关于“AK8S-11 Configuring Pod Autoscaling and NodePools”的评价

The activities in this lab does not relate to the topics that were covered in the previous lessons (Nodes AutoScaling using gcloud and GCP console)

Ayman Q. · 已于 about 6 years前审核

Nice

David G. · 已于 about 6 years前审核

Jesus I. · 已于 about 6 years前审核

Víctor M. · 已于 about 6 years前审核

alex k. · 已于 about 6 years前审核

Chung W. · 已于 about 6 years前审核

Bibhuti N. · 已于 about 6 years前审核

Alex L. · 已于 about 6 years前审核

Jiahao J. · 已于 about 6 years前审核

Yinghao L. · 已于 about 6 years前审核

Xiaobo L. · 已于 about 6 years前审核

Momoi H. · 已于 about 6 years前审核

Murilo Z. · 已于 about 6 years前审核

Amar Y. · 已于 about 6 years前审核

Michel Rafael P. · 已于 about 6 years前审核

Carlos R. · 已于 about 6 years前审核

Denys N. · 已于 about 6 years前审核

Pokutnia D. · 已于 about 6 years前审核

andrei t. · 已于 about 6 years前审核

Fredrik J. · 已于 about 6 years前审核

Sharadraj H. · 已于 about 6 years前审核

Initial cluster was too weak to accomplish lab tasks. When I scaled load deployment up to 4 replicas, 3 additional pods for web deployment were not able to be scheduled. I got the following errors: "Pod errors: Unschedulable", "0/2 nodes are available: 2 Insufficient cpu". Only after I scaled load deployment to 0 replicas, web deployment pods were scheduled properly. That is why logically it would be much better to do task 3 at the beginning of the lab, as cluster with 4 nodes was able to schedule all pods correctly. The second bad aspect is connected to theoretical part. We don't have enough theory about settings in spec.containers.resources.limits section. And we don't have knowledge about how to monitor resources consumption properly. But this information is crucial for understanding how autoscaling works.

Alexey I. · 已于 about 6 years前审核

Akilan S. · 已于 about 6 years前审核

Mariia K. · 已于 about 6 years前审核

Andrii P. · 已于 about 6 years前审核