Response time is represented by a line in some New Relic APM charts. It is the duration of a transaction from the perspective of the requester, and not necessarily the total sum of all time spent within a transaction.
Response time and total time
The relationship between response time and total amount of time is a reflection of how much concurrency is taking place among the instrumented code in your application:
- If response time is equal to the total time spent, you may be using a blocking or synchronous computing model.
- If response time is less than the total time spent, you are taking advantage of concurrency with a non-blocking or asynchronous computing model.
- If response time is greater than the total time spent, this may be due to resource contention. For example, if the response work runs in a highly contended thread pool, other jobs must finish prior to completing the response. The time spent waiting for resources is not currently tracked by the New Relic agent.
In order to view decoupled response time metrics, upgrade your agent to the supported version as applicable:
- Java 3.22.0 or higher
- .NET: Version requirements depend on async mode
- Node.js 1.25.0 or higher
- PHP 5.4.0 or higher
- Python 188.8.131.52 or higher
- Ruby 184.108.40.2069 or higher