Introducing Analytics- Digital Transformation with IBM API Connect
This section will introduce the analytics that comes with API Connect and how to begin customizing it for your needs. The analytics that come with API Connect is implemented with open source from the Elastic Stack (ELK – Elasticsearch, Log Stash, and Kibana). In version 5 of API Connect, this component was installed on the management server. With version v10, analytics is separated from the management service and provides greater flexibility and better performance. Analytics servers are associated with Gateways. The administrator configures this relationship in the Topology view in the Cloud Manager. The Gateway collects the metric data for each API executed and sends the data to the associated analytics server.
Architect Tip
While this book is primarily for architects and developers, understanding some basic planning and configurations for analytics can be valuable knowledge. For instance, you should consider using MQ queueing to ensure your Gateway analytics aren’t lost due to the loss of network storage. This will provide extra reliability in maintaining your analytics. You should also be aware that you can have multiple Gateways associated with a single Analytics service, but keep in mind that each Gateway can be associated with only one Analytics service. There are other planning considerations an architect should be aware of so you should review the IBM documentation on planning the Analytics deployment: https://www.ibm.com/docs/en/api-connect/10.0.x?topic=deployment-planning-your-analytics.
Understanding the analytics initial setup
Although the deployment of the Analytics server is outside the scope of this book there are a few contextual bits of information that will help you better understand the scope of the analytics provided by API Connect. Having this background will help you understand how analytics are captured and certain organizational traits that will make your analytics more aligned to your deployment. Next, you will learn, at a high level, how the analytics server is set up in the Cloud Manager topology.
Associating the Analytics server in the Cloud Manager topology
After your API Connect components (Cloud Manager/API Manager, Gateways, Portal, and Analytics server) are installed the cloud administrator will set up within the default availability zone the API Connect components. You can set up components in different availability zones but that is a topic beyond the scope of this book.
In Figure 15.12, you see the Topology page from the Cloud Manager user interface. What you are seeing are the components that are implemented in the Default Availability Zone. The Cloud Manager is not shown because it exists at a higher level and is the manager of the overall topology across the availability zone.
Figure 15.12 – Topology showing associating Analytics with a Gateway
Looking closer at Figure 15.12 you will see two references to DataPower Gateways. DPGatewayV5 is for the version 5 compatible Gateway and DPGatewayAPIGW is for the new higher performance Gateway simply known as the API Gateway. You also see a single Analytics Service that was deployed. In this simple case, there is only one analytics server. That is not a limitation because you can have more analytics servers deployed in different availability zones or the same availability zone as the other services. There is also a Portal server, which you learned about earlier in this chapter.
When the administrator is configuring the topology, one task is to associate a Gateway to an Analytics server. In Figure 15.12, you see that the DPGatewayV5 service is associated with the analytics-services service. So that means all analytics generated by the DPGatewayV5 Gateway will be sent to that specific Analytics Service. You’ll also notice that DPGatewayAPIGW is not associated yet because it shows the Associate analytics service link.
If the administrator wants to associate that Gateway with the same analytics service, they simply click the link and will be presented with a popup to perform the association. Figure 15.13 shows that popup:
Figure 15.13 – Associate an Analytics server with a Gateway
Once this association is completed the Cloud Manager interacts with the Gateway to inform the Gateway to send its analytics to the associated analytics service. Once the Gateway is associated with an analytics service it cannot be associated with another analytics service.
You have been briefly introduced to how the topology setup is accomplished and should have a good idea of how Gateways send analytics to a designated analytics service. There are more sophisticated scenarios that can be configured. If you are interested in diving even deeper into this subject, you can review https://www.ibm.com/docs/en/api-connect/10.0.x?topic=topology-creating-availability-zone to learn more about the flexibility of availability zones.
Now it’s appropriate to begin learning how to use the Analytics user interface within API Connect. You’ll be learning where you can find the Analytics user interfaces and how to create or modify existing dashboards and visualizations.