Apdex is an industry standard to measure users' satisfaction with the response time of web applications and services. It's a simplified Service Level Agreement (SLA) solution that helps you see how satisfied users are with your app. In contrast, average response time and other traditional metrics can be skewed by a few very long responses.
To see how all your applications, services, containers, cloud services, hosts, and other entities work together, use New Relic One.
Apdex is a measure of response time based against a set threshold. It measures the ratio of satisfactory response times to unsatisfactory response times. The response time is measured from an asset request to completed delivery back to the requestor.
The application owner defines a response time threshold
T. All responses handled in
T or less time satisfy the user.
For example, if
T is 1.2 seconds and a response completes in 0.5 seconds, then the user is satisfied. All responses greater than 1.2 seconds dissatisfy the user. Responses greater than 4.8 seconds frustrate the user.
You can define multiple Apdex T values for each of these:
New Relic Browser monitors the end-user experience for these apps. To define end-user Apdex settings for these apps, use the Browser UI. New Relic labels them as End user on the APM Overview page and the Browser Overview page.
These are transactions important to your business that you choose to monitor. To define Apdex settings for key transactions, use the Key transactions page in APM, then view on the Key transactions Overview page.
Apdex tracks three response counts:
- Satisfied: The response time is less than or equal to T.
- Tolerating: The response time is greater than T and less than or equal to 4T. In this example, 4 x 1.2 = 4.8 seconds as the maximum tolerable response time.
- Frustrated: The response time is greater than 4T.
Your configuration file's
apdex_f value is four times your app server's Apdex T value. This threshold is useful, for example, with transaction traces. For more information, see the configuration file documentation for your New Relic agent.
The time calculation will change based on your own app's T setting. In the following example, T = 1.2 seconds.
|Level||Multiplier||Time (T Example = 1.2)|
|Satisfied||T or less||<= 1.2 seconds|
|Tolerated||>T, <= 4T||Between 1.2 and 4.8 seconds|
|Frustrated||> 4T||Greater than 4.8 seconds|
After you define your Apdex levels, use any of New Relic's resources to help identify and troubleshoot changes that indicate poor customer experiences with your app.
The Apdex score is a ratio value of the number of satisfied and tolerating requests to the total requests made. Each satisfied request counts as one request, while each tolerating request counts as half a satisfied request. An Apdex score varies from 0 to 1, with 0 as the worst possible score (100% of response times were Frustrated), and 1 as the best possible score (100% of response times were Satisfied).
Example Apdex score:
During a 2 minute period a host handles 200 requests. The Apdex threshold T = 0.5 seconds (500ms). This value is arbitrary and is selected by the user.
- 170 of the requests were handled within 500ms, so they are classified as Satisfied.
- 20 of the requests were handled between 500ms and 2 seconds (2000 ms), so they are classified as Tolerating.
- The remaining 10 were not handled properly or took longer than 2 seconds, so they are classified as Frustrated.
The resulting Apdex score is 0.9:
(170 + (20/2))/200 = 0.9.
Errors frustrate your users. Any request that raises a server-side error is considered a frustrating response, no matter how fast it returns to the user. Nobody likes to see a "500: Application Error" page. This makes it possible to have an average response time that is well below the Apdex T but still have a poor Apdex score.
The dissatisfaction percentage is the percentage of the total dissatisfaction experienced by the app's users that is contributed by this transaction.
Example Apdex dissatisfaction calculation:
Frustrations(App) + Tolerations(App)/2
If a transaction is always frustratingly slow but rarely visited, it will not contribute much to the app's total dissatisfaction. Conversely, if a transaction normally is fast (for example, Apdex 0.95) but has high throughput, this may contribute a large proportion of the app's dissatisfaction simply because it contributes a large proportion of your app's traffic.