Skip to main content

Create a Job

Deploy batch tasks and scheduled operations in minutes.

Prerequisites​

Before creating a job, ensure you have:

  • A project created in your organization
  • A Docker image ready to deploy
  • (Optional) A Docker registry configured for private images

Step-by-Step Guide​

1. Navigate to Jobs​

  1. Select your project from the project switcher
  2. Click Jobs in the left sidebar
  3. Click Create Job

2. Choose Job Type​

Select the type of job you want to create:

TypeDescriptionUse Case
JobRuns once immediatelyMigrations, one-time scripts, data imports
CronJobRuns on a scheduleBackups, reports, periodic data processing

3. Basic Configuration​

Job Name​

Enter a unique name for your job. This name is used for:

  • Identification in the jobs list
  • Log filtering
  • Execution tracking
Valid: data-processor, backup-job, nightly-report
Invalid: Data Processor, backup_job, nightly.report
Naming Convention

Use lowercase letters, numbers, and hyphens. Start with a letter.

Group (Optional)​

Enter an optional group name to visually bundle related jobs in the list. Jobs sharing a group are folded into a single collapsible row.

  • Free text, up to 63 characters.
  • Fixed at creation — cannot be changed later.

Docker Image​

Specify the Docker image containing your job script:

FormatExample
Docker Hubpython:3.12-slim
GitHub Registryghcr.io/org/my-job:v1.0
Custom Registryregistry.example.com/job:latest

For private registries, configure your registry credentials first.

Container Command (Optional)​

Override the command that runs in the container. Enter the complete command as you would type it in a terminal.

python /app/process.py --batch-size=1000
/bin/sh -c 'echo Starting && /app/run.sh'

Leave empty to use the image's default CMD/ENTRYPOINT.

4. Schedule Configuration (CronJob Only)​

If you selected CronJob, configure the schedule:

Cron Expression​

Use standard cron syntax: minute hour day-of-month month day-of-week

ExpressionDescription
* * * * *Every minute
*/5 * * * *Every 5 minutes
0 * * * *Every hour at minute 0
0 0 * * *Daily at midnight
0 9 * * 1-5Weekdays at 9:00 AM
0 0 1 * *First day of every month

Quick presets are available in the form for common schedules.

Timezone​

Select the timezone for schedule evaluation (default: UTC).

Timezone Selection

Use UTC for server-side jobs. Use local timezones for business-hour related tasks.

Concurrency Policy​

Control behavior when a new execution is due while a previous run is still active:

PolicyBehaviorBest For
AllowRun new job alongside existingIndependent tasks
ForbidSkip if previous still runningPrevent overlap
ReplaceStop running, start newAlways run latest

History Limits​

SettingDefaultDescription
Successful Jobs History3Completed executions to keep
Failed Jobs History1Failed executions to keep

5. Resources & Execution​

Compute Class​

Select the appropriate compute class for your job:

ClassCPU:RAM RatioMax CPUMax RAMBest For
Eco1:23 cores6 GBSimple scripts, testing
Standard1:415 cores58 GBGeneral batch jobs
Production1:46 cores28 GBCritical operations
Prod. Compute1:26 cores16 GBCPU-intensive processing
Prod. Memory1:86 cores58 GBData-heavy operations
ARM1:48 cores32 GBCost-optimized processing

Resource Allocation​

Configure resources using sliders:

  • CPU: Adjust in millicores (1000 mCPU = 1 core)
  • Memory: Automatically calculated based on CPU:RAM ratio
  • Ephemeral Storage: Temporary storage for job files (100 MiB - 10 GiB)
Resource Allocation

Memory is automatically adjusted when you change CPU to maintain the compute class ratio. A 15% tolerance is allowed.

Job Configuration​

SettingDefaultRangeDescription
Restart PolicyNeverOnFailure, NeverHow to handle container failures
Completions11-100Total successful runs needed
Parallelism11-100Maximum concurrent instances
Backoff Limit30-100Retry attempts before failure
TimeoutNone≥ 1 secondMaximum run duration before the execution is killed

Restart Policy Options:

  • OnFailure: Restart container on failure, retry up to backoff limit
  • Never: Don't restart, mark as failed immediately

6. Environment & Files​

Environment Variables​

Add configuration via environment variables:

  1. Click Add Variable
  2. Enter the key and value
  3. Toggle Secret for sensitive values
Example KeyExample Value
DATABASE_URLpostgresql://...
BATCH_SIZE1000
LOG_LEVELinfo

Security Context (Optional)​

Configure security restrictions:

FieldDescription
Run As User IDNumeric UID (e.g., 1000)
Run As Group IDNumeric GID (e.g., 1000)
Strong Isolation

All jobs benefit from strong isolation by default. Security context is only needed for compliance requirements.

7. Create the Job​

  1. Review your configuration in the summary
  2. Click Create
  3. For one-time jobs, execution starts immediately
  4. For CronJobs, the first execution occurs at the next scheduled time

Job Detail Tabs​

One-Time Jobs​

TabDescription
OverviewStatus, configuration summary, execution details
ConfigurationFull job configuration (image, resources, environment)
ObservabilityCPU, memory, and resource usage metrics
LogsReal-time and historical job output
EventsKubernetes events related to the job

CronJobs​

TabDescription
OverviewStatus, schedule, next/last execution, configuration summary
ConfigurationFull cron job configuration
ExecutionsRun history with status, duration, and logs for each execution
ObservabilityCPU, memory, and resource usage metrics
LogsReal-time and historical job output
EventsKubernetes events related to the cron job

After Creation​

Monitor Execution​

  1. View the job in the Jobs list
  2. Check the Executions tab for run history (CronJobs)
  3. View Logs for output
  4. Check Events for lifecycle events

For CronJobs​

  • Run Now: Trigger an immediate execution
  • Suspend: Pause scheduling
  • Resume: Restart scheduling

Example Configurations​

Database Backup (CronJob)​

ParameterValue
Namedb-backup
TypeCronJob
Imagepostgres:16-alpine
Schedule0 2 * * * (Daily at 2 AM)
TimezoneUTC
Compute ClassEco
CPU500 mCPU
Memory1024 MiB
ConcurrencyForbid

Command:

pg_dump -h $DB_HOST -U $DB_USER -d $DB_NAME | gzip > /backup/$(date +%Y%m%d).sql.gz

Data Processing Job​

ParameterValue
Namedata-processor
TypeJob
Imagepython:3.12-slim
Compute ClassStandard
CPU2000 mCPU
Memory8192 MiB
Completions10
Parallelism3
Backoff Limit2

Command:

python /app/process.py --job-index=$JOB_COMPLETION_INDEX

Report Generation (CronJob)​

ParameterValue
Nameweekly-report
TypeCronJob
Imagenode:20-alpine
Schedule0 8 * * 1 (Mondays at 8 AM)
TimezoneEurope/Paris
Compute ClassEco
CPU1000 mCPU
Memory2048 MiB
ConcurrencyReplace

Troubleshooting​

Job Never Starts​

  • Check the Events tab for scheduling issues
  • Verify image exists and is accessible
  • For CronJobs, verify schedule expression is valid

Job Fails Immediately​

  • Check Logs for error messages
  • Verify command syntax is correct
  • Ensure required environment variables are set

Job Keeps Restarting​

  • Backoff limit may be too high
  • Check if restart policy should be Never
  • Look for transient errors in logs

Image Pull Errors​

  • Verify image name and tag
  • Check registry credentials
  • Ensure image is built for correct architecture (AMD64 vs ARM)

Next Steps​