The New Face of Analytics
An Intro to APIM 4.0 Analytics
API Manager 4.0 went out 5th of May, offering a bunch of exciting features. Among them, moving Analytics to the cloud was one important change, which allowed customers to enjoy the full benefits of the Analytics without going through the pain of setting it up. Along with the move, the Analytics Portal (or the Dashboards) underwent a complete revamp, making the UIs more intuitive and user-friendly.
Through this post, I thought of highlighting the new features, explaining the reasons behind the cloud move
From on-premise to the Cloud
Those who are familiar with previous API Manager releases might remember that Analytics was offered as an on-premise distribution. Though the standalone distribution helped customers to quickly set up and try out, creating production-grade setups required a bit more work. Things like configuring High Availability, running profiles separately, sometimes required customers to know internal details about the platform. And whenever technologies used for the Analytics changed customers had to change their DevOps processes, which often turned to be a little too much for showing API Stats.
The cloud version takes this burden away. If you’ve tried out the new version, you’ll realize that all it takes to use Analytics is getting a token from the Portal and putting it in a config file. No downloading is needed. Since the processing and storage happen in the cloud, you don’t have to worry about setting up and managing the setup.
The New Look and Feel — A window to the API Ecosystem
Another change the cloud move prompted is the new look and feel. While it provides certain interactive features like zoomable charts, clickable legends, and navigation options, the greatest value it offers is in the intuitive way charts are arranged. The new portal gives a broader lens to look at different parts of the API ecosystem.
API Ecosystem is the combination of API Providers, End-users, Consumers (Application Developers), partners, and the whole bunch of people who contribute to increasing the value proposition of APIs. Even the DevOps and Infra teams that take care of the system underneath become major actors of it. Since it’s the collaboration between different parties that evolves the Ecosystem, providing the stats each party is looking for is important to its organic growth. The new portal concisely provides these stats.
For API Providers
If APIs are thought of as Digital products, then API Providers are those who build the products, whose objective is to create value by converting an organizational asset to an API. Typically they’d like to know how well their APIs are adopted, whether the APIs are used by a range of Consumers or limited to a particular segment, and how much traffic APIs are receiving. As they make changes and roll out newer versions, they’d like to compare, how well the latest versions are doing.
The traffic page offers these insights, allowing API Providers to filter traffic by API and Application. The pie chart shows the traffic composition by the Application, which is useful to get an idea about the adoption. If needed, the view can be switched to an overtime chart.
The Cache page is a new addition, which is helpful to determine whether caches are used effectively. Cache hits would appear when Response Caching is enabled in the APIs and responses are served from the cache.
For DevOps/Operations
The operations team who are working to keep the system up and running would have to look at stats like error rate, the latency of the APIs. If there are too many errors, or if the APIs are taking long to respond then it can impact the SLAs. When an API is degrading, they quickly have to find out what the cause is. With the new Latency and Error views provided, the operations team can quickly identify whether the problem is with the API or with the Backend.
Latency breakdown helps to narrow down an issue further, by showing request and response mediation latencies, which effectively gives the latency added by the Gateway. If the request mediation increases, even when the API doesn’t have a custom In-Sequence, it usually means Key Validation is taking time — something which should be fixed at the infra level.
For API Consumers
If APIs are products, then Applications are the stores that arrange and sell them. When a store owner knows how to arrange a product to best show its value, both the store and the product owners get the benefit.
Similarly, API Consumers (Application Developers) need statistics to gauge how their Applications are doing. Both Traffic and Report pages give an Application-wise breakdown, which is helpful to find Applications that are generating traffic. On the Error page, errors generated by the Application can be filtered, through which quality issues and the needs to upgrade can be identified.
Since the Traffic page allows filtering Traffic by an Application, the Consumers can assess whether the end-users are following the expected user journey. Through some careful analysis, the points where the flow breaks can be identified.
Know thy Users
Knowing where your end-users come from is crucial to growing your API program. Details like which Platforms they are using, whether the majority of users come from a web page or a mobile device, what sort of browsers they use, help to decide how to improve reach further. Details on the Devices page offer some insights, through which customers can get some of the above stats.
Invite and Share
Riding on the power of Choreo, the new Analytics makes a change in sharing data and managing access. When setting up Analytics for multiple environments, with the on-premise version, a separate setup would have to be created, which will only show stats related to a particular environment. If users need to view stats in multiple environments, they would have to log in to multiple portals. However, in the new solution, switching between environments is a matter of changing a drop-down.
With the previous versions, data for tenants were presented in siloed views. Only users within the tenant were able to see the tenant’s data. While technically this separation was correct, practically tenancy in APIM was used to represent different departments of a larger organization. Which often prompted a set of privileged users to view data across tenants. To better cater to this, the new platform lets users switch between tenants through another drop-down.
The new Analytics Solution is a part of Choreo, which means, that features like inviting users are available innately. If you’re an Admin User setting up Analytics for your organization, then you can select users who wish to see stats. The person who configures Analytics doesn’t need to be an Admin of the APIM setup.
Give it a try
Didn’t I mention configuring Analytics has never been easier? You simply log into the Cloud portal, create a config key for the environment, and put it in the deployment.toml. That’s it. If you’re not believing, you can check the official document. When the Gateways startup they’ll talk to a service and get configs needed to Publish events.
I hope you got a fair idea about the new Analytics, features it offers, and the troubles it saves you from. You can try it either by downloading API Manager 4.0 or by managing APIs through Choreo. If you need any help, you can reach the team through the Slack Channel or the Contact Form.
