This article applies to:
- Product edition: All versions
- Feature Category: Integration.
Tanzu Application Service is a popular platform for building cloud-native applications. This integration uses the VMware Tanzu Observability by Wavefront Nozzle v4. The integration offers TAS data egress, system dashboards, and alerts. A TAS environment with many foundations can be difficult to debug. Tanzu Observability can help.
It has multiple chart which help to identify usage of sources and performance of application.
Problem : Why is the percentage of “Application CPU% “ chart is so high for this application?
Description : Here, we are going to describe more about “ Application CPU %" Chart in TAS integration.
This chart help as a list of Application Instances ranked by the highest utilization of their CPU entitlement.
It could be possible that the value of “Application CPU%” chart is more than 100%. % > 100 is what customer can expect if the diego cell has spare CPU capacity and the container is trying to use more than its entitled.
If the host has spare CPU, it doesn’t throttle the CPU usage. But if the host has more demand on its CPU, it would try to throttle all the containers “fairly” based on their entitlement.
The more than 100% usage also indicates that if the cell were to be fully CPU utilized, then that app would be waiting on requested CPU time because it wouldn't be able to "steal" that from other running containers. Theoretically that could degrade the performance of that application.
Please go through below doc to understand more about TAS integration.
Solution: The solution there would be to scale up the Diego cells, either horizontally or vertically to add CPU processing power. But if the application is functioning fine then that indicates you have sufficient free CPU capacity on the cells and no need to do anything.
If performance of application is not working as expected than raise a case with Aria Operations for Applications support team.