Spark/Hadoop workloads which use Hadoop MR Committer v1/v2 algorithm upload objects to a temporary prefix in a bucket. These objects are 'renamed' to a different prefix on Job commit. Object storage admins are forced to configure separate ILM policies to expire these objects and their versions to reclaim space.

### Solution

        try {
            processHelper.sendCommand("nonexistent", "test command");
            fail("Expected JobNotFoundException");
        } catch (JobNotFoundException e) {
            assertTrue(e.getMessage().contains("Job for nonexistent is not found"));
        }
    }

    public void test_startProcess_basicCommand() {
        String sessionId = "test_session";
        List<String> cmdList = Arrays.asList("echo", "hello");

        """.trimIndent()

    artifactRules = artifactRuleOverride ?: defaultArtifactRules
    paramsForBuildToolBuild(buildJvm, os, arch)
    params {
        // The promotion job doesn't have a branch, so %teamcity.build.branch% doesn't work.
        param("env.BUILD_BRANCH", "%teamcity.build.branch%")
    }

    vcs {
        root(AbsoluteId(VersionedSettingsBranch.fromDslContext().vcsRootId()))

			<version>${commons.fileupload.version}</version>
		</dependency>

		<!-- scheduler -->
		<dependency>
			<groupId>org.lastaflute.job</groupId>
			<artifactId>lasta-job</artifactId>
			<version>${lasta.job.version}</version>
		</dependency>

		<!-- mail -->
		<dependency>
			<groupId>org.dbflute.mail</groupId>
			<artifactId>mailflute</artifactId>

    public void sendCommand(final String sessionId, final String command) {
        final JobProcess jobProcess = runningProcessMap.get(sessionId);
        if (jobProcess == null) {
            throw new JobNotFoundException("Job for " + sessionId + " is not found.");
        }
        try {
            final OutputStream out = jobProcess.getProcess().getOutputStream();
            IOUtils.write(command + "\n", out, Constants.CHARSET_UTF_8);

### API Change

- Added field .status.uncountedTerminatedPods to the Job resource. This field is used by the job controller to keep track of finished pods before adding them to the Job status counters.
  
  Pods created by the job controller get the finalizer batch.kubernetes.io/job-tracking
  

Why I'm Leaving Flask - author: Mike Moritz author_link: https://medium.com/@mike.p.moritz link: https://medium.com/@mike.p.moritz/using-docker-compose-to-deploy-a-lightweight-python-rest-api-with-a-job-queue-37e6072a209b title: Using Docker Compose to deploy a lightweight Python REST API with a job queue - author: '@euri10' author_link: https://gitlab.com/euri10 link: https://gitlab.com/euri10/fastapi_cheatsheet title: A FastAPI and Swagger UI visual cheatsheet - author: Uber Engineering author_link:...

- Pods owned by a Job now uses the labels `batch.kubernetes.io/job-name` and `batch.kubernetes.io/controller-uid`.
  The legacy labels `job-name` and `controller-uid` are still added for compatibility. ([#114930](https://github.com/kubernetes/kubernetes/pull/114930), [@kannon92](https://github.com/kannon92))

        createFileConfig();
        logger.info("FileConfig is created");
        refresh();
        fileConfigId = getFileConfigIds(NAME_PREFIX).get(0);

        createJob();
        logger.info("Job is created");
        refresh();

        startJob(NAME_PREFIX);

        waitJob(NAME_PREFIX);
        refresh();
    }

    @BeforeEach
    protected void init() {
        refresh();
    }

- Improved performance of the job controller when handling job delete events. ([#127378](https://github.com/kubernetes/kubernetes/pull/127378), [@hakuna-matatah](https://github.com/hakuna-matatah))