- Industry: Finance, SaaS
- Challenge: Airtrans heavily uses on-demand vehicle booking application platform that is built on a legacy infrastructure and development tools. There are three different mobile front end clients send data to the booking platform including rider, driver and admin applications. Airtrans was experiencing system crashes and multiple errors, with database failures on a regular basis. They were also experiencing issues with data loading into the client applications in a timely fashion, as well as issues with the overall database backup and recovery procedures. Airtrans wanted to implement best practices to efficiently process all booking requests, fleet and driver management data, and improve Recovery Time Objectives and Recovery Point Objectives.
Airtrans Limo and Tours provide a variety of transportation solutions for individuals and groups. It operates one of the top 10 largest chauffeured fleets in the country and is metropolitan Toronto’s largest provider of luxury transportation. Since its incorporation in 1998, the company has grown from five vehicles to 250 vehicles today.
The Challenge:
Airtrans heavily uses on-demand vehicle booking application platform that is built on a legacy infrastructure and development tools. There are three different mobile front end clients send data to the booking platform including rider, driver and admin applications. Airtrans was experiencing system crashes and multiple errors, with database failures on a regular basis. They were also experiencing issues with data loading into the client applications in a timely fashion, as well as issues with the overall database backup and recovery procedures. Airtrans wanted to implement best practices to efficiently process all booking requests, fleet and driver management data, and improve Recovery Time Objectives and Recovery Point Objectives.
The solution:
When Cloudism was engaged to fix the stability and performance of the existing platform, we realized that fixing it requires more effort, time and money than building a solution based on serverless architecture. Cloudism has developed a Proof Of Concept using AWS Lambda in 2 weeks to demonstrate the benefits of serverless architecture such as faster development, easier operational management, scaling, and reduction in operational costs. Also we have used the combination of Amazon Aurora and DynamoDB to provide better business continuity, Recovery Time Objectives and Recovery Point Objectives.
As Airtrans was pleased with how the Cloudism team designed and developed the POCwhile keeping deliverables on track, provided its approval to build the enterprise level solution using the AWS services. Cloudism built the solution mainly using Lambda, API Gateway, DynamoDB, Aurora, Amazon S3, CloudFront, SNS,and Amazon Cognito User Pool.
Here are core AWS services that were used to meet the application platform requirements:
- JavaScript from client applications exchange the data with a backend API built through API Gateway and AWS Lambda.
- Amazon Cognito for user authentication and management with the help of secured backend API.
- CloudFront is utilized to forward certain paths (e.g. /rest/*) to API gateway, and serve static content from a S3 bucket.
- Amazon S3 for hosting the static website content like HTML, media files, CSS, JavaScript which acts as a front end artifacts.
- Aurora ServerlessSQL database from Amazonto handle relational data and perform complex data queries from the Lambda functions.
- Multi AZ deployments and read replicas are enabled to provide better Business Continuity, RTO and RPO
- DynamoDB for storing copy of booking requests through Lambda functions. Any update in the DB stream publishes a new SNS topic with the message
The Outcome
After processing between 300 and 500 thousand requests per month, only less than 0.0002 percent requests were dropped. The platform has reduced the number of timeout errors by 99%. Airtransalso had substantial cost saving of 70% by moving to serverless architecture. Performance tests revealed significant improvement in latency; and data loading into client applicationswere reduced to less than 3 seconds. In addition, here are some of additional benefits that have been realized:
- Reduced Operations costs – No infrastructure to maintain, still provides ability to scale up or down
- High Availability and Uptime – Reduced system crashes and improved availability to 99.95%
- Better User Experience – CloudFront to deliver multimedia content, and Cognito to support Social and Enterprise Identify Federationhas provided better user experience
- Improved RTO and RPO – Multi AZ deployments, automated backups and read replicas provide better RTO and RPO
- Improved Security – Cognito supports multi-factor authentication (MFA), and encryption of data both in-transit and in-storage
Rexman Mariadas
CEO at Airtrans Limo and Tours
Our mission is to empower organizations to thrive in the cloud…
About Us
