We can do more than you expect
HIGH LOAD BETTING PLATFORM
US BettingOnline betting platform are widely used, so that users can bet from any of their personal devices. The platform is built to handle large amounts of traffic and data while maintaining high performance and reliability.
Challenge
The betting sector is a dynamic and growing business, with new competitive entries each year. It has now become a fiercely competitive industry of fickle users, regulatory pressure, and an increasing user base.
ARCHITECTURE REVIEW
Integrations with bettable content provider
Integrations with stats/visuals provider
Integrations with casino providers
Supports multiple sites/brands
Is optimized for speed (web sockets)
Is built using horizontal scaling and supports member grows
I Is built using powerful message broker so supports providers growths
Has Big Data
Databases migrated to RDS. Also migrated GitLab to gitlab.com to use its SaaS
Is flexible enough to support b2b, b2c and embedded environments
Is integrated with payment gateway hub that offers different convenient deposit / withdrawal method types
Is a subject of seamless integration to external wallets
Is hosted on cloud
Is monitored
Key concepts
Multiple location regions
User facing region with its own dynamic caches is represented on the diagram and is working to ensure fastsest dynamic content apearance for the end users and close to the customer CDN and security logic
Shards for users
- users are split to groups and all the transactions and computations are done on the shard, with no dependency on other shards
- the cumulative parameters and configs are synchronized via the message broker to the admin/backoffice platform
- configurations and global events are done through the same queing tier
User local caches and websockets
Duplicate blocks and columns to add more features.
Microservices that are independent and working through a message broker
Every integration is an isolated microservice that is using
- distributed cache for state synchronization
- message broker for event synchronization
Big Data / Analytics
standard big data patterns are applied onto gathering and distributing data from multiple shards and providing common interface to the cumulative and analytical data
Event / Content manager
A combo tool that allows to configure events based on the offerings on the provider
Global Backoffice
Global admin tool is used for managing users, content and event provider
Casino lobby is managed there as well as company risks and p&l.
Provider manager
A hub for all the providers, their configs the content they offer and their manager tool that allows to configure events based on the offerings on the provider
Casino lobby / casino adapters
Casino providers are treated / configured separately and grouped into configurable casino lobby
Payment gateway
Needed for b2c mostly, but could be extended for uncommon b2b payment methods
Provider adapters
standard big data patterns are applied onto gathering and distributing data from multiple shards and providing common interface to the cumulative and analytical data
- Stats
- Video
- Differnent sports
- E-sports
- Virtual sports
- Exchange
- Sports boo
Seamless
Flexible member account structure allows using an external wallet as a transaction approval source, allowing the system to be integrated into bigger platforms through seamless iframe
Autoscaling stateless cluster to support peak loads
Sports betting is hugely dependent on some timed events, so the load is significantly higher onto some dedicated time slots throughout the year
Autoscaling stateless servers group allows adding more compute power only when it’s needed, thus making overall platform very cost-effective
Result
55%
IT infrastructure cost reductions
65%
IT staff productivity benefits
75%
User productivity benefits
95%
Business productivity benefits