Instana Application Metrics API

The endpoints of this group retrieve the metrics for defined applications, discovered services and endpoints. ### Mandatory Parameters **metrics** A list of metric objects that define which metric should be returned, with the defined aggregation. Each metrics objects consists of minimum two items: 1. *metric* select a particular metric to get a list of available metrics query the [catalog endpoint](#operation/getApplicationCatalogMetrics). 2. *aggregation* depending on the selected metric different aggregations are available e.g. SUM, MEAN, P95. The aforementioned [catalog endpoint](#operation/getApplicationCatalogMetrics) gives you the metrics with the available aggregations. **Note**: The above mentioned list of available metrics with its supported metrics can also be found in the section **Supported Aggregation on Application metrics** below. ### Optional Parameters **metrics** Default you will get an aggregated metric with for the selected timeframe * *granularity* * If it is not set you will get a an aggregated value for the selected timeframe * If the granularity is set you will get data points with the specified granularity **in seconds** * The granularity should not be greater than the `windowSize` (important: `windowSize` is expressed in **milliseconds**) * The granularity should not be set too small relative to the `windowSize` to avoid creating an excessively large number of data points (max 600) **pagination** if you use pagination you most probably want to fix the timeFrame for the retrieved metrics 1. *page* select the page number you want to retrieve 2. *pageSize* set the number of applications you want to return with one query **order** You can order the returned items alphanumerical by label, either ascending or descending 1. *by* if the granularity is set to 1 you can use the metric name eg. "latency.p95" to order by that value 1. *direction* either ascending or descending **timeFrame** As in our UI you can specify the timeframe for metrics retrieval. ``` windowSize to (ms) (unix-timestamp) <----------------------| ``` The timeFrame might be adjusted to fit the metric granularity so that there is no partial bucket. For example, if the query timeFrame is 08:02 - 09:02 and the metric granularity is 5 minutes, the timeFrame will be adjusted to 08:05 - 09:00. The adjusted timeFrame will be returned in the response payload. If the query does not have any metric with granularity, a default granularity will be used for adjustment. To narrow down the result set you have four options to search for an application. **nameFilter | applicationId | serviceId | endpointId** * *nameFilter:* filter by name with "contains" semantic. * *applicationId:* search directly for an application by applicationId * *serviceId:* search for applications that include a particular service by serviceId * *endpointId:* search for applications that include a particular endpoint by endpointId ### Defaults **metrics** * *granularity:* 1 **order** * by application label ascending. **timeFrame** ``` "timeFrame": { "windowSize": 60000, "to": {current timestamp} } ``` **nameFilter | applicationId | serviceId | endpointId** * no filters are applied in the default call ## Supported Aggregation on Application Metrics | Metric | Description | Allowed Aggregations | |------------------|--------------------------------------------------------------------------------------------|----------------------| | `calls` | Number of received calls | `PER_SECOND`, `SUM` | | `erroneousCalls` | The number of erroneous calls |`PER_SECOND`, `SUM` | | `latency` | Latency of received calls in milliseconds | `P25`, `P50`, `P75`, `P90`, `P95`, `P98`, `P99`, `SUM`, `MEAN`, `MAX`, `MIN` | | `errors` | Error rate of received calls. A value between 0 and 1 | `MEAN` | | `applications` | The number of Application Perspectives |`DISTINCT_COUNT` | | `services` | The number of Services |`DISTINCT_COUNT` | | `endpoints` | The number of Endpoints |`DISTINCT_COUNT` | | `http.1xx` | Counts the number of occurrences of HTTP status codes where 100 <= status code <= 199 |`PER_SECOND`, `SUM` | | `http.2xx` | Counts the number of occurrences of HTTP status codes where 200 <= status code <= 299 |`PER_SECOND`, `SUM` | | `http.3xx` | Counts the number of occurrences of HTTP status codes where 300 <= status code <= 399 |`PER_SECOND`, `SUM` | | `http.4xx` | Counts the number of occurrences of HTTP status codes where 400 <= status code <= 499 |`PER_SECOND`, `SUM` | | `http.5xx` | Counts the number of occurrences of HTTP status codes where 500 <= status code <= 599 |`PER_SECOND`, `SUM` |