Skip to main content

JobScheduler

Handles tasks lifecycle. Extends @perseid/server Engine class.

instanceId​

Job scheduler instance unique id.

protected instanceId: Id;

logsPath​

Path to the tasks logs directory.

protected logsPath: string;

bucketClient​

Bucket client.

protected bucketClient: BucketClient;

availableSlots​

Amount of initially available slots for that scheduler to run jobs.

protected availableSlots: number;

jobs​

List of registered jobs.

protected jobs: Record<string, JobScript>;

tasksRegistry​

Running tasks registry.

protected tasksRegistry: Record<string, {
  worker: Worker;
  _status: DataModel['tasks']['_status'];
}>;```

---

## constructor

Class constructor.

```typescript
public constructor(
  logger: Logger,
  databaseClient: DatabaseClient,
  bucketClient: BucketClient,
  settings: JobSchedulerSettings,
);

Parameters​

  • logger: Logger to use.
  • databaseClient: Database client to use to store/fetch data.
  • bucketClient: Bucket client to use to store logs files.
  • settings: Jobs scheduler settings.

Usage​

const jobScheduler = new JobScheduler(logger, databaseClient, bucketClient, {
  availableSlots: 10,
  logsPath: '/tmp/logs',
  jobs: {
    testJob: async (taskId, metaData, log): Promise<void> => {
      await Promise.resolve();
      log.info(`Hello from ${String(taskId)}!`);
      log.info(metaData.lastCompletedAt);
    },
  },
})

withAutomaticFields​

Returns updated payload with automatic fields.

protected withAutomaticFields<Resource extends keyof DataModel>(
  resource: Resource,
  existingResource: DataModel[Resource] | null,
  payload: Payload<DataModel[Resource]>,
  context: CommandContext<DataModel>,
): Promise<Payload<DataModel[Resource]>>;

Parameters​

  • resource: Type of resource for which to generate automatic fields.
  • existingResource: Existing resource being updated, if applicable, null otherwise.
  • payload: Payload to update.
  • context: Command context.

Returns​

Payload with automatic fields.

Usage​

await jobScheduler.withAutomaticFields(
  'tasks',
  { _id: new Id(), _status: 'IN_PROGRESS' },
  { _status: 'COMPLETED' },
  {}
);

reSchedulePeriodicTask​

Schedules next execution for the given task, if it is periodic.

protected reSchedulePeriodicTask(task: DataModel['tasks'], taskCompleted: boolean): Promise<void>;

Parameters​

  • task: Task to re-schedule.
  • taskCompleted: Whether task successfully completed.

Usage​

await jobScheduler.reSchedulePeriodicTask({ _id: new Id(), _status: 'IN_PROGRESS' }, false);

uploadLogs​

Uploads logs file of task to a persistent storage.

protected uploadLogs(task: DataModel['tasks']): Promise<void>;

Parameters​

  • task: Task for which to upload logs file.

Usage​

await jobScheduler.uploadLogs({ _id: new Id(), _status: 'IN_PROGRESS' });

executeTask​

Executes task.

protected executeTask(task: DataModel['tasks']): Promise<void>;

Parameters​

  • task: Task to execute.

Usage​

await jobScheduler.executeTask({ _id: new Id(), _status: 'PENDING' });

closeTask​

Closes task, performing all post-processing operations.

protected closeTask(
  task: DataModel['tasks'],
  status: DataModel['tasks']['_status'],
): Promise<void>;

Parameters​

  • task: Task to close.
  • status: Status to update task with.

Usage​

await jobScheduler.closeTask({ _id: new Id(), _status: 'IN_PROGRESS' }, 'COMPLETED');

processPendingTasks​

Processes candidate pending tasks.

protected processPendingTasks(): Promise<void>;

Usage​

await jobScheduler.processPendingTasks();

processRunningTasks​

Processes tasks in progress.

protected processRunningTasks(): Promise<void>;

Usage​

await jobScheduler.processRunningTasks();

runJob​

Runs the job passed as a command line argument to the script. This method is meant to be called in its own dedicated script, and should not be mixed up with run.

public static runJob(
  jobs: JobSchedulerSettings['jobs'],
  logsPath: string,
  logLevel: 'debug' | 'info' | 'warn' | 'error' | 'fatal',
): Promise<void>;

Parameters​

  • jobs: List of jobs to register.
  • logsPath: Path to the tasks logs directory.
  • logLevel: Minimum logging level (all logs below that level won't be logs).

Usage​

await JobScheduler.runJob({
  testJob: async (taskId, metaData, log): Promise<void> => {
    await Promise.resolve();
    log.info(`Hello from ${String(taskId)}!`);
    log.info(metaData.lastCompletedAt);
  },
}, '/tmp/logs', 'debug');

run​

Executes job scheduler.

public run(): Promise<void>;

Usage​

await jobScheduler.run();