Package org.apache.sling.event.jobs

Interface JobManager

  • @ProviderType
public interface JobManager
    The job manager is the heart of the job processing.

    The job manager allows to create new jobs, search for jobs and get statistics about the current state.

    The terminology used in the job manager is slightly different from common terminology: Each job has a topic and a topic is associated with a queue. Queues can be created through configuration and each queue can process one or more topics.

    Since:
    3.0

    • Method Detail

      • getStatistics

        Statistics getStatistics()
        Return statistics information about all queues.
        Returns:
        The statistics.

      • getTopicStatistics

        Iterable<TopicStatistics> getTopicStatistics()
        Return statistics information about job topics.
        Returns:
        The statistics for all topics.

      • getQueue

        Queue getQueue​(String name)
        Return a queue with a specific name (if running)
        Parameters:
        name - The queue name
        Returns:
        The queue or null

      • getQueues

        Iterable<Queue> getQueues()
        Return an iterator for all available queues.
        Returns:
        An iterator for all queues.

      • addJob

        Job addJob​(String topic,
           Map<String,​Object> properties)
        Add a new job If the topic is null or illegal, no job is created and null is returned. If properties are provided, all of them must be serializable. If there are non serializable objects in the properties, no job is created and null is returned. A job topic is a hierarchical name separated by dashes, each part has to start with a letter, allowed characters are letters, numbers and the underscore. The returned job object is a snapshot of the job state taken at the time of creation. Updates to the job state are not reflected and the client needs to get a new job object using the job id. If the queue for processing this job is configured to drop the job, null is returned as well.
        Parameters:
        topic - The required job topic.
        properties - Optional job properties. The properties must be serializable.
        Returns:
        The new job - or null if the job could not be created.
        Since:
        1.2

      • getJobById

        Job getJobById​(String jobId)
        Return a job based on the unique id. The returned job object is a snapshot of the job state taken at the time of the call. Updates to the job state are not reflected and the client needs to get a new job object using the job id.
        Parameters:
        jobId - The unique identifier from Job.getId()
        Returns:
        A job or null
        Since:
        1.2

      • removeJobById

        boolean removeJobById​(String jobId)
        Removes the job even if it is currently in processing. If the job exists and is not in processing, it gets removed from the processing queue. If the job exists and is in processing, it is removed from the persistence layer, however processing is not stopped.
        Parameters:
        jobId - The unique identifier from Job.getId()
        Returns:
        true if the job could be removed or does not exist anymore. false otherwise.
        Since:
        1.2

      • getJob

        Job getJob​(String topic,
           Map<String,​Object> template)
        Find a job - either queued or active. This method searches for a job with the given topic and filter properties. If more than one job matches, the first one found is returned which could be any of the matching jobs. The returned job object is a snapshot of the job state taken at the time of the call. Updates to the job state are not reflected and the client needs to get a new job object using the job id.
        Parameters:
        topic - Topic is required.
        template - The map acts like a template. The searched job must match the template (AND query).
        Returns:
        A job or null
        Since:
        1.2

      • findJobs

        Collection<Job> findJobs​(JobManager.QueryType type,
                         String topic,
                         long limit,
                         Map<String,​Object>... templates)
        Return all jobs of a given type. Based on the type parameter, either the history of jobs can be returned or unfinished jobs. The type parameter can further specify which category of jobs should be returned: for the history either succeeded jobs, cancelled jobs or both in combination can be returned. For unfinished jobs, either queued jobs, started jobs or the combination can be returned. If the history is returned, the result set is sorted in descending order, listening the newest entry first. For unfinished jobs, the result set is sorted in ascending order. The returned job objects are a snapshot of the jobs state taken at the time of the call. Updates to the job states are not reflected and the client needs to get new job objects.
        Parameters:
        type - Required parameter for the type. See above.
        topic - Topic can be used as a filter, if it is non-null, only jobs with this topic will be returned.
        limit - A positive number indicating the maximum number of jobs returned by the iterator. A value of zero or less indicates that all jobs should be returned.
        templates - A list of filter property maps. Each map acts like a template. The searched job must match the template (AND query). By providing several maps, different filters are possible (OR query).
        Returns:
        A collection of jobs - the collection might be empty.
        Since:
        1.2

      • stopJobById

        void stopJobById​(String jobId)
        Stop a job. When a job is stopped and the job consumer supports stopping the job processing, it is up to the job consumer how the stopping is handled. The job can be marked as finished successful, permanently failed or being retried.
        Parameters:
        jobId - The job id
        Since:
        1.3

      • retryJobById

        Job retryJobById​(String jobId)
        Retry a cancelled job. If a job has failed permanently it can be requeued with this method. The job will be removed from the history and put into the queue again. The new job will get a new job id. For all other jobs calling this method has no effect and it simply returns null.
        Parameters:
        jobId - The job id.
        Returns:
        If the job is requeued, the new job object otherwise null

      • createJob

        JobBuilder createJob​(String topic)
        Fluent API to create, start and schedule new jobs
        Parameters:
        topic - Required topic
        Returns:
        A job builder
        Since:
        1.3

      • getScheduledJobs

        Collection<ScheduledJobInfo> getScheduledJobs()
        Return all available job schedules.
        Returns:
        A collection of scheduled job infos
        Since:
        1.3

      • getScheduledJobs

        Collection<ScheduledJobInfo> getScheduledJobs​(String topic,
                                              long limit,
                                              Map<String,​Object>... templates)
        Return all matching available job schedules.
        Parameters:
        topic - Topic can be used as a filter, if it is non-null, only jobs with this topic will be returned.
        limit - A positive number indicating the maximum number of jobs returned by the iterator. A value of zero or less indicates that all jobs should be returned.
        templates - A list of filter property maps. Each map acts like a template. The searched job must match the template (AND query). By providing several maps, different filters are possible (OR query).
        Returns:
        All matching scheduled job infos.
        Since:
        1.4