問題文
A retailer's order intake API runs on CloudHub on a single 1 vCore worker in one region. A recent incident took the whole application offline for eleven minutes while the platform moved it after a hardware failure. The retailer wants that class of outage removed. What should the architect change first?
選択肢
- Deploy the application to more than one worker, because CloudHub then spreads those workers across two or more data centers and the shared load balancer keeps sending requests to the workers that are still up
- Increase the worker to 4 vCores, because CloudHub then gives that worker more memory and disk and the shared load balancer keeps sending requests to it while the platform repairs the underlying hardware
- Enable persistent queues on the application, because CloudHub then keeps the messages while the worker is unavailable and the shared load balancer keeps sending requests to it once it returns
- Enable automatic restarts on the application, because CloudHub then brings the worker back without an operator and the shared load balancer keeps sending requests to it as soon as it is up