Technical features of the Marvin_ portal

The dashboard in the Marvin_ portal shows you the status of your applications at a glance.

Some of the insights:

• the last measured response and processing time;
• exceedance of KPI's (and SLA's);
• errors that occurred in click paths;
• trends in response times and errors.

Application tests are composed of multiple steps, which means each step can be measured individually. By plotting these steps separately it's easy to see where any performance issues or other problems are occurring. The well-structured detail pages provide you with extra information about a test run, including screenshots taken at the time of measurement.

In a test you can configure a Status Threshold for each step. This controls the displayed status of that step (and the overall test), which is based on a violation of the run time or error count limits. Two limits can be set to mark the status as either "warning" or "critical". If desired, the thresholds can be automatically calculated based on the results of the previous 30 days.

It is possible to specify multiple notifications for each test with their own separate destinations. You can choose to receive a notification only after a certain number of repeated violations (or successes) of the configured rules. Once a notification for a problem is sent, you can also choose to regularly receive repeat notifications until the problem is resolved.

Marvin_ Application Monitoring uses OpenRunner to perform benchmarks and automated functional tests. This makes it possible to test even the most heavy front-end web applications. OpenRunner can very accurately spot functional changes and gives a precise report of the response and processing times.

The break down screens give detailed information about the processing that takes place inside the web browser. This contains not just the response times for requests, but also the amount of time taken to process specific DOM elements. This enables you to easily go from "Which step was slow?" to "Why was this step slow?"

The wallscreen is designed to securely (authentication is token-based) show the status of your applications on large monitoring screens. All the display settings are stored in the web address (URL) which makes sharing of configurations easy. This means each team can configure a monitoring screen with their own selection of applications, and control the level of detail.

Through an API annotations can be added to the graphs. By means of the provided API this is integrated with your CI/CD flow. Some examples of useful annotations are deploys and maintenance windows. This makes it easy to see whether (and how) a change affected your application.