A container app serving a public API should add replicas as soon as any single replica is handling about 100 requests at the same moment. Which scale rule configuration matches this requirement?
Show answer and explanation
Answer: A
An HTTP scale rule in Container Apps measures concurrent requests per replica, and setting concurrentRequests to 100 tells the platform to add replicas once each one reaches that load. The mechanism is powered by KEDA behind the scenes, but the scale rule is what you configure. A cron rule reacts to the clock, a queue rule reacts to message backlog, and revision label weights only divide traffic between revisions without changing replica count.
Why the other options are wrong
- B. A cron rule scales on a fixed schedule rather than on live request load, so it would not respond when concurrency rises at unpredictable times.
- C. A queue rule scales on the number of pending messages, which does not represent concurrent HTTP requests handled by each replica.
- D. Revision label weights control traffic split between revisions and do not change how many replicas run, so they cannot add capacity.