UmbraDatastore/k8s/base/06-import-job.yaml

# Copyright 2026 Umbra Tools (umbratools.com)
# SPDX-License-Identifier: Apache-2.0

---
# Job to import existing productions data into PostgreSQL
# Run once after initial deployment: kubectl apply -f 06-import-job.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: umbratrans-import-productions
  namespace: umbratrans
  labels:
    app: umbratrans-import
    component: import
  annotations:
    # Allow job to run for a long time
    batch.kubernetes.io/job-tracking: "true"
spec:
  ttlSecondsAfterFinished: 86400  # Keep job for 24h after completion
  backoffLimit: 3
  activeDeadlineSeconds: 7200  # 2 hours max
  template:
    metadata:
      labels:
        app: umbratrans-import
        component: import
      annotations:
        prometheus.io/scrape: "false"
    spec:
      serviceAccountName: umbratrans-worker
      restartPolicy: OnFailure
      securityContext:
        runAsNonRoot: true
        runAsUser: 1000
        runAsGroup: 1000
        fsGroup: 1000
      initContainers:
        - name: wait-for-db
          image: postgres:16-alpine
          command: ["sh", "-c", "until pg_isready -h umbratrans-db-rw -U umbratrans -d umbratrans; do sleep 5; done"]
          env:
            - name: PGPASSWORD
              valueFrom:
                secretKeyRef:
                  name: umbratrans-db-credentials
                  key: password
      containers:
        - name: import
          image: umbratrans-worker:latest
          imagePullPolicy: IfNotPresent
          command: ["python", "-m", "umbratrans_web.cli", "import-productions"]
          args:
            - "--organization"
            - "umbra"  # Organization slug
            - "--confirm"
          envFrom:
            - secretRef