Watch a video: Get an overview of how Apigee's Analytics Services lets you view API trends, spot anomalies, see relationships between different metrics, and view top performers, which helps you be proactive and responsive in managing your API ecosystem.
Data and metrics are key to managing the performance of your API program, because you can't manage what you can't measure. For information about the different data that is important to the API team, developers, the API operations team, and others who are involved in making your API program successful, see Analytics Services overview.
Apigee collects default information as data passes through it. The data includes API call information (like URL, IP, and User ID), latency, errors, and so on. This information is displayed in the out-of-box-reports. You can include additional information using policies. For example, Apigee can collect data from message headers, query parameters, and portions of the request and response extracted from XML or JSON. This data can be displayed in custom reports. See Build custom reports.
See the troubleshooting topics for ideas of how to use metrics to manage your API business.
The Apigee Enterprise dashboard gives you a comprehensive, high-level view of the traffic, apps, products, and developers in your API ecosystem.
The main Dashboard page displays overall API traffic for the time period and environment you select, and shows the number of successes and the number of errors as part of the total traffic. It also displays information about developer signups and traffic, as well as traffic numbers for the top developer apps. You can choose to display data for the last hour, today, the last week or month, or select a custom date range from a calendar. You can also choose to aggregate the displayed data by minute, hour, day, week, or month.
Click the Top Performance View icon to show charts for the top APIs, top developers and developer apps, and top products.
Overall, the charts display metrics that help you manage both the business and operational aspects of your APIs.
- API traffic monitoring metrics are useful for managing the business side of your APIs. These metrics are available from the Dashboard page and provide a high-level view of all the APIs in your organization. They give you insight into which APIs are most popular with customers and how much traffic is flowing through all your APIs.
- API performance metrics are useful for managing the operations side of an organization. These metrics are available on the Dashboard tab and provide insight into traffic, response time, and errors per API.
|All API traffic||
All the API traffic (throughput) for an enterprise over the selected period of time.
Throughput = traffic = the number of API requests and responses seen by the enterprise.
|Developer Engagement||This section visually shows the ratio among traffic levels for high, medium, low, and inactive developers using your APIs for the selected time period. Also shown are the total number of developers who have signed up to use your APIs, and the percentage change of developer signups between this week and last week.|
|Developer Apps||Traffic for your top-performing developer apps for the selected time period is displayed in this chart.|
The APIs with the highest traffic over the selected time period.
|Top Developer Apps||
The developer apps with the highest traffic over the selected time period.
The API products with the highest traffic over the selected time period.
(API products are groups or collections of APIs that are logically grouped into sets for the purposes of provisioning access, logging or metrics generation with the ability to view and maintain control at the group level.)
Developers with the highest traffic over the selected time period.
Viewing chart details
The charts provide a number of ways to see specific data numbers:
- Hover your mouse pointer over data points for a popup window that displays details.
- Click the Detail button in the API Traffic chart to see data details for specific APIs.
- Zoom in on charts by clicking and dragging across chart areas.
Notice that the details chart also identifies and lists statistics for the APIs that have the most and least traffic and most and least errors over the selected period.
The following metrics are available to help monitor API performance. You'll find these on the API Proxies page. You can drill down on specific APIs to see more granular metrics.
|Traffic||Also known as throughput. The number of API requests and resulting responses seen by the enterprise.|
|Average Response Time||The time an API takes to respond to an incoming request.|
|Average Target Response Time||
The time that elapses between the request entering the Apigee system from the Apigee gateway to the response arriving at the Apigee gateway from the Apigee system.
The workflow for the servicing of an API request is as follows:
|Average Endpoint Response Time||The average time it takes the target endpoint to respond to an incoming request for the selected period.|
|Maximum Response Time||The slowest response time for the selected period.|
|Error Rate||The fraction of all API requests that are unsuccessful, that is, the request does not deliver a response as desired by the end user.|
|Average Data Exchange||The size of request and response. That is, the amount of data that is transferred in both directions as a request for an API is services and a response is generated and delivered to the calling entity.|
If you have multiple resources in an API proxy, you can track metrics on any and all resources for comparison on the API proxy page. Metrics include traffic (and a bar chart that visually illustrates traffic trends), average response time, error rate, and the average size of the request and response in KB (average data exchange), which can also be displayed in a chart on the page.
