A global manufacturing company is developing an Azure AI solution to provide real-time translation for internal communications across different language-speaking teams. The solution will utilize Azure Cognitive Services Translator. The company needs to ensure that the translation service is highly available and responsive for users worldwide, minimizing latency regardless of their geographic location. Which deployment strategy should be employed for the Azure Cognitive Services Translator resource?
- AUse a single Translator resource in the closest Azure region to the company headquarters.
- BDeploy a single Translator resource in a central Azure region and use Azure Front Door.
- CDeploy the Translator service as a container on-premises in each major office location.
- DDeploy multiple Translator resources in different Azure regions and use Azure Traffic Manager.
Show answer & explanationAnswer & explanation
Correct answer: D. Deploy multiple Translator resources in different Azure regions and use Azure Traffic Manager.
Deploying multiple Translator resources in different Azure regions and using Azure Traffic Manager is the best strategy. Traffic Manager is a DNS-based load balancer that directs user requests to the closest healthy endpoint (Translator resource in this case) based on various routing methods, such as performance or geographic, ensuring high availability and minimizing latency for a global user base.
Why the other options are wrong
- A. A single resource, even if closest to headquarters, does not ensure low latency or high availability for a global user base outside that region.
- B. A single resource, even with Azure Front Door, introduces a single point of failure for the backend service and may still result in high latency for users far from that central region.
- C. Deploying containers on-premises provides low latency for local users but creates a distributed management burden and doesn't leverage Azure's global infrastructure for broader availability and scalability efficiently.
Global Distribution with Azure Traffic Manager
Using Azure Traffic Manager to distribute user traffic across multiple globally deployed Azure resources to ensure high availability, responsiveness, and low latency for a worldwide user base.
- DNS-based traffic load balancer.
- Routes users to the closest healthy endpoint.
- Supports various routing methods (e.g., performance, geographic, priority).
Memory trick: Traffic Manager 'Spreads' AI services for global speed.