在使用多区域部署原型的云架构中，应用可在两个或更多 Google Cloud 区域中运行。应用数据会复制到架构的所有区域。为了确保快速、同步地复制数据，区域通常位于一个大洲内。
下图显示了在两个 Google Cloud 区域中运行的应用的云拓扑：
上图显示了在两个 Google Cloud 区域中独立运行的两个独立多层应用堆栈。在每个区域中，该应用都在 3 个可用区运行。系统会复制两个区域中的数据库。如果工作负载具有较低的恢复点目标 (RPO)，或者需要高跨区域数据一致性，则数据库复制需要同步。否则，数据库可以异步复制。用户请求使用 DNS 路由政策路由到区域负载平衡器。
如果两个区域中的任何一个发生中断，则 DNS 会将用户请求路由到另一个区域的负载均衡器。
[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC)：2024-11-20。"],[[["The multi-regional deployment archetype involves running an application in two or more Google Cloud regions, with data replicated across all regions, typically within a continent for fast, synchronous replication."],["This archetype is recommended for business-critical applications requiring high availability and robustness against region outages, ensuring minimal downtime and near-zero recovery time objective (RTO) when data is replicated synchronously."],["Multi-regional deployments offer low latency for users within a specific geographical area by allowing a global load balancer to route requests to other regions during an outage, maintaining performance."],["It can help meet data residency and operational sovereignty requirements by deploying applications in specific Google Cloud regions and using geofenced DNS routing policies to ensure data stays within a defined geographic area."],["While this approach provides high availability, considerations include the potentially higher cost of cloud resources and network traffic, as well as the increased operational complexity due to managing redundant resources across locations."]]],[]]
