- The HugePageStorageMediumSize feature gate is now on by default allowing usage of multiple sizes huge page resources on a container level. ([#90592](https://github.com/kubernetes/kubernetes/pull/90592), [@bart0sh](https://github.com/bart0sh)) [SIG Node]

* Fixes validation of VolumeAttachment API objects created with inline volume sources. ([#80945](https://github.com/kubernetes/kubernetes/pull/80945), [@tedyu](https://github.com/tedyu))
* Fix kubelet errors in AArch64 with huge page sizes smaller than 1MiB ([#78495](https://github.com/kubernetes/kubernetes/pull/78495), [@odinuge](https://github.com/odinuge))

    Bu adımda cache kullanmak, geliştirme sırasında image'ı tekrar tekrar build ederken size çok **zaman** kazandırır; her seferinde bağımlılıkları **indirip yüklemek** zorunda kalmazsınız.

5. `./app` dizinini `/code` dizininin içine kopyalayın.

        ' binary GPU code, or as "sm_xy" to only include the binary '
        'code.\nPlease note that each additional compute capability '
        'significantly increases your build time and binary size, and that '
        'TensorFlow only supports compute capabilities >= 3.5 [Default is: '
        '%s]: ' % default_cuda_compute_capabilities)
    hermetic_cuda_compute_capabilities = get_from_env_or_user_or_default(

			accessKey:          credentials.AccessKey,
			secretKey:          credentials.SecretKey,
			expectedRespStatus: http.StatusNoContent,
		},
		// Test case - 2.
		// Setting the content length to be more than max allowed size.
		// Expecting StatusBadRequest (400).
		{
			bucketName:         bucketName,
			bucketPolicyReader: bytes.NewReader(fmt.Appendf(nil, bucketPolicyTemplate, bucketName, bucketName)),

- Kube-log-runner: Added the `-log-file-size` parameter to rotate log output into a new file once it reached a certain size. Introduced `-log-file-age` to enable automatic removal of old output files, and `-flush-interval` to support periodic flushing. ([#127667](https://github.com/kubernetes/kubernetes/pull/127667), [@zylxjtu](ht...

      LockGraphNode node = lock.getLockGraphNode();
      // Iterate in reverse because locks are usually locked/unlocked in a
      // LIFO order.
      for (int i = acquiredLockList.size() - 1; i >= 0; i--) {
        if (acquiredLockList.get(i) == node) {
          acquiredLockList.remove(i);
          break;
        }
      }
    }
  }

    
- Kubeadm: improve the validation of serviceSubnet and podSubnet.
  ServiceSubnet has to be limited in size, due to implementation details, and the mask can not allocate more than 20 bits.
  PodSubnet validates against the corresponding cluster "--node-cidr-mask-size" of the kube-controller-manager, it fail if the values are not compatible.

* don't sync daemonsets with selectors that match all pods ([#23223](https://github.com/kubernetes/kubernetes/pull/23223), [@mikedanese](https://github.com/mikedanese))
* Trusty: Avoid reaching GCE custom metadata size limit ([#22818](https://github.com/kubernetes/kubernetes/pull/22818), [@andyzheng0831](https://github.com/andyzheng0831))

     *
     * @param <K> the key type of {@link IdentityHashMap}
     * @param <V> the value type of {@link IdentityHashMap}
     * @param expectedMaxSize the expected maximum size of the map
     * @return a new instance of {@link IdentityHashMap}
     * @see IdentityHashMap#IdentityHashMap(int)
     */
    public static <K, V> IdentityHashMap<K, V> newIdentityHashMap(final int expectedMaxSize) {