Currently we do not monitor if the resource usage or the performance of Ankaios was degraded by a commit or a PR. With time we could get into the situation where Ankaios is not lean and low-profile enough for the target use-case.
To prevent this, we can add automated KPI checks for the performance and the resource usage that are executed as part of the CI pipeline.
Goals
Monitor the performance and resource usage of Ankaios to prevent gradual degradation.
We could connect real HW as an external runner and check in a nightly build if the performance (startup time) and the resource usage (RAM, CPU) is getting worst.
Description
Currently we do not monitor if the resource usage or the performance of Ankaios was degraded by a commit or a PR. With time we could get into the situation where Ankaios is not lean and low-profile enough for the target use-case.
To prevent this, we can add automated KPI checks for the performance and the resource usage that are executed as part of the CI pipeline.
Goals
Monitor the performance and resource usage of Ankaios to prevent gradual degradation.
Final result
Summary
To be filled when the final solution is sketched.
Tasks