Option C provides the lowest-change migration path for the application tier. Rehosting the existing Windows workload on Windows Server ECS instances preserves its operating-system environment while distributing instances across availability zones improves resilience against zone-level infrastructure failures.

For the database tier, Option A provides the stronger managed architecture. Alibaba Cloud explicitly supports using DTS to migrate self-managed Oracle databases to PolarDB for PostgreSQL (Compatible with Oracle), including schema, full-data, and incremental-data migration for minimal migration downtime.

PolarDB ' s Oracle-compatible edition is specifically engineered to run Oracle workloads with minimal application modification. It supports Oracle-compatible data types, SQL constructs, PL/SQL features, OCI connectivity, packages, sequences, synonyms, and related functionality.

It also provides built-in high availability: failures are detected automatically and workloads can fail over to healthy nodes. This provides a more complete managed availability architecture than simply deploying independent self-managed Oracle servers across zones, which would additionally require Oracle-level replication/failover configuration.

Options B and D require substantial application refactoring and therefore conflict with the explicit requirement to minimize development changes.

Study Guide reference: Building Highly Available, Performant Cloud Architecture — workload migration, ECS multi-zone design, DTS, and PolarDB Oracle compatibility.

==================================================

Option B reduces interference between transaction processing and read-heavy workloads inside PolarDB. PolarDB ' s read/write splitting automatically routes writes to the primary node while distributing eligible reads across read-only nodes. This allows reporting and ordinary SELECT traffic to consume separate compute capacity instead of competing with critical transactional writes on the primary node. Alibaba Cloud specifically identifies reporting, analytics, and read-heavy applications as appropriate read/write-splitting scenarios.

Option D provides stronger isolation for computationally expensive analytics. DTS can continuously synchronize PolarDB for MySQL data into MaxCompute using initial full synchronization followed by ongoing incremental change capture. Analysts can then execute large-scale analytical workloads against MaxCompute without placing those queries on the production transactional database.

Running queries on ECS does not inherently remove load from PolarDB because the SQL operations would still execute against the database. Introducing MongoDB merely to execute analytical queries creates a second database technology without an appropriate replication or analytical architecture.

Together, PolarDB read/write splitting optimizes operational read traffic while DTS-to-MaxCompute isolates heavyweight analytical processing, producing an effective HTAP-style separation of concerns.

Study Guide reference: Core Infrastructure Deep Dive — PolarDB, read/write splitting, DTS, MaxCompute, and analytical workload offloading.

==================================================

The strongest architecture for an existing e-commerce platform experiencing predictable but substantial traffic spikes is OSS/CDN + ALB + Auto Scaling. Auto Scaling is specifically designed to increase ECS capacity when traffic or resource utilization rises and remove unnecessary instances after demand subsides. Alibaba Cloud documents this as a core mechanism for handling traffic spikes while avoiding permanent over-provisioning.

ALB complements Auto Scaling by distributing requests across healthy ECS backend instances. As scaling creates additional instances, they can be automatically added to the ALB server group, allowing traffic to use the enlarged compute pool immediately.

Static images and videos should be moved to OSS and accelerated through CDN. CDN caches static resources on geographically distributed edge nodes, substantially reducing requests reaching the ECS application tier and lowering origin bandwidth and processing load during sales events.

Function Compute is highly elastic, but introducing it merely to supplement an existing server-based application normally requires application restructuring. It is therefore less directly applicable than the three mechanisms specifically designed to scale and offload the existing architecture.

Study Guide reference: Building Highly Available, Performant Cloud Architecture — Auto Scaling, ALB, CDN offloading, and elastic ECS architecture.

==================================================

The workload has an extreme daily usage differential, making serverless PolarDB more appropriate than a fixed database configuration. PolarDB Serverless automatically scales compute resources up or out when CPU, memory, or database load increases and scales them down when demand subsides. This allows the database tier to accommodate millions of morning users without permanently paying for peak capacity throughout the remainder of the day.

The application front end is a single static page, so ECS instances are unnecessary. OSS can host HTML, CSS, JavaScript, images, and other static files directly. Placing Alibaba Cloud CDN in front of the OSS bucket caches those files at globally distributed edge nodes and serves requests from locations near users. Alibaba Cloud recommends OSS + CDN for high-concurrency static websites because it reduces access latency and removes most origin load.

