## Changelog since v1.28.10

## Changes by Kind

### API Change

- Improved scheduling performance when many nodes, and prefilter returns 1-2 nodes (e.g. daemonset)
  

			if len(j.PartIndices) == len(fi.Parts) {
				fi.Parts[i].Index = j.PartIndices[i]
			}
		}
	}

	// fi.Erasure.Checksums - is left empty since we do not have any
	// whole checksums for many years now, no need to allocate.

	fi.Metadata = make(map[string]string, len(j.MetaUser)+len(j.MetaSys))
	for k, v := range j.MetaUser {
		// https://github.com/google/security-research/security/advisories/GHSA-76wf-9vgp-pj7w

)

const letterBytes = "abcdefghijklmnopqrstuvwxyz01234569"
const (
	letterIdxBits = 6                    // 6 bits to represent a letter index
	letterIdxMask = 1<<letterIdxBits - 1 // All 1-bits, as many as letterIdxBits
	letterIdxMax  = 63 / letterIdxBits   // # of letter indices fitting in 63 bits
)

// Random number state.
// We generate random temporary file names so that there's a good

		if name == "" {
			continue
		}

		fileName = part.FileName()

		// Multiple values for the same key (one map entry, longer slice) are cheaper
		// than the same number of values for different keys (many map entries), but
		// using a consistent per-value cost for overhead is simpler.
		maxMemoryBytes := 2 * int64(10<<20)
		maxMemoryBytes -= int64(len(name))
		maxMemoryBytes -= mapEntryOverhead

semi-abstract instruction set and turn it into instructions to be
input to the linker.
If you want to see what the instructions look like in assembly for a given architecture, say amd64, there
are many examples in the sources of the standard library, in packages such as
<a href="/pkg/runtime/"><code>runtime</code></a> and
<a href="/pkg/math/big/"><code>math/big</code></a>.

        for (int i = 0; i < DRAIN_THRESHOLD * 2; i++) {
          map.get(keyOne);
          assertTrue(segment.recencyQueue.size() <= DRAIN_THRESHOLD);
        }

        // get over many different keys

        for (int i = 0; i < DRAIN_THRESHOLD * 2; i++) {
          map.put(new Object(), new Object());
        }
        assertTrue(segment.recencyQueue.isEmpty());

- Improved scheduling performance when many nodes, and prefilter returns 1-2 nodes (e.g. daemonset)
  

* Update kubectl help for 1.2 resources ([#23305](https://github.com/kubernetes/kubernetes/pull/23305), [@janetkuo](https://github.com/janetkuo))
* Support addon Deployments, make heapster a deployment with a nanny. ([#22893](https://github.com/kubernetes/kubernetes/pull/22893), [@Q-Lee](https://github.com/Q-Lee))

type scriptedReader []func(p []byte) (n int, err error)

func (sr *scriptedReader) Read(p []byte) (n int, err error) {
	if len(*sr) == 0 {
		panic("too many Read calls on scripted Reader. No steps remain.")
	}
	step := (*sr)[0]
	*sr = (*sr)[1:]
	return step(p)
}

func newScriptedReader(steps ...func(p []byte) (n int, err error)) io.Reader {

    )
    val url = server.url("/a")
    assertThat(get(url).body.string()).isEqualTo("a")
    val i = cache.urls()
    assertThat(i.next()).isEqualTo(url.toString())
    i.remove()

    // Too many calls to remove().
    assertFailsWith<IllegalStateException> {
      i.remove()
    }
  }

  @Test
  fun elementEvictedBetweenHasNextAndNext() {
    // Put a response in the cache.