Monitor key requests
Key requests are requests that need special attention, either because they're a critical measure of the success of your business (for example, a login request or a shopping-cart checkout request) or because they provide vital technical functionality that your application relies on.
Key requests feature long-term metric history and dedicated dashboard tiles for charting and direct access from your dashboard. Alerting is always enabled for key requests, even when they contribute less than 1% of throughput. They also provide custom thresholds.
Key requests are highlighted in the Key requests section of each service overview page. This visibility is particularly valuable for low-volume, high-importance requests that would otherwise appear at the bottom of the Top requests section of a service overview page.
To tag a specific request as a key request
In the Dynatrace menu, select Services.
Select the relevant service from the list.
On the service overview page, select View (View requests, View dynamic web requests, or View resources).
Scroll down to the Top requests section and select a request you want to mark as a key request.
On the request overview page, open the browse menu (…) and select Mark as key request.
Show key requests on a dashboard
To create a dashboard tile for a specific request
- In the Dynatrace menu, select Services.
- Select the relevant service from the list.
- On the service overview page, select View (View requests, View dynamic web requests, or View resources).
- Scroll down to the Key requests section and select a request you want to show on a dashboard.
- On the request overview page, open the browse menu (…) and select Pin to dashboard. A new request-specific tile that shows the most important metrics for that particular request is then added to your dashboard.
Long-term data for key requests
Dynatrace enables you to chart any request that it detects during monitoring. By default, detailed history of all requests is retained for 10 days. Longer-term historical data is maintained for requests that you manually identify as key requests. Trend lines for key requests are retained perpetually, but the granularity of long-term history is gradually reduced over time. See the list of available granularities below.
|Retention period||Interval granularity|
|0–14 days||1 minute|
|14–28 days||5 minutes|
|28–400 days||1 hour|
|400 days–5 years||1 day|
Anomaly detection with key requests
Dynatrace assumes that low-volume requests are of less importance than high-volume and key requests. This means that requests that contribute less than 1% to the overall load of a service won't raise alerts unless their impact is significant enough that the service's overall response time or failure rate is affected. Because this default treatment is not appropriate for all low-volume requests, you should manually tag any important low-volume requests as key requests to ensure that they have standard alerting thresholds.
Request-specific alerting thresholds
Because certain requests may have specific response-time and failure-rate patterns, while others may have strict SLA thresholds, Dynatrace enables you to define custom alerting thresholds when anomalies are detected related to the performance of key requests. If set, key-request-level thresholds override service-level thresholds. To learn how to set request-level thresholds, see Thresholds for a specific web request.
Calculated service metric
As an alternative way to focus on particular requests, you can create a calculated service metric, based on the requests you need. This approach provides you more flexibility with alerting—you can use the calculated metric just like any built-in metric provided by Dynatrace.