- Potentially breaking change - Updating the polling interval for Windows stats collection from 1 second to 10 seconds ([#116546](https://github.com/kubernetes/kubernetes/pull/116546), [@marosset](https://github.com/marosset)) [SIG Node and Windows]

  E0303 17:25:03.436781 63839 summary_sys_containers.go:47] Failed to get system container stats for "/docker": failed to get cgroup stats for "/docker": failed to get container info for "/docker": unknown container "/docker"

- Kubelet/stats: set INFO log level for stats not found in cadvisor memory cache error. ([#125656](https://github.com/kubernetes/kubernetes/pull/125656), [@gyuho](https://github.com/gyuho))

* Bumped Heapster version to 1.4.1: ([#50642](https://github.com/kubernetes/kubernetes/pull/50642), [@piosz](https://github.com/piosz))
    * handle gracefully problem when kubelet reports duplicated stats for the same container (see [#47853](https://github.com/kubernetes/kubernetes/pull/47853)) on Heapster side
    * fixed bugs and improved performance in Stackdriver Sink

## API Changes

*   Rename `SqueezeDims` attribute to `Axis` in C++ API for Squeeze op.
*   `Stream::BlockHostUntilDone` now returns Status rather than bool.
*   Minor refactor: move stats files from `stochastic` to `common` and remove
    `stochastic`.

## Known Bugs

*   Using XLA:GPU with CUDA 9 and CUDA 9.1 results in garbage results and/or
    `CUDA_ILLEGAL_ADDRESS` failures.

* client-go and kubectl no longer write cached discovery files with world-accessible file permissions ([#77874](https://github.com/kubernetes/kubernetes/pull/77874), [@yuchengwu](https://github.com/yuchengwu))
* Check if container memory stats are available before accessing it ([#77656](https://github.com/kubernetes/kubernetes/pull/77656), [@yastij](https://github.com/yastij))

- kubelet: use the CRI stats provider if `PodAndContainerStatsFromCRI` feature is enabled ([#126488](https://github.com/kubernetes/kubernetes/pull/126488), [@haircommander](https://github.com/haircommander)) [SIG Node]

- Fix a bug where CRI implementations that use cAdvisor stats provider (CRI-O) don't evict pods when their logs exceed ephemeral storage limit. ([#108115](https://github.com/kubernetes/kubernetes/pull/108115), [@haircommander](https://github.com/haircommander)) [SIG Node]

- Fixed a bug in the CSI metrics that does not return not supported error when a CSI driver does not support metrics. ([#79851](https://github.com/kubernetes/kubernetes/pull/79851), [@jparklab](https://github.com/jparklab))
- Fix disk stats in LXD using ZFS storage pool and CRI-O missing network metris bug ([#81972](https://github.com/kubernetes/kubernetes/pull/81972), [@dashpole](https://github.com/dashpole))

## Notable Features

### Beta

### Feature

- Add --dry-run to kubectl delete, taint, replace ([#88292](https://github.com/kubernetes/kubernetes/pull/88292), [@julianvmodesto](https://github.com/julianvmodesto)) [SIG CLI and Testing]
- Add huge page stats to Allocated resources in "kubectl describe node" ([#80605](https://github.com/kubernetes/kubernetes/pull/80605), [@odinuge](https://github.com/odinuge)) [SIG CLI]