*   `tf.distribute`:
    *   Support added for global sync `BatchNormalization` by using the newly
        added `tf.keras.layers.experimental.SyncBatchNormalization` layer. This
        layer will sync `BatchNormalization` statistics every step across all
        replicas taking part in sync training.
    *   Performance improvements for GPU multi-worker distributed training using

- GCE Internal LoadBalancer sync loop will now release the ILB IP address upon sync failure. An error in ILB forwarding rule creation will no longer leak IP addresses. ([#97740](https://github.com/kubernetes/kubernetes/pull/97740), [@prameshj](https://github.com/prameshj)) [SIG Cloud Provider and Network]...

- golang.org/x/crypto: 32db794 → 8634188
- golang.org/x/mod: v0.4.2 → 86c51ed
- golang.org/x/net: 491a49a → 1e63c2f
- golang.org/x/oauth2: 2bc19b1 → d3ed0bb
- golang.org/x/sync: 036812b → 886fb93
- golang.org/x/sys: f4d4317 → v0.3.0
- golang.org/x/term: 6886f2d → v0.3.0
- golang.org/x/text: v0.3.7 → v0.5.0
- golang.org/x/time: 1f47c86 → 90d013b
- golang.org/x/tools: d4cc65f → v0.1.12

- StatefulSet controllers no longer force a resync every 30 seconds when nothing has changed. ([#75622](https://github.com/kubernetes/kubernetes/pull/75622), [@jonsabo](https://github.com/jonsabo))

) {
	// used for storing the uploadID's parsed on concurrent HTTP requests for NewMultipart upload on the same object.
	testUploads := struct {
		sync.Mutex
		uploads []string
	}{}

	objectName := "test-object-new-multipart-parallel"
	var wg sync.WaitGroup
	for i := 0; i < 10; i++ {
		wg.Add(1)
		// Initiate NewMultipart upload on the same object 10 times concurrrently.
		go func() {

- Fixed bug in CPUManager with race on container map access ([#97427](https://github.com/kubernetes/kubernetes/pull/97427), [@klueska](https://github.com/klueska)) [SIG Node]

  your DRA Pods could be affected by this bug. ([#125643](https://github.com/kubernetes/kubernetes/pull/125643), [@sanposhiho](https://github.com/sanposhiho)) [SIG Scheduling and Testing]
- Fix endpoints status out-of-sync when the pod state changes rapidly ([#125675](https://github.com/kubernetes/kubernetes/pull/125675), [@tnqn](https://github.com/tnqn)) [SIG Apps, Network and Testing]

tchbook:before{content:"\f5c3"}.fa-swift:before{content:"\f8e1"}.fa-swimmer:before{content:"\f5c4"}.fa-swimming-pool:before{content:"\f5c5"}.fa-symfony:before{content:"\f83d"}.fa-synagogue:before{content:"\f69b"}.fa-sync:before{content:"\f021"}.fa-sync-alt:before{content:"\f2f1"}.fa-syringe:before{content:"\f48e"}.fa-table:before{content:"\f0ce"}.fa-table-tennis:before{content:"\f45d"}.fa-tablet:before{content:"\f10a"}.fa-tablet-alt:before{content:"\f3fa"}.fa-tablets:before{content:"\f490"}.fa-t...

package cmd

import (
	"bytes"
	"context"
	"encoding/xml"
	"fmt"
	"io"
	"math/rand"
	"net/http"
	"net/url"
	"reflect"
	"regexp"
	"runtime"
	"strings"
	"sync"
	"testing"
	"time"

	"github.com/dustin/go-humanize"
	jwtgo "github.com/golang-jwt/jwt/v4"
	"github.com/minio/minio-go/v7/pkg/set"
	xhttp "github.com/minio/minio/internal/http"
	"github.com/minio/pkg/v3/policy"

    * The vSphere cloud provider has a cleanup function with a loop,
    * inside of which a write lock was intended to prevent races against
    * VM creation, but the usage of the sync.RWMutex type is incorrect,
    * plus is placed in a defer call so the locking is effectively
    * nonexistent.  This patch adds an anonymous function call inside