)

// tagKey returns the tag-based key to use in the pkgCache.
// It is a comma-separated string; the first part is dir, the rest tags.
// The satisfied tags are derived from context but only those that
// matter (the ones listed in the tags argument plus GOOS and GOARCH) are used.
// The tags list, which came from go/build's Package.AllTags,
// is known to be sorted.
func tagKey(dir string, context *build.Context, tags []string) string {

In this case, if you had **multiple containers**, by default, when Prometheus came to **read the metrics**, it would get the ones for **a single container each time** (for the container that handled that particular request), instead of getting the **accumulated metrics** for all the replicated containers.

		HTTPStatusCode: http.StatusBadRequest,
	},
	ErrInvalidSSECustomerParameters: {
		Code:           "InvalidArgument",
		Description:    "The provided encryption parameters did not match the ones used originally.",
		HTTPStatusCode: http.StatusBadRequest,
	},
	ErrKMSNotConfigured: {
		Code:           "NotImplemented",
		Description:    "Server side encryption specified but KMS is not configured",

		anonPools = make([]string, len(poolsMatches))
		idxMap := map[int]string{
			1: "spfx",
			3: "ssfx",
		}
		for pi, poolsMatch := range poolsMatches {
			// Replace the server prefix/suffix with anonymized ones
			for idx, lbl := range idxMap {
				if len(poolsMatch[idx]) > 0 {
					poolsMatch[idx] = fmt.Sprintf("%s%d", lbl, crc32.ChecksumIEEE([]byte(poolsMatch[idx])))
				}
			}

- Added the implementation for `PodRecreationPolicy` to wait for the creation of pods once the existing ones are fully terminated. ([#117015](https://github.com/kubernetes/kubernetes/pull/117015), [@kannon92](https://github.com/kannon92))

    * These `dependencies` are run before the normal parameter dependencies. And normal dependencies are run too. They can be combined.
    * Dependencies declared in a router are executed first, then the ones defined in *path operation decorators*, and then the ones declared in normal parameters. They are all combined and executed.

- Introduced new metrics removing the redundant subsystem in kube-apiserver pod logs metrics and deprecate the original ones:
  - kube_apiserver_pod_logs_pods_logs_backend_tls_failure_total becomes kube_apiserver_pod_logs_backend_tls_failure_total