Auto scaling for reliable delivery
This solution uses ALB and Auto Scaling to intelligently distribute network traffic and dynamically adjust server resources. This architecture improves application high availability and throughput, optimizes resource utilization, and reduces costs.
Intended customers
Users who handle high-concurrency traffic
Users who need to respond quickly to market changes
Users who seek to optimize resources and control costs
Solution background
High availability and Auto Scaling
In modern application architectures, high availability is a key factor for business continuity and the foundation for horizontal scaling. Horizontal scaling prevents resource waste by precisely matching resource supply with demand. It automatically scales in resources during low-traffic periods to reduce costs, and rapidly scales out server instances during traffic peaks to ensure service quality and user experience.
Solution advantages
Alibaba Cloud solution vs. a traditional approach
In a traditional approach, you build your own load balancing nodes and manually scale resources to meet business demand. This method leads to high hardware costs, complex management, and poor maintainability, while also requiring a significant investment in time and personnel. We recommend using the Alibaba Cloud ALB and Auto Scaling solution for horizontal application scaling.
Traditional approach
Alibaba Cloud ALB + Auto Scaling
Complex deployment
A traditional high-availability deployment is often complex. It requires manual configuration of multi-tiered redundant systems and a dedicated operations team for ongoing maintenance.
Deployment and maintenance complexity
Simplified configuration
Alibaba Cloud simplifies configuration by providing highly integrated tools and services that ensure application high availability.
High fixed costs
You must purchase and maintain equipment, and keep all hardware running even during low-load periods. This results in high fixed costs.
Cost efficiency
Reduces unnecessary costs
With a pay-as-you-go model, you can automatically scale resources based on actual demand to avoid unnecessary costs.
Slow fault recovery
Recovery is typically slow and requires manual intervention to inspect and replace faulty hardware or restart systems.
Fault recovery
Fast automatic failover
Alibaba Cloud offers fast, automatic failover capabilities that help you achieve near-zero downtime recovery.
Poor scalability
Scalability is limited. You must plan and purchase extra resources in advance to handle potential traffic growth, which often leads to idle resources.
Scalability
Excellent scalability
This solution offers excellent scalability to handle traffic fluctuations. The services automatically adjust resources in real time without manual intervention.
Improve availability with ALB and ESS
High concurrency
A single Application Load Balancer (ALB) instance supports up to 10 million concurrent connections and 1 million new connections per second (CPS). A single ALB instance provides a Layer 7 processing capacity of up to 1 million queries per second (QPS).
High availability
ALB offers an availability service level agreement (SLA) of up to 99.995%. Auto Scaling (ESS) provides health checks. If an Elastic Compute Service (ECS) instance becomes unhealthy, ESS automatically replaces it with a new instance of the same type.
Optimize costs with Auto Scaling
Auto Scaling eliminates the need for manual resource adjustments. It removes the requirement to provision resources in advance and ensures that idle resources are released promptly. ESS performs scaling tasks at the right time to reduce resource ownership costs.
Architecture and deployment
On-demand scaling: Automatic elasticity for stable delivery
2766666
This solution uses Application Load Balancer (ALB) and Auto Scaling (ESS) to improve application availability. When ALB receives a user request, it routes traffic to a target group based on the request content. The ALB server group is associated with an ESS group. This allows ESS to automatically adjust the number of specified ECS instances to meet fluctuating business demands.
30 minutes
CNY 10(This estimate assumes you select the recommended resource specifications and run the solution for no more than 60 minutes. Actual costs may vary based on traffic usage. The final cost is subject to the quote displayed in the console and your final bill.)
ECSSLBAuto Scaling
Recommended solutions