The limitations I notice include the instability of the agent, which crashes, and the application itself consuming memory under heavy loads. Additionally, the customizations are not easy to build, especially for new parsers and functionalities that are just supported out-of-the-box. Splunk AppDynamics needs to be more user-friendly so that even someone who does not know what is happening should be able to easily find the root cause of issues, as the tool is specifically designed for application developers, not from a user perspective for problem identification. My experience with the application performance monitoring capabilities of Splunk AppDynamics indicates that it is complex. The correlation from the front end to the back end is not working properly. In an application using the EUM feature and having the app agent, the app agent gives me the transaction data, but the EUM data does not correlate to the backend transactions, and I do not know why. It can be very complex to find the root cause in those scenarios, as there is a disparity in metrics, making it challenging. Even though Splunk AppDynamics is trying, my very large environment presents hurdles that medium-scale organizations might not experience. After spending around three to four years trying to get value from it, I still feel a lack of satisfaction. To get closer to a rating of 10 in the future, I would like to see whatever features are being provided made to work, particularly in large environments. Splunk AppDynamics claims it can do 100 things, but if those things are not functioning cohesively, it is ineffective. Instead of offering 100 features that do not work together, it would be better to focus on a smaller set of features that operate effectively. Changes happen daily with new applications and technologies, and Splunk AppDynamics needs to adapt more swiftly to keep up. The feature that helps detect and protect against vulnerabilities does raise concerns. I experience vulnerabilities in Splunk AppDynamics that I communicate to support for fixes, and they often respond by indicating that some fixes are available in new versions. This creates a continuous cycle of raising issues and waiting for resolutions through version updates.
