Utilizzo RegisterJobDefinition con un AWS SDK o una CLI

Per registrare una definizione di lavoro

Questo esempio registra una definizione di processo per un semplice processo contenitore.

Comando:

aws batch register-job-definition --job-definition-name sleep30 --type container --container-properties '{ "image": "busybox", "vcpus": 1, "memory": 128, "command": [ "sleep", "30"]}'

Output:

{
    "jobDefinitionArn": "arn:aws:batch:us-east-1:012345678910:job-definition/sleep30:1",
    "jobDefinitionName": "sleep30",
    "revision": 1
}

    /**
     * Registers a new job definition asynchronously in AWS Batch.
     * <p>
     * When using Fargate as the compute environment, it is crucial to set the
     * {@link NetworkConfiguration} with {@link AssignPublicIp#ENABLED} to
     * ensure proper networking configuration for the Fargate tasks. This
     * allows the tasks to communicate with external services, access the
     * internet, or communicate within a VPC.
     *
     * @param jobDefinitionName the name of the job definition to be registered
     * @param executionRoleARN the ARN (Amazon Resource Name) of the execution role
     *                         that provides permissions for the containers in the job
     * @param cpuArch a value of either X86_64 or ARM64 required for the service call
     * @return a CompletableFuture that completes with the ARN of the registered
     *         job definition upon successful execution, or completes exceptionally with
     *         an error if the registration fails
     */
    public CompletableFuture<String> registerJobDefinitionAsync(String jobDefinitionName, String executionRoleARN, String image, String cpuArch) {
        NetworkConfiguration networkConfiguration = NetworkConfiguration.builder()
                .assignPublicIp(AssignPublicIp.ENABLED)
                .build();

        ContainerProperties containerProperties = ContainerProperties.builder()
                .image(image)
                .executionRoleArn(executionRoleARN)
                .resourceRequirements(
                        Arrays.asList(
                                ResourceRequirement.builder()
                                        .type(ResourceType.VCPU)
                                        .value("1")
                                        .build(),
                                ResourceRequirement.builder()
                                        .type(ResourceType.MEMORY)
                                        .value("2048")
                                        .build()
                        )
                )
                .networkConfiguration(networkConfiguration)
               .runtimePlatform(b -> b
                        .cpuArchitecture(cpuArch)
                        .operatingSystemFamily("LINUX"))
                .build();

        RegisterJobDefinitionRequest request = RegisterJobDefinitionRequest.builder()
                .jobDefinitionName(jobDefinitionName)
                .type(JobDefinitionType.CONTAINER)
                .containerProperties(containerProperties)
                .platformCapabilities(PlatformCapability.FARGATE)
                .build();

        CompletableFuture<String> future = new CompletableFuture<>();
        getAsyncClient().registerJobDefinition(request)
                .thenApply(RegisterJobDefinitionResponse::jobDefinitionArn)
                .whenComplete((result, ex) -> {
                    if (ex != null) {
                        future.completeExceptionally(ex);
                    } else {
                        future.complete(result);
                    }
                });

        return future;
    }