**replicate the containers** in the **available machines** taking into account the amount of memory needed by them, and the amount available in the machines in the cluster.

If your application is **simple**, this will probably **not be a problem**, and you might not need to specify hard memory limits. But if you are **using a lot of memory** (for example with **machine learning** models), you should check how much memory you are consuming and adjust the **number of containers** that runs...

        /*
         * We'd like to simply create our own TypeVariable with the newly resolved bounds. There's
         * just one problem: Starting with JDK 7u51, the JDK TypeVariable's equals() method doesn't
         * recognize instances of our TypeVariable implementation. This is a problem because users
         * compare TypeVariables from the JDK against TypeVariables returned by TypeResolver. To

```Python hl_lines="1  5-7" title="app/dependencies.py"
{!> ../../docs_src/bigger_applications/app_an/dependencies.py!}
```

////

//// tab | Python 3.8+ non-Annotated

/// tip

Prefer to use the `Annotated` version if possible.

///

```Python hl_lines="1  4-6" title="app/dependencies.py"
{!> ../../docs_src/bigger_applications/app/dependencies.py!}
```

////

/// tip

 *       running on it. This class, by contrast, does not expose an {@code Executor} API.)
 * </ul>
 *
 * <p>If you don't need the features of this class, you may prefer {@code newSequentialExecutor} for
 * its simplicity and ability to accommodate interruption.
 *
 * @since 26.0
 */
@ElementTypesAreNonnullByDefault
@J2ktIncompatible
public final class ExecutionSequencer {

		},
		// Test case - 10.
		// Case with proper components
		{
			bucketName: bucketName,
			objectName: "etc/path/proper/.../etc",
			byteRange:  "",
			accessKey:  credentials.AccessKey,
			secretKey:  credentials.SecretKey,

			expectedContent: encodeResponse(getAPIErrorResponse(ctx,
				getAPIError(ErrNoSuchKey),
				getGetObjectURL("", bucketName, "etc/path/proper/.../etc"),
				"", "")),

have both been migrated.

### gRPC Probes Graduate to Beta

With Kubernetes 1.24, the [gRPC probes functionality](https://github.com/kubernetes/enhancements/issues/2727)
has entered beta and is available by default. You can now [configure startup, liveness, and readiness probes](https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/#configure-probes) for your gRPC app

 *
 * <p>There are three primary interfaces provided to represent graphs. In order of increasing
 * complexity they are: {@link Graph}, {@link ValueGraph}, and {@link Network}. You should generally
 * prefer the simplest interface that satisfies your use case. See the <a
 * href="https://github.com/google/guava/wiki/GraphsExplained#choosing-the-right-graph-type">

 *
 * <p>There are three primary interfaces provided to represent graphs. In order of increasing
 * complexity they are: {@link Graph}, {@link ValueGraph}, and {@link Network}. You should generally
 * prefer the simplest interface that satisfies your use case. See the <a
 * href="https://github.com/google/guava/wiki/GraphsExplained#choosing-the-right-graph-type">

   * We get plan2 as a follow-up to plan1, typically retry the same IP but different TLS.
   * We get plan3 as a retry of plan0, which was canceled when it lost the race.
   *
   * This test confirms that we prefer to do the TLS follow-up (plan2) before the TCP retry (plan3).
   * It also confirms we enforce the 250 ms delay in each race.
   */
  @Test
  fun tcpConnectionsRaceAfterTlsFails() {
    val plan0 = routePlanner.addPlan()

 * guarantees detailed below.
 *
 * <p><b>Warning:</b> avoid <i>direct</i> usage of {@link ImmutableCollection} as a type (just as
 * with {@link Collection} itself). Prefer subtypes such as {@link ImmutableSet} or {@link
 * ImmutableList}, which have well-defined {@link #equals} semantics, thus avoiding a common source
 * of bugs and confusion.
 *