Option A can scale but retains fixed provisioned database resources and therefore provides less elasticity than the serverless model. Option D creates unnecessary compute infrastructure for content that requires no server-side execution.

The architecture therefore separates an infinitely scalable static delivery tier from an elastic serverless persistent database tier.

Study Guide reference: Building Highly Available, Performant Cloud Architecture — PolarDB Serverless, OSS static hosting, CDN, and automatic resource scaling.

==================================================

The threats described exist at different layers, so the architecture should apply layered protection. WAF is the correct control for public HTTP/API traffic because it analyzes application-layer requests and blocks threats such as SQL injection, cross-site scripting, command injection, brute-force attacks, and other OWASP-style attacks. Alibaba Cloud explicitly supports integrating WAF with API Gateway and recommends disabling direct access paths afterward so clients cannot bypass WAF.

Large DDoS attacks require a dedicated volumetric mitigation service. Anti-DDoS Proxy scrubs malicious network and transport-layer traffic before clean traffic is forwarded toward the application infrastructure. Alibaba Cloud ' s Anti-DDoS architecture supports protected services behind Server Load Balancer resources and provides port-forwarding mechanisms for non-HTTP workloads.

Alibaba Cloud specifically recommends combining Anti-DDoS and WAF when an application needs protection against both large volumetric attacks and sophisticated application-layer exploits.

WAF should protect the HTTP/API application boundary rather than being treated as a general Layer-4 NLB firewall. Security Center provides workload security and posture management, while basic DDoS protection alone is less suitable for the question ' s explicitly stated large and sophisticated attacks.

Study Guide reference: Securing Workloads on Alibaba Cloud — WAF, Anti-DDoS, API Gateway security, and defense-in-depth.

==================================================

CDN should form the content-delivery layer for cacheable video assets. Alibaba Cloud CDN distributes content to geographically dispersed edge nodes so viewers retrieve cached objects from nearby POPs rather than repeatedly connecting to a distant origin. Alibaba Cloud explicitly supports video-oriented CDN acceleration and edge delivery, reducing both end-user latency and origin bandwidth demand.

Auto Scaling addresses peak-time compute demand. It monitors workload conditions and automatically increases ECS capacity during traffic spikes, then scales resources back when traffic decreases. Alibaba Cloud specifically identifies streaming platforms with unpredictable demand as a typical Auto Scaling scenario.

Global Accelerator complements CDN and Auto Scaling where viewers or application endpoints are separated by long-distance or cross-border networks. GA routes traffic through Alibaba Cloud ' s optimized global BGP network, reducing latency, jitter, and packet loss. Alibaba documentation explicitly identifies video streaming as a latency-sensitive scenario that benefits from this architecture.

Tair can improve authentication or session performance, but it does not directly address video delivery quality or regional network performance. Therefore, CDN + Auto Scaling + Global Accelerator most directly satisfies the streaming requirement.

Study Guide reference: Delivering Services and Content on Alibaba Cloud — CDN, Global Accelerator, Auto Scaling, video delivery, and global traffic optimization.

==================================================

The fundamental decision is the trade-off between infrastructure control and infrastructure abstraction.

A conventional managed/dedicated Kubernetes environment gives architects greater control over node pools, instance types, resource placement, operating-system configuration, network topology, specialized hardware, and related infrastructure choices. This makes Option B appropriate when a regulated fintech application requires exact control over the compute and networking environment.

ACK Serverless takes the opposite approach. Alibaba Cloud manages the underlying node infrastructure, and workloads run without customers purchasing or maintaining worker nodes. Alibaba ' s documentation states that ACK Serverless uses pay-as-you-go resource consumption and can scale pods rapidly, making it particularly appropriate for variable or unpredictable workloads.

Alibaba Cloud also explicitly describes ACK Serverless as Kubernetes-compatible infrastructure where customers focus on application development while paying according to CPU and memory resources requested by workloads.

Option A is too absolute: dedicated Kubernetes infrastructure does not inherently guarantee lower application latency; latency depends on workload, topology, instance configuration, networking, and scaling behavior. Option C is also incorrect because both managed and serverless ACK environments integrate deeply with Alibaba Cloud networking services.

Study Guide reference: Core Infrastructure Deep Dive — ACK deployment models, serverless containers, cluster control, elasticity, and cost optimization.

==================================================