
GKE Autopilot, with Yochay Kiriaty
Kubernetes Podcast from Google
00:00
Will Gk Audo Pilot Handle This Optimization?
The debdadog container survey says that almost half of containers are using less than 30 % of the cpu that they've actually requested. With cobernetes, you have an allocatable resource, right? Ip and memory, and then there is the utilization. The alocatable basically becomes the hundred%. And what wehave found is the ondelocatable people, pretty much as the daladog server says,. We have exactly the same data beha in our fleet that shows that customers are doing really poor job in a location. This week, i believe this we call uspic an autopilet. Will obviously pick that up by default as well. So it's enable
Transcript
Play full episode