Custom delete policy

Do you have a recurring job that triggers every 5 minutes and only does a small check to see whether something needs to be processed? 95% of the time there is probably nothing to do and the job succeeds immediately. But, all these recurring jobs fill up your JobRunr Dashboard (a huge amount of succeeded and deleted jobs) and also fill up your database.

JobRunr already allows you to configure the deletion policy for all jobs in the JobRunr configuration. By default, succeeded jobs will go to the deleted state after 36 hours and they will then be permanently deleted after 72 hours.

Thanks to the deleteOnSuccess and deleteOnFailure attributes on the @Job annotation you know have fine-grained control per Job on when it will move from the Succeeded state to the Deleted state and when the Job will be deleted permanently.

Usage

Using Java 8 lambdas

Specifying a custom delete policy for a job is easy thanks to the Job annotation. Just add it to your service method and specify when you want your job to be deleted.

@Job(deleteOnSuccess="PT10M!PT10M")
public void startImportingFilesIfPresent() {
    if(Files.list(importDirectory).findAny()) {
        BackgroundJob.enqueue(() -> fileImportService.import(Files.list(importDirectory).collect(toList())));
    }
}

Using JobRequest and JobRequestHandler

When you are using a JobRequest and a JobRequestHandler, just add the @Job annotation to the run method of the handler.

public class ImportFilesJobRequest implements JobRequest {

    public final String folder;

    public ImportFilesJobRequest(String folder) {
        this.folder = folder;
    }

    @Override
    public Class<? extends JobRequestHandler> getJobRequestHandler() {
        return ImportFilesJobRequestHandler.class;
    }
}

@Component
public class ImportJobRequestHandler implements JobRequestHandler<ImportFilesJobRequest> {

    @Job(deleteOnSuccess="PT10M!PT10M")
    public void run(ImportFilesJobRequest importFiles) {
        // business logic here
    }
}

Using the JobBuilder pattern

If you prefer the JobBuilder pattern, you can again pass the deleteOnSuccess and deleteOnFailure to the builder.

jobScheduler.create(aJob()
        .withDeleteOnSuccess(Duration.ofHours(1))
        .withDeleteOnFailure(Duration.ofHours(8))
        .withDetails(() -> System.out.println("This job will move to the deleted state after 1 hour if it succeeded and after 8 hours if it failed."));

The deleteOnSuccess and deleteOnFailure attribute accept the following string format: duration1!duration2:

• duration1 (optional): the duration in ISO8601 format before the job will move from Succeeded or Failed to Deleted
• ! (optional): a separator
• duration2 (optional): the duration in ISO8601 format before the job will be permanently deleted

Below is a table with some examples:

As you can specify both deleteOnSuccess and deleteOnFailure, this means you can keep failed jobs around longer (to see what went wrong). If you don’t specify deleteOnFailure, it falls back to JobRunr’s default which means the job will stay in the Failed state until manual action is taken.