omitted in short mode and when `externalDocs` is absent. ([#136988](https://github.com/kubernetes/kubernetes/pull/136988), [@pedjak](https://github.com/pedjak)) [SIG CLI]
- Kubectl: `kubectl describe node` now lists aggregated **ResourceSlices** when the `ResourceSlice` API is present, detailing slice name, driver, and pool. ([#131744](https://github.com/kubernetes/kubernetes/pull/131744), [@ArangoGutierrez](https://github.com/ArangoGutierrez)) [SIG CLI]

- K8s docker config json secrets are now compatible with docker config desktop authentication credentials files ([#82148](https://github.com/kubernetes/kubernetes/pull/82148), [@bbourbie](https://github.com/bbourbie))
- Kubelet and aggregated API servers now use v1 TokenReview and SubjectAccessReview endpoints to check authentication/authorization. ([#84768](https://github.com/kubernetes/kubernetes/pull/84768), [@liggitt](https://github.com/liggitt))

- The sample-apiserver aggregated conformance test has updated to use the Kubernetes v1.17.0 sample apiserver ([#84735](https://github.com/kubernetes/kubernetes/pull/84735), [@liggitt](https://github.com/liggitt)) [SIG API Machinery, Architecture, CLI and Testing]

    }

    @Override
    public CacheStats stats() {
      SimpleStatsCounter aggregator = new SimpleStatsCounter();
      aggregator.incrementBy(localCache.globalStatsCounter);
      for (Segment<K, V> segment : localCache.segments) {
        aggregator.incrementBy(segment.statsCounter);
      }
      return aggregator.snapshot();
    }

    @Override
    public void cleanUp() {

- The `RemoteRequestHeaderUID` feature moves to beta and is now enabled by default. This makes the kube-apiserver propagate UIDs in the `X-Remote-Uid` header in requests to the aggregated API servers. The header is not honored by default for incoming requests, but that can be enabled by setting the `--requestheader-uid-headers` flag explicitly. ([#130560](https://github.com/kubernetes/kubernetes/pull/130560), [@stlaz](https://github.com/stlaz))...

		m.Merge(&mLocal)

		mRemote := collectRemoteMetrics(ctx, madmin.MetricsBatchJobs, collectMetricsOpts{})
		m.Merge(&mRemote)

		if m.Aggregated.BatchJobs == nil {
			return metrics
		}

		for _, mj := range m.Aggregated.BatchJobs.Jobs {
			jtype := toSnake(mj.JobType)
			var objects, objectsFailed float64
			var bucket string
			switch madmin.BatchJobType(mj.JobType) {

- Added: Log Line for Debugging possible merge errors for kubelet related Config requests. ([#124389](https://github.com/kubernetes/kubernetes/pull/124389), [@holgerson97](https://github.com/holgerson97))
- Aggregated Discovery v2beta1 fixture is removed in `./api/discovery`. Please use v2 ([#127008](https://github.com/kubernetes/kubernetes/pull/127008), [@Jefftree](https://github.com/Jefftree)) [SIG API Machinery]

        this.executeLifecycle = executeLifecycle;
    }

    /**
     * @param aggregator <code>true</code> if the Mojo uses the Maven project and its child modules,
     * <code>false</code> otherwise.
     */
    public void setAggregator(boolean aggregator) {
        this.aggregator = aggregator;
    }

    /**
     * @return <code>true</code> if the Mojo uses the Maven project and its child modules,

    [FTRL paper](https://research.google.com/pubs/archive/41159.pdf).
*   `Optimizer.__init__` now accepts a `gradient_aggregator` to allow for
    customization of how gradients are aggregated across devices, as well as
    `gradients_transformers` to allow for custom gradient transformations (such
    as gradient clipping).
*   Improvements to Keras preprocessing layers:

		return e.aggregate.runningSum, err

	case aggFnMin:
		if !e.aggregate.seen {
			// No rows were seen by MIN
			return FromNull(), nil
		}
		return e.aggregate.runningMin, nil

	case aggFnMax:
		if !e.aggregate.seen {
			// No rows were seen by MAX
			return FromNull(), nil
		}
		return e.aggregate.runningMax, nil

	case aggFnSum:
		// TODO: check if returning 0 when no rows were seen