Job

Table of Contents

• Job: Running Your Tasks
• How to Create a Job
• Daestro Job Environment Variables
• How to Manage a Job
  • Cancel a Job
  • Re-submit a Job

Job: Running Your Tasks

A Job is a single, live instance of a Job Definition. If a Job Definition is the blueprint, then a Job is the actual house being built. Every time you want to run a task—like processing a file or running a report—you create a new Job.

Each Job is sent to a Job Queue, which then finds a suitable server (from a Compute Environment) to run it on.

---

How to Create a Job

You can start a new job at any time from the Daestro Console.

1. From the Dashboard, click on Add New Job.
2. Fill out the form with the details for this specific run:
   • Name (Optional): You can give this specific job a unique name. If you leave it blank, Daestro will generate a random one for you.
   • Job Definition (Required): Select the blueprint for the task you want to run from the list of definitions you’ve already created.
   • Job Queue (Required): Choose which queue this job should be sent to. The queue determines the job’s priority and what kind of server it will run on.
   • Command (Optional): You can override the command from the Job Definition for this specific run.
   • Command Parameters (Optional): If your command uses placeholders, you can provide or override their values here.
   • Environment Variables (Optional): You can add or override environment variables for this specific job.
   • Schedule At (Optional): If you want the job to run later, you can pick a specific date and time. If you leave this blank, the job will start immediately.

---

Daestro Job Environment Variables

In each job Daestro injects some Daestro specific environment variables which are listed below:

DAESTRO_JOB_ID: Id of the running job.

DAESTRO_JOB_RUN_ID: Job run id of currently running job.

DAESTRO_JOB_QUEUE_NAME: Name of the job queue to which job belongs.

DAESTRO_COMPUTE_SPAWN_ID: Id of the compute spawn where job is running.

DAESTRO_COMPUTE_ENV_NAME: Name of the compute environment which is used to spawn the compute.

DAESTRO_CLOUD_PROVIDER: Name of the cloud provider where it’s running.

DAESTRO_INSTANCE_TYPE: Instance type code within the cloud provider.

DAESTRO_PUBLIC_IPV4: Public IPv4 address of the server.

DAESTRO_PUBLIC_IPV6: Public IPv6 address of the server.

It is not guaranteed that all keys mentioned above will be present.

How to Manage a Job

After a job is created, you can track its status and manage it from the Jobs page.

Cancel a Job

You can cancel a job if it’s no longer needed. A job can be cancelled only if its status is queued (waiting to run) or running. Once a job has finished, failed, or been cancelled, it cannot be cancelled again.

Re-submit a Job

You can easily run a job again by re-submitting it. This is useful if a job failed and you want to retry it, or if you just need to run the exact same task again. You can re-submit any job as long as it is not currently in the queue or running.