Setting Up & Using API Analytics & Observability
Learn how to configure Revenium's analytics suite, including API usage & performance reporting and alerting.
API analytics are created automatically when using any of our API gateway connectors. (i.e. Kong, Mulesoft, aws, azure, and many others). API analytics are not generated for non-API data sources, though you can add your own performance data to your usage calls via our metering API (see example here).
If you are using Revenium for usage-based metering & billing without an API data source, you will find the product analytics guide more useful.
What Can you Do with Revenium API Analytics?
Revenium API analytics helps to answer questions like these:
API Performance
Which of my APIs has the highest error rates? Do these always happen at the same time?
When did our performance problems begin and which customers have been affected?
Which endpoints regularly deliver a poor customer experience?
How did overall performance change after our last product release?
Is the API performance adequate for international customers?
Customer Usage Data
Which customers have the highest usage?
Which customers have shown a drop in usage in the last 30 days? Did this correlate to any changes made by our team?
Setting Up API Analytics
1. Connect Your Data
Use one of our API gateway connectors in this section to begin sending performance data to Revenium. If you want to send transactions with analytics data to Revenium via the command line, you can use the examples here.
2. Confirm Successful Data Connection
Within 2 minutes of sending transaction data to Revenium, the transactions will be visible in the analytics logs section of Revenium. API calls to Revenium's metering API that return a 400 or 500 response will provide troubleshooting information in the API response. For any transactions successfully sent to Revenium, you will also find informational logs in the System Logs.
3. Review Auto-Detected Data Sources
Revenium will auto-create data sources for each different URL it observes from your traffic, after which you can refine the auto-detected sources as needed.
Note that when creating or editing your API data sources, you can define the sources to be as generic or specific as you want by refining the filters available on the Source creation form. For example, the API in Figure 1 could be broadened to include all methods (vs. POSTS only in the image), or the Resource field could be extended to include only a more specific endpoint (i.e. https://my-api.com/v2/order/DE/shipping
vs https://my-api.com/v2/order
).
Analytics will be captured distinctly for each source you create, so if you wished to store analytics on GET transactions separately from POST transactions, you could do so by cloning the source below and changing only the method.
4. Review Analytics Transactions in Revenium
Once step #2 has completed successfully, all of Revenium's API Analytics reporting will be generated and automatically updated as new transactions are generated from your system. Once setup, you will be able to report on the following:
Abnormal transaction latency or error rates
The API experience by customer, which helps to identify if API-problems are customer-specific or global
Compare performance over different periods of time, which helps to quickly identify the effects of code changes on the customer experience
View traffic by geography to understand how performance varies depending on the location of your customers
5. Set Up Alerting for API Performance & Availability
Revenium allows you to set up alerting for a variety of performance conditions. For example, you can configure:
alerts for slow application or gateway response times
for APIs that stop processing traffic
for abnormal error rates
Each of these alerts can be configured globally (across all data sources) on configured on each source individually.
If your API Sources will be added to Products for usage-based billing, you can also follow the guide for setting up Product Analytics.
Last updated