Client Profile
The client is a travel technology company headquartered in Texas, USA, that operates a travel marketplace and provides SaaS-driven solutions to travel professionals. Having a global presence with offices in multiple countries, their digital platform offers a wide array of travel inventory and tools, including AI-powered booking assistance and serves a large network of travel advisors and other users.
Business Challenge
As the business scaled up, there was considerable load on site and services experienced issues with latency, autoscaling and load balancers. CI/CD was not set up resulting in deployment issues, version maintenance, rollback and downtime issues. Services also experienced issues with cache memory due to inter-connection issues. Multiple such scattered issues affected negatively on user experience, performance, uptime and cost.
LogusIMS Solution
LogusIMS deployed a comprehensive DevOps team to streamline, stabilize and improvise the scattered infrastructure operations.
Key elements of Solution
- Implemented Jenkins CI/CD system and maintained all yaml files in GitLab to track changes ensuring version control, easy rollback and controlled-quicker-stable deployments.
- HPA auto-scaling was set-up so that whenever there is an increase in load, another node/pod will be created automatically.
- Deployment configurations were tuned to ensure zero downtime and seamless transition of traffic from old pod to new pod
- RabbitMQ (messaging queue service) was re-configured at cluster level resulting in high availability
- Redis (cache memory service) was re-configured at Redis cluster level achieving faster read & write operations, data persistence & replication which are crucial for data safety and high availability
- Created a load balancer at cluster IP level which helped to divert load to micro services evenly. This helped in avoidance of unnecessary load balancer creation at microservice level, ensuring savings in resources & cost.