## Known Issues

Populated via [v1.5.0 known issues / FAQ accumulator](https://github.com/kubernetes/kubernetes/issues/37134)

* CRI [known issues and

## Known Issues

- Completed pods lose logs across node upgrade ([#32324](https://github.com/kubernetes/kubernetes/issues/32324))
- Pods are deleted across node upgrade ([#32323](https://github.com/kubernetes/kubernetes/issues/32323))
- Secure master -> node communication ([#11816](https://github.com/kubernetes/kubernetes/issues/11816))

  }

  /**
   * Verify that we don't retry connections on certificate verification errors.
   *
   * http://code.google.com/p/android/issues/detail?id=13178
   */
  @Flaky
  @Test
  fun connectViaHttpsToUntrustedServer() {
    // Flaky https://github.com/square/okhttp/issues/5222
    server.useHttps(handshakeCertificates.sslSocketFactory())
    server.enqueue(MockResponse()) // unused
    assertFailsWith<IOException> {

    enableProtocol(Protocol.HTTP_2)
    postBodyRetransmittedAfterAuthorizationFail("abc")
  }

  /** Don't explode when resending an empty post. https://github.com/square/okhttp/issues/1131  */
  @Test
  fun postEmptyBodyRetransmittedAfterAuthorizationFail() {
    postBodyRetransmittedAfterAuthorizationFail("")
  }

  @Test
  fun postEmptyBodyRetransmittedAfterAuthorizationFail_HTTPS() {

    testResponseCaching(TransferKind.END_OF_STREAM)
  }

  /**
   * Skipping bytes in the input stream caused ResponseCache corruption.
   * http://code.google.com/p/android/issues/detail?id=8175
   */
  private fun testResponseCaching(transferKind: TransferKind) {
    val mockResponse =
      MockResponse.Builder()
        .addHeader("Last-Modified: " + formatDate(-1, TimeUnit.HOURS))

            "BUG: The gradients prefix have been unexpectedly altered when "
            "adding the nodes to the graph. This is a bug. Please file an "
            "issue at https://github.com/tensorflow/tensorflow/issues.");
        return;
      }
      // We have a convoluted scheme here: Using the C++ graph construction API
      // to add potentially many nodes to the graph without running the checks

       * bugs are themselves just symptoms of the same underlying problem):
       *
       * https://github.com/typetools/checker-framework/issues/3030
       *
       * https://github.com/typetools/checker-framework/issues/3236
       */
      String key = (String) requireNonNull(e.nextElement());
      /*
       * requireNonNull is safe because the key came from propertyNames...
       *

       * bugs are themselves just symptoms of the same underlying problem):
       *
       * https://github.com/typetools/checker-framework/issues/3030
       *
       * https://github.com/typetools/checker-framework/issues/3236
       */
      String key = (String) requireNonNull(e.nextElement());
      /*
       * requireNonNull is safe because the key came from propertyNames...
       *

### Bug or Regression

- DRA: fixed several issues related to "allocationMode: all" ([#127565](https://github.com/kubernetes/kubernetes/pull/127565), [@pohly](https://github.com/pohly)) [SIG Node]
- Fix bug where PodCIDR was released before node was deleted ([#128305](https://github.com/kubernetes/kubernetes/pull/128305), [@adrianmoisey](https://github.com/adrianmoisey)) [SIG Apps and Network]

	for {
		n, err := io.CopyN(xioutil.Discard, r.Body, 128*humanize.KiByte)
		if err != nil && err != io.EOF && err != io.ErrUnexpectedEOF {
			// would mean the network is not stable. Logging here will help in debugging network issues.
			if time.Since(connectTime) < (globalNetPerfMinDuration - time.Second) {
				adminLogIf(ctx, err)
			}
		}
		totalRx += n
		if err != nil || ctx.Err() != nil || totalRx > 100*humanize.GiByte {