Which metric best tells you a worker fleet should scale up?

Options

• ✗ CPU usage on the message broker
• ✓ The age of the oldest message, or the backlog per worker Correct
• ✗ The number of producers
• ✗ The average message size

Why

Message age measures what users feel: how long work waits. Worker CPU can be low while workers wait on slow downstream calls, so it can miss a growing backlog.

Review this in your daily deck → All cards in Queues