# Default service response time alerting

**URL:** <https://community.dynatrace.com/t/default-service-response-time-alerting/47967>\
**Category:** Alerting\
**Tags:** services-classic, anomaly-detection, problems-classic\
**Created:** [September 25, 2024, 8:32am UTC](https://community.dynatrace.com/t/default-service-response-time-alerting/47967 "2024-09-25T08:32:53Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![g\_kat](https://dub1.discourse-cdn.com/dynatrace/user_avatar/community.dynatrace.com/g_kat/32/34573_2.png) [@g\_kat](https://community.dynatrace.com/u/g_kat)\
**Post date:** [September 25, 2024, 8:32am UTC](https://community.dynatrace.com/t/default-service-response-time-alerting/47967/1 "2024-09-25T08:32:53Z")

</div>

Hi,

This maybe a rudimentary question but I wanted to be completely sure:

Let’s say we have a Web Request Service that has X service methods/endpoints. I leave the anomaly detection for response time on default values. Will Dynatrace raise an alert whenever 1 of those service methods deviates from the default baseline, or it will wait until all service methods deviate from the baseline?

I think is the former, but I think I read somewhere about a corner case(if an endpoint contributes less than 1% to the load is ignored, or something along those lines).

Thanks,

George

---

<div class="post-metadata">

**Author:** ![AntonPineiro](https://dub1.discourse-cdn.com/dynatrace/user_avatar/community.dynatrace.com/antonpineiro/32/49_2.png) [@AntonPineiro](https://community.dynatrace.com/u/AntonPineiro)\
**Post date:** [September 25, 2024, 9:24am UTC](https://community.dynatrace.com/t/default-service-response-time-alerting/47967/2 "2024-09-25T09:24:38Z")

</div>

Hi,

You can check/change anomaly detection for services. It will affect all request in relation to that service.

If you want different threshold for a single request, you need to mark it as [key request](https://docs.dynatrace.com/docs/shortlink/key-requests#create-ui), in advance. Then, you can [Request-specific alerting thresholds](https://docs.dynatrace.com/docs/shortlink/key-requests#alerting).

Best regards

---

<div class="post-metadata">

**Author:** ![g\_kat](https://dub1.discourse-cdn.com/dynatrace/user_avatar/community.dynatrace.com/g_kat/32/34573_2.png) [@g\_kat](https://community.dynatrace.com/u/g_kat)\
**Post date:** [September 25, 2024, 9:29am UTC](https://community.dynatrace.com/t/default-service-response-time-alerting/47967/3 "2024-09-25T09:29:29Z")

</div>

Hi Anton,

I don’t want a different threshold, quite the opposite I want all requests to obey the overall threshold. But, I want to be sure that it will raise a problem for the service if even 1 endpoint is problematic.

Key requests are definitely a solution, but I wanted to double check whether I need them, since if by default Dynatrace alerts on endpoint deviation I don’t have a need for them.

Thanks,

George

---

<div class="post-metadata">

**Author:** ![AntonPineiro](https://dub1.discourse-cdn.com/dynatrace/user_avatar/community.dynatrace.com/antonpineiro/32/49_2.png) [@AntonPineiro](https://community.dynatrace.com/u/AntonPineiro)\
**Post date:** [September 26, 2024, 9:21am UTC](https://community.dynatrace.com/t/default-service-response-time-alerting/47967/4 "2024-09-26T09:21:35Z")

</div>

Hi,

One of benefits of key request is “Alerting is always enabled for key requests, even when they contribute less than 1% of throughput”.

If response time is higher than normal, it would raise a problem, no matter&nbsp;throughput.

Best regards