Observing Your Apps Running on SAP BTP

When providing and running applications in the cloud, ops teams can face different issues and challenges within their cloud setup.

Click on the numbers below to discover some examples of potential issues.

To operate a cloud application reliably, it's key to handle such issues in time.

Therefore, the main focus of the operations capabilities in SAP BTP is to provide the DevOps teams with the right means (tools and approaches) to keep their solutions up and running.

To do this, they need to be notified about potential issues in time.

Also, to cope with issues outside of working hours, DevOps teams need to automate remediation actions that can fix the issues.

Our friends at Rotating Banana developed an SAPUI5 Web application deployed on the SAP BTP Cloud Foundry environment. However, the app becomes randomly inaccessible for different users, and as of now, there are not any alerts sent out. As you can imagine, this is not well received by clients, managers, and the DevOps team.

As a member of the Ops team in charge to run this application, you're tasked to solve the issue using local technical monitoring and automation services (that complement SAP Cloud ALM as central operations platform of SAP BTP) within the SAP BTP DevOps portfolio:

1. SAP Alert Notification service with its main focus on alerting.
2. SAP Automation Pilot with its main focus on Ops automation.

To be able to implement the notification and automation services, you need first to know what is their function and how they work. Discover in the next sections the details for each of the two services before resuming the task at Rotating Banana.

Watch the video animation below to find out more about this service.

The service is designed to work with low latency, even when it's under a heavy workload.

Watch the video animation below to find out more about the SAP Automation Pilot.

Getting back to Rotating Banana, remember that you're tasked with implementing an integrated monitoring and availability checks solution for a SAPUI5 app deployed in SAP BTP.

Mind that the solution architecture has the following details:

• The cloud application does provide a health-check endpoint.
• SAP Automation Pilot will be used as automation engine that will trigger periodically health-check status calls to this app health-check end point.
• Based on the responses received by SAP Automation Pilot, there will be a condition within the SAP Automation Pilot so that:
  • If the HTTP response status code is "200", the application is considered accessible and therefore healthy - no alert is sent out.
  • If the HTTP response status code is different than "200", the application is considered not to be accessible - an alert is to be sent out to the DevOps team.

Click on the connecting arrows in the interactive application below to visualize details.

To realize the desired solution, you will perform the following operations (use the links to explore the actions in detail):

1. Create and maintain a command within SAP Automation Pilot that calls the application health-check endpoint every 5 minutes.
2. Set a condition within the SAP Automation Pilot command that sends a notification to SAP Alert Notification service if the response is different than "200". For this, SAP Automation Pilot shall produce a custom event that gets consumed by SAP Alert Notification service.
3. In SAP Alert Notification service, create a subscription that is configured to filter received events that fall into predefined conditions. These special events shall fire a specific action, in our case the sending of an email to the DevOps team.

You have gained the basic understanding about monitoring and automated remediation capabilities provided by two services within the SAP BTP.

Moreover, you have learned how to implement by yourself a basic health cloud application monitoring mechanism by tools provided by SAP to support your DevOps strategy without the need of any coding from your side.

In case you want to learn more about this example or implement it in your landscape, you can find a detailed step-by-step implementation guide on GitHub.