* and halfSquare are integers, this is equivalent to testing whether or not x <=
         * halfSquare. (We have to deal with overflow, though.)
         *
         * If we treat halfSquare as an unsigned long, we know that
         *            sqrtFloor^2 <= x < (sqrtFloor + 1)^2
         * halfSquare - sqrtFloor <= x < halfSquare + sqrtFloor + 1

   * remaining 32 bits interpreted as an IPv4 address. These are conventionally represented in
   * string literals as {@code "::192.168.0.1"}, though {@code "::c0a8:1"} is also considered an
   * IPv4 compatible address (and equivalent to {@code "::192.168.0.1"}).
   *
   * <p>For more on IPv4 compatible addresses see section 2.5.5.1 of <a target="_parent"

   * are thus effectively static during the lifetime of a cache entry.
   *
   * <p>When the weight of an entry is zero it will not be considered for size-based eviction
   * (though it still may be evicted by other means).
   *
   * <p><b>Important note:</b> Instead of returning this as a {@code CacheBuilder}

  - A cluster-scoped object with an ownerReference referencing a uid of a namespaced kind is now consistently treated as though that owner is not resolvable, and the child object is ignored by the garbage collector. ([#92743](https://github.com/kubernetes/kubernetes/pull/92743),...

     * present in the cache. Active values consist of live values, which are returned by cache
     * lookups, and dead values, which have been evicted but awaiting removal. Non-active values
     * consist strictly of loading values, though during refresh a value may be both active and
     * loading.
     */
    boolean isActive();
  }

  /** Placeholder. Indicates that the value hasn't been set yet. */

    assertThat(stream1.writeBytesMaximum)
      .isEqualTo(Settings.DEFAULT_INITIAL_WINDOW_SIZE.toLong())

    // Another stream should be able to send data even though 1 is blocked.
    val stream2 = connection.newStream(headerEntries("b", "banana"), true)
    val out2 = stream2.sink.buffer()
    out2.writeUtf8("foo")
    out2.flush()
    assertThat(connection.writeBytesTotal)

    // such that none of the subsequent iterations observed it, despite the fact that at every point
    // in time it was present somewhere int the map. This becomes increasingly unlikely as
    // CONTAINS_VALUE_RETRIES increases, though without locking it is theoretically possible.
    Segment<K, V, E, S>[] segments = this.segments;
    long last = -1L;
    for (int i = 0; i < CONTAINS_VALUE_RETRIES; i++) {
      long sum = 0L;

      .assertRequestUrl(server.url("/"))
      .assertSentRequestAtMillis(request2SentAt, request2ReceivedAt)
      .assertReceivedResponseAtMillis(request2SentAt, request2ReceivedAt)

    // Check the cache response. Even though it's a miss, we used the cache.
    cacheMiss
      .cacheResponse()
      .assertCode(200)
      .assertHeader("Donut", "a")
      .assertHeader("ETag", "v1")
      .assertRequestUrl(server.url("/"))

- Fixes a problem with 63-second or 1-second connection delays with some VXLAN-based
  network plugins which was first widely noticed in 1.16 (though some users saw it
  earlier than that, possibly only with specific network plugins). If you were previously
  using ethtool to disable checksum offload on your primary network interface, you should

- The v1 Endpoints API is now officially deprecated (though still fully supported). The API will not be removed, but all users should use the EndpointSlice API instead. ([#130098](https://github.com/kubernetes/kubernetes/pull/130098), [@danwinship](https://github.com/danwinship)) [SIG API Machinery and Network]...