Catching the wave of Blockchain and NFT, our client has requested a dedicated team from OpsSpark to help them deploy the whole system with dynamic scaling and security guarantees. The system consists of a Landing & Dashboard page, a Unity Game platform, Back-end, and Blockchain components.

• All provisions and deployments have to be done as fast as possible to adapt to changes
• The whole system is designed to dynamically adapt from hundreds to thousands of concurrent users
• The NFT game platform is able to handle up to thousands of concurrent players
• Site performances have been optimized and cached so that users are happy with the loading time
• It’s very challenging to set up, maintain and monitor Blockchain on-chain services, to keep them updated to the network
• The internal components are accessed by limited IP addresses in an encrypted tunnel
• The system is able to detect fraud access and limit illegal incoming connections, in order to avoid DDOS attacks

• System stack of GCP Firebase, GCP LoadBalancer & Instance group, GCP Amor, StackDriver, GCP Storage, etc are deployed
• Instant group for backend is dynamically scaled based on in-coming traffic
• Firebase configurations are optimized in caching and compression
• CDN strategy for caching and optimizing static and dynamic contents is built
• Backup & restore database strategy is delivered
• Firewalls are provisioned and managed by Terraform
• Secret Manager is advised to be used to store confidential data
• To ensure sure no Blockchain components downtime, monitoring scripts are established and DPR for all on-chain services are implemented

• The system is in good shape and can handle a dynamic incoming amount of users
• The game platform can now handle thousands of concurrent users without lag
• The website wins an A for GTMetrix Grade and LCP (largest content element) is less than two seconds
• Deployment time is kept under three minutes without downtime
• The Recovery Time Objective duration is under 15 minutes, with 24/7 online support for the OpsSpark team
• Abnormal and fraud access can be traced in Stackdriver logs