Alerting on Zero Incoming Requests in Dynatrace

[Sohel_Rashid]

Is there any capability in Dynatrace to track incoming request volume per minute and generate an alert when no requests are detected for a defined duration?

[MaximilianoML · Accepted solution]

Hello, @Sohel_Rashid 

Yes! You can create a metric event on a built-in request-count metric for the target service or endpoint and trigger an alert when the value stays below 1 for a defined window. If needed, you can also enable missing data alerting to detect a complete stop in incoming traffic.

[Sohel_Rashid]

Hi @MaximilianoML 

Thanks for the confirmation! That helps a lot.
I had one follow-up question—regarding the Query definition type, should I use a Metric selector or a Metric key for this use case?

Also, could you please guide me with a sample Metric selector query to monitor request count and trigger an alert when it drops below 1 (or no traffic is received) over a defined time window? That would be really helpful.

[MaximilianoML · Accepted solution]

Hello, @Sohel_Rashid

I’d usually start with Metric key if you only want a simple statc alert on one built-in request-count metric. Metric key events monitor a single raw metric and support only static thresholds, while Metric selector events are better when you want extra filtering, aggregation, or more flexible query logic. Also, Dynatrace service request metrics are collected with 1-minute granularity, which fits your “requests per minute / no traffic” scenario well.

If you want to scope it just to one specific service, then using Metric selector is often the most practical option. A selector would be something like this below:

builtin:service.requestCount.total
:filter(in("dt.entity.service",entitySelector("type(~"SERVICE~"),entityName.equals(~"YOUR_SERVICE_NAME~")")))
:splitBy()
:sum

[Sohel_Rashid]

Hi @MaximilianoML 

Thank you for the detailed explanation and guidance.

I’ve created the metric event for the “Registration Controller” service using the appropriate configuration. Could you please review it once and confirm if everything looks correct?

Thanks again for your support.

[MaximilianoML]

Hello, @Sohel_Rashid!

I reviewed your docx and looks good, you can check that is there a part saying "Alert Preview" and it's really useful in this case. Now the next step is to observe the behavior of the metric and analyze the alarms (if creates or not, and why).

[Sohel_Rashid]

Hi @MaximilianoML 

Thanks for the guidance and for reviewing the document.

I checked the configuration, but I’m not seeing the “Alert Preview” option on my end.

I’ll proceed with monitoring the metric behavior and analyze the alerts as suggested.

Thanks again for your support.