Enhancement name:
Pod Level Resource Specifications
Enhancement status (alpha/beta/stable):
currently in alpha, trying for beta in 1.34
SIG
Node, Autoscaling
k/enhancements repo issue #:
PR #’s:
Promotion of feature gate to beta - https://github.com/kubernetes/kubernetes/pull/132999
actively being reviewed, waiting for LGTM and approval labels - https://github.com/kubernetes/kubernetes/pull/132634
has LGTM labels, waiting on approvals https://github.com/kubernetes/kubernetes/pull/132430
might get approved before the code freeze - https://github.com/kubernetes/kubernetes/pull/131089
Additional time needed (in days):
6 (until Wednesday)
Reason this enhancement is critical for this milestone:
Postponing the beta launch of this feature would drastically slow both adoption and our ability to gather essential feedback.
Risks from adding code late: (to k8s stability, testing, etc.)
The risk is minimal as this feature requires setting resources in pod spec at pod-level explicitly, because of which the code path is isolated.
The feature is actively reviewed and tested.
Risks from cutting enhancement: (partial implementation, critical customer usecase, etc.)
This enhancement is key for AI/ML workloads and boosts resource use. It handles their unpredictable resource needs, especially when it's hard to decide resource allocation upfront for each container. It also lays the groundwork for the Dynamic Containers proposal. We haven't been able to drive adoption of this feature as it is still in alpha, delaying the beta further would slow down the feature adoption.
No concerns from sig-autoscaling.
--
You received this message because you are subscribed to the Google Groups "Autoscaling Kubernetes" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubernetes-sig-auto...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/kubernetes-sig-autoscaling/CAJo%3DGk4QLq%3D64RXFRuAkqUTUviQ2PFf%3Dma%3Dr_smCVjXSVcAwWA%40mail.gmail.com.