You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jan 11, 2023. It is now read-only.
For ACS- and RP- created kubernetes clusters, agent nodes are created with the default value for allocatable pods of 110. For large VM sizes, that could be not desirable. A better way would be to calculate an ideal value for different VM sizes, and add the --max-pods flag to parts/kuberneteskubelet.service. This could be part of the OrchestratorProfile.KubernetesConfig options.
The text was updated successfully, but these errors were encountered:
I'd rather it be part of the agent pool profile, since that's where SKU is set. We could allow the user to override a default of 110 rather than try to do some magical calculation. And I'd argue that if there were a magic calculation, it would be based on mem/cpu and live in kubelet itself rather than here.
For ACS- and RP- created kubernetes clusters, agent nodes are created with the default value for allocatable pods of 110. For large VM sizes, that could be not desirable. A better way would be to calculate an ideal value for different VM sizes, and add the --max-pods flag to parts/kuberneteskubelet.service. This could be part of the OrchestratorProfile.KubernetesConfig options.
The text was updated successfully, but these errors were encountered: