+Another useful metric to consider for scaling is the time between when the latest message was sent and when it was processed, also known as "critical time". This is helpful for scenarios where a queue may have thousands of messages in it, but the processing is fast enough to keep up, giving a "critical time" of only a couple of seconds, which may be more than enough for something like an email sending endpoint. Third-party libraries like [NServiceBus](https://docs.particular.net/nservicebus/operations/opentelemetry#meters-emitted-meters) emit this and other useful metrics via OpenTelemetry.
0 commit comments