Adding metrics for URIs
To collect metrics on individual URIs in an API proxy:
- In the API Platform menu, select APIs, then click the API proxy you want.
- In the Performance section of the API proxy page, click + URI Pattern.
- Select the method/verb for which you want metrics.
- Enter a Pattern for a URI that is beyond the Default Proxy Endpoint Base Path listed at the top of the page. (The URI pattern must be part of the proxy.pathSuffix, as described in Variables reference). Patterns begin with a forward slash and can include the asterisk (*) wildcard as part of the pattern.
For example, given a proxy base path of /v1/inventory and resource paths of
you could collect metrics on the following URIs:
- /1 - Collects metrics on the /1 URI
- /abc/123 - Collects metrics on the /abc/123 URI
- /abc/*/dec - Collects metrics on the /abc/123/dec and /abc/456/dec URIs
- After adding a URI pattern, click the check mark icon.
- In the chart at the bottom of the page, click URI patterns in the right pane to show and hide them.
After adding URI patterns, it may take up to an hour for metrics to be available on them. Also, if you delete a URI pattern, metrics for that pattern are no longer collected.
If you see a dramatic rise (a "spike") or a dramatic drop in the traffic, you can get further detail, by checking the Investigate Anomalies checkbox on the API Proxies page and clicking on a point in the chart that corresponds to the spike or drop. In response, you'll be able to view the traffic pattern before, at, and after the spike or drop. You can display the Investigate Anomalies data by app, developer, resource, client IP address, or target URL.
In addition to the Investigate Anomalies checkbox, there are two checkboxes displayed for the APIs on the API Proxies page or for an API on its detail page.
|Show Moving Averages||Check this checkbox to view a moving average for the API. You can check this checkbox for multiple APIs to view a moving average that includes the set of these APIs. A moving average is a series of averages taken over sucessive subsets of a complete set of data. It's especially useful in viewing trends. The moving average is displayed as a band whose limits are +-20% of the calculated moving average data points.|
|Show Alerts||Select this checkbox (on the API Proxies page) to view the number of times that the moving average for the API exceeded the +-20% limit.|
The Geo Map lets you track traffic across locations. Depending on the metric you choose, you use this information to see the quality of services across geos or growth trends by country. You can see all API traffic or individual APIs on different environments over a specific period of time.
Growth trends by location
Use the Message Count metric to see where your traffic is coming from. By using different time periods you can see which countries generate the most traffic at which times. You can also record traffic rates over different time periods to see trends in growth.
- On the Analytics tab, click GeoMap.
- Choose the Message Count from the Metric menu. You can drill down on individual countries to see which country is generating the most traffic.
Quality of Service by location
Use the Average Response Time and Average Endpoint Response Time metrics to see how your backend services are performing by country.
- On the Analytics tab, click GeoMap.
- Choose the response time metric from the menu.
- Choose either Average response time to see the overall quality of service or Average Endpoint response time to see the quality of service for your backend services.
The Traffic Composition report provides insights into the most valuable entities of an API program - apps, developers, APIs, and resources. Use the report to detect business problems such as lower traffic trends or diminishing contribution from key apps and developers. You can also get early notification of new entities that contribute to API traffic, and take actions to respond — for example, you can determine which new developers are highly contributing and include them in nurturing programs.
To view the Traffic Composition report:
- On the Analytics tab, click Traffic Composition.
- Choose a type of entity for the report: APIs, Apps, Developers, or Products.
For each entity type, the report displays the following charts:
|Overall Traffic Trends||The number of messages for each entity of the selected type over time — for example, the number of messages for each API over time.|
|Relative Traffic Contribution||The relative amount of traffic contributed by each entity of the selected type over time — for example, the relative contribution of each app over time.|
|Traffic Trend for entity||Click on a segment of a Relative Traffic Contribution chart to view the traffic trend for the corresponding entity. For example, click on a segment that displays the relative traffic contribution for a specific developer to view the traffic trend for that developer.|
In addition to the analytics available online through Apigee, all organization administrators are automatically subscribed to receive daily analytics summary reports through email.
As an organization administrator, you can unsubscribe through an opt out link in the summary report.
You can export out-of-the-box reports and custom reports to CSV, PDF, or PNG format. Each report includes an export menu where you can select a format for the export.