How to Ensure Availability and Elasticity in High-Concurrency Scenarios on Huawei Cloud Singapore CN2

Introduction: In the context of cross-border operations and low-latency requirements, how to ensure availability and scalability in high-concurrency scenarios within Huawei Cloud’s Singapore CN2 network environment is a core challenge that architects and operations teams must address. This article provides actionable recommendations on architecture design, network optimization, auto-scaling, data consistency, and operational automation, balancing stability with cost efficiency. It is suitable as a reference for deployment in the Asia-Pacific region with connectivity to the Chinese mainland.

Architectural Design Principles: Prioritize both availability and resilience

When designing high-concurrency systems on Huawei Cloud Singapore cn2, the principles of layering, decoupling, and reusability should be followed. Separate the frontend, applications, cache, and persistence layer. Use stateless application instances combined with shared storage or external session management to facilitate horizontal scaling and rapid failover, thereby improving overall availability and resilience.

Multi-AZ and Cross-Region Disaster Recovery Strategies

Prioritize multi-availability zone deployment to isolate hardware and network failures between availability zones. For critical business operations, cross-regional disaster recovery (DR) plans are designed, with regular off-site drills and recovery tests conducted to ensure rapid switching in the event of single-point failures or regional disruptions, thereby reducing the risk of business interruptions.

Auto-scaling and Load Balancing Practices

Achieve on-demand scaling by combining Auto Scaling with ELB. Set scaling policies based on business metrics such as request count, CPU usage, and queue length, and combine them with warm-up and cool-down strategies to reduce fluctuations ； Load balancing configuration health checks ensure that traffic is directed only to available instances.

Network and Interconnection Optimization: Utilizing CN2 properties

In scenarios with extensive interactions with the Chinese mainland, CN2 exports usually provide a more stable and low-latency transmission path. Making sensible choices regarding exits, configuring bandwidth pools, optimizing routing and MTU, and tuning parameters for long-lived connections and concurrent connections can significantly improve the experience of cross-border high-concurrency access.

Caching and Data Layering: Reduce backend pressure

Reduce the load on the origin server through multi-level caching (CDN + application-side caching + distributed caching). Store hot data in local or distributed caches (such as Redis), and use protection strategies against cache penetration, avalanche, and breakdown to avoid sudden surges in database traffic under high-concurrency scenarios.

High availability of databases and read-write separation

The database layer utilizes master-slave replication, semi-synchronous or synchronous replication mechanisms to achieve high availability, and combines read-write separation and sharding strategies to alleviate the load on a single database. Regular backups and offsite backups (backups to different locations) are essential means to ensure data security and recovery capabilities.

Asynchronous processing and peak shaving

Asynchronously process non-critical requests, use message queues to decouple from peak traffic, and smooth out load through rate limiting, delayed queues, degradation, and batch processing. Combine traffic throttling strategies (such as queuing, draining, and circuit breaking) to protect the stability of core services during sudden high concurrency.

Monitoring, Early Warning, and Automated Operations Management

Establish a unified monitoring system that covers networks, hosts, applications, middleware, and business metrics, with multi-level alerts and self-healing scripts. Combine with automated operations and maintenance (IaC, CI/CD, rolling releases) to reduce manual intervention and improve fault response speed and recovery efficiency.

Security and DDoS Protection

In high-concurrency scenarios, security challenges arise simultaneously; therefore, DDoS protection, WAF, network ACLs, and strict access control should be deployed. Rate-limit traffic spikes and manage allowlists/blocklists, while implementing identity authentication and traffic auditing for critical interfaces to prevent malicious requests from affecting availability.

Testing, Drilling, and Deployment Strategies

Regularly conduct stress testing, chaos engineering exercises, and failover drills to verify the effectiveness of scaling strategies and disaster recovery procedures. Blue-green or gray-release methods are used to reduce the risks associated with new versions, along with rollback mechanisms to ensure that stable services can be quickly restored in case of regressions or failures.

Summary and Recommendations

Summary: To ensure high concurrency availability and elasticity in Huawei Cloud’s Singapore CN2 environment, it is necessary to make coordinated efforts across five dimensions: architecture, network, data, operations and maintenance, and security. It is recommended to prioritize multi-availability zone deployment, auto-scaling, and multi-level caching. Combined with a robust monitoring and testing framework, continuous optimization of the pipeline and capacity planning should be carried out to ensure stable and reliable service performance under high-concurrency scenarios.