* NodeController doesn't evict Pods if no Nodes are Ready ([#25571](https://github.com/kubernetes/kubernetes/pull/25571), [@gmarek](https://github.com/gmarek))

          call: Call,
          response: Response,
        ) {
          val bytes = response.body.byteStream()
          assertThat(bytes.read()).isEqualTo('a'.code)
          assertThat(bytes.read()).isEqualTo('b'.code)
          assertThat(bytes.read()).isEqualTo('c'.code)

          // This request will share a connection with 'A' cause it's all done.

		if er.readN == 0 && Equal(etag, er.checksum) {
			return er
		}
	}
	if len(forceMD5) != 0 {
		return &Reader{
			src:      r,
			md5:      NewUUIDHash(forceMD5),
			checksum: etag,
		}
	}
	return &Reader{
		src:      r,
		md5:      md5.New(),
		checksum: etag,
	}
}

// Read reads up to len(p) bytes from the underlying
// io.Reader as specified by the io.Reader interface.

*   Taint key `unreachable` is now in GA.
*   Taint key `notReady` is changed to `not-ready`, and is also now in GA.
*   These changes are automatically updated for taints. Tolerations for these taints must be updated manually. Specifically, you must:
    *   Change `node.alpha.kubernetes.io/notReady` to `node.kubernetes.io/not-ready`
    *   Change `node.alpha.kubernetes.io/unreachable` to  `node.kubernetes.io/unreachable`

      InputStream in2 = closer.register(other.openStream());
      while (true) {
        int read1 = ByteStreams.read(in1, buf1, 0, buf1.length);
        int read2 = ByteStreams.read(in2, buf2, 0, buf2.length);
        if (read1 != read2 || !Arrays.equals(buf1, buf2)) {
          return false;
        } else if (read1 != buf1.length) {
          return true;
        }
      }
    } catch (Throwable e) {

tchedCount","setMatched","contextBackup","dirrunsUnique","token","compiled","filters","unique","getText","isXML","selectors","until","truncate","is","siblings","n","rneedsContext","rsingleTag","winnow","qualifier","self","rootjQuery","parseHTML","ready","rparentsprev","guaranteedUnique","children","contents","prev","sibling","cur","targets","l","closest","index","prevAll","add","addBack","parents","parentsUntil","nextAll","nextUntil","prevUntil","contentDocument","content","reverse","rnothtmlwhi...

    - [Container Images](#container-images-10)
  - [Changelog since v1.28.0](#changelog-since-v1280)
  - [Urgent Upgrade Notes](#urgent-upgrade-notes)
    - [(No, really, you MUST read this before you upgrade)](#no-really-you-must-read-this-before-you-upgrade)
  - [Changes by Kind](#changes-by-kind-10)
    - [Deprecation](#deprecation)
    - [API Change](#api-change-3)
    - [Feature](#feature-8)

* Retry 'connection refused' errors when setting up clusters on GCE. ([#57394](https://github.com/kubernetes/kubernetes/pull/57394), [@mborsz](https://github.com/mborsz))

* YAMLDecoder Read now returns the number of bytes read ([#57000](https://github.com/kubernetes/kubernetes/pull/57000), [@sel](https://github.com/sel))

    * To avoid that negative effect only use per pod metrics from pods that are:
    * - ready (so metrics about them should be valid), or
    * - unready but creation and last readiness change timestamps are apart more than 10s (pods that have formerly been ready and so metrics are in at least some cases (pod becoming unready because of overload) very useful).

    - [Container Images](#container-images-15)
  - [Changelog since v1.25.0](#changelog-since-v1250)
  - [Urgent Upgrade Notes](#urgent-upgrade-notes)
    - [(No, really, you MUST read this before you upgrade)](#no-really-you-must-read-this-before-you-upgrade)
  - [Changes by Kind](#changes-by-kind-15)
    - [Deprecation](#deprecation)
    - [API Change](#api-change-8)
    - [Feature](#feature-15)