infrastructure. This lowered the risk, as instances automatically werer ecognized and were secured, upon their launch. The feature also covered the following aspects of security:
- Integrity Monitoring
- Web Reputation
- Intrusion Detection and Prevention
- Bidirectional Stateful Firewall
- Log Inspection
Blazeclan also assisted in charting out a disaster recovery strategy and roadmap for the customer.
Benefits Achieved by the Customer
- High Availability: The customer achieved a highly available systemthat could meet and cater to more than 200 concurrent users.
- Scalability: The customer achieved the ability to scale the application as and when required by leveraging auto-scaling and load balancing features of AWS.
- Improved Performance: The bottlenecks at the server level were removed and the response time was increased.
|Amazon Route 53