### Container Images

All container images are available as manifest lists and support the described
architectures. It is also possible to pull a specific architecture directly by
adding the "-$ARCH" suffix  to the container image name.

name | architectures
---- | -------------

* Fix possible panic in PodAffinityChecker ([#33086](https://github.com/kubernetes/kubernetes/pull/33086), [@ivan4th](https://github.com/ivan4th))

min/disruptions/) feature in 1.4 (i.e. created `PodDisruptionBudget` objects), then **BEFORE**  upgrading from 1.4 to 1.5, you must delete all `PodDisruptionBudget` objects (`policy/v1alpha1/PodDisruptionBudget`) that you have created. It is not possible to delete these objects after you upgrade, and their presence will prevent you from using the beta PodDisruptionBudget feature in 1.5 (which uses `policy/v1beta1/PodDisruptionBudget`). If you have already upgraded, you will need to downgrade the...

	}
	// compare first to rest
	for _, str := range strs[1:] {
		xfixl := len(xfix)
		strl := len(str)
		// short-circuit empty strings
		if xfixl == 0 || strl == 0 {
			return ""
		}
		// maximum possible length
		maxl := xfixl
		if strl < maxl {
			maxl = strl
		}
		// compare letters
		if pre {
			// prefix, iterate left to right
			for i := 0; i < maxl; i++ {
				if xfix[i] != str[i] {

  // is a blocking operation for async eager; custom devices should avoid
  // retrieving shapes of tensors they wrap until the custom device tensor's
  // shape is explicitly requested where possible.
  int (*num_dims)(void* data, TF_Status* status);

  // Computes the axis length at `dim_index`.
  int64_t (*dim)(void* data, int dim_index, TF_Status* status);

  void (*deallocator)(void* data);

			if configRetriableErrors(err) {
				retryInterval := time.Duration(r.Float64() * float64(time.Second))
				logger.Info("Waiting for all MinIO IAM sub-system to be initialized.. possible cause (%v) (retrying in %s)", err, retryInterval)
				time.Sleep(retryInterval)
				continue
			}

          distinctElements++;
        }
        this.totalCount += count;
        return (right.height == initHeight) ? this : rebalance();
      }

      // adding count to me!  No rebalance possible.
      result[0] = elemCount;
      long resultCount = (long) elemCount + count;
      checkArgument(resultCount <= Integer.MAX_VALUE);
      this.elemCount += count;
      this.totalCount += count;

			return
		}
		for _, op := range ops {
			switch op {
			case madmin.CredentialsUpdateType:
				if !globalSiteReplicationSys.isEnabled() {
					// credentials update is possible only in bucket replication. User will never
					// know the site replicator creds.
					tgt.Credentials = target.Credentials
					tgt.TargetBucket = target.TargetBucket
					tgt.Secure = target.Secure

	}
}

// forceCompact will force compact the cache of the top entry.
// If the number of children is more than limit*100, it will compact self.
// When above the limit a cleanup will also be performed to remove any possible abandoned entries.
func (d *dataUsageCache) forceCompact(limit int) {
	if d == nil || len(d.Cache) <= limit {
		return
	}
	top := hashPath(d.Info.Name).Key()
	topE := d.find(top)
	if topE == nil {

* Update kube-dns to Version 1.14.8 that includes only small changes to how Prometheus metrics are collected. ([#57918](https://github.com/kubernetes/kubernetes/pull/57918), [@rramkumar1](https://github.com/rramkumar1))
* Fixes a possible deadlock preventing quota from being recalculated ([#58107](https://github.com/kubernetes/kubernetes/pull/58107), [@ironcladlou](https://github.com/ironcladlou))