* Mark multi-scheduler graduation to beta and then v1. ([#38871](https://github.com/kubernetes/kubernetes/pull/38871), [@k82cn](https://github.com/k82cn))
* Scheduler treats StatefulSet pods as belonging to a single equivalence class. ([#39718](https://github.com/kubernetes/kubernetes/pull/39718), [@foxish](https://github.com/foxish))

### SIG Scheduling

- The equivalence class cache has been redesigned to be a two level cache, resulting in a significant increase in scheduling throughput and performance. ([#65714](https://github.com/kubernetes/kubernetes/pull/65714), [@resouer](https://github.com/resouer))

* Use monotonically increasing generation to prevent scheduler equivalence cache race. ([#67308](https://github.com/kubernetes/kubernetes/pull/67308), [@cofyc](https://github.com/cofyc))

   * implicit).
   */
  protected Ordering() {}

  // Instance-based factories (and any static equivalents)

  /**
   * Returns the reverse of this ordering; the {@code Ordering} equivalent to {@link
   * Collections#reverseOrder(Comparator)}.
   *
   * <p><b>Java 8+ users:</b> Use {@code thisComparator.reversed()} instead.
   */

  @CanIgnoreReturnValue
  @Override
  Hasher putLong(long l);

  /** Equivalent to {@code putInt(Float.floatToRawIntBits(f))}. */
  @CanIgnoreReturnValue
  @Override
  Hasher putFloat(float f);

  /** Equivalent to {@code putLong(Double.doubleToRawLongBits(d))}. */
  @CanIgnoreReturnValue
  @Override
  Hasher putDouble(double d);

  /** Equivalent to {@code putByte(b ? (byte) 1 : (byte) 0)}. */
  @CanIgnoreReturnValue

  /**
   * Determines if the given iterable contains no elements.
   *
   * <p>There is no precise {@link Iterator} equivalent to this method, since one can only ask an
   * iterator whether it has any elements <i>remaining</i> (which one does using {@link
   * Iterator#hasNext}).
   *
   * <p><b>{@code Stream} equivalent:</b> {@code !stream.findAny().isPresent()}
   *
   * @return {@code true} if the iterable contains no elements

   *
   * <p><b>Comparison to {@code java.util.Optional}:</b> this method is equivalent to Java 8's
   * {@code Optional.ofNullable}.
   */
  public static <T> Optional<T> fromNullable(@Nullable T nullableReference) {
    return (nullableReference == null) ? Optional.<T>absent() : new Present<T>(nullableReference);
  }

  /**
   * Returns the equivalent {@code com.google.common.base.Optional} value to the given {@code

   *
   * <p><b>Comparison to {@code java.util.Optional}:</b> this method is equivalent to Java 8's
   * {@code Optional.ofNullable}.
   */
  public static <T> Optional<T> fromNullable(@Nullable T nullableReference) {
    return (nullableReference == null) ? Optional.<T>absent() : new Present<T>(nullableReference);
  }

  /**
   * Returns the equivalent {@code com.google.common.base.Optional} value to the given {@code

   */
  @Nullable Range<C> rangeContaining(C value);

  /**
   * Returns {@code true} if there exists a non-empty range enclosed by both a member range in this
   * range set and the specified range. This is equivalent to calling {@code
   * subRangeSet(otherRange)} and testing whether the resulting range set is non-empty.
   *
   * @since 20.0
   */
  boolean intersects(Range<C> otherRange);

  /**

When the Kotlin DSL was introduced, we made an effort to add non-Groovy equivalents for all APIs.
This has been mostly done, but there remain a few holdouts (fixing these is out of scope).

To keep the Groovy DSL ergonomic, we generate methods as necessary from the non-Groovy equivalents.

Doing this provides the following specific benefits: