Attention given to the provision of stable service in times of heavy load and to the provision of the service worldwide

Summary

Even though the usual number of accesses is usually low, it explodes several times a time, so we had been operating an expensive, dedicated server that assumes a high load. However, because response degrades when intensive accesses occur, when unexpected accesses occur, the server went down due to insufficient capacity. Previously, the way to handle this problem was restricting access to the site. However by moving to the AWS environment, we could build an architecture that speedily handles spikes in access without restricting access. To take into account access from different countries and various types of images, we have also deployed a CDN so that large images can be sent. As a result, we could greatly hold down the operational cost associated with a dedicated server, and eliminated concerns about the server going down.

AWS Environment BeansTalk,EC2,S3,ELB,AutoScaling,RDS,CloudFront,CloudWatch,SNS,Route53
Feature The operational cost of infrastructure was greatly reduced by linking AWS's high performance service with the Web system.
Our Role Direction, server design/architecture, management program, operational maintenance
Production Period about 1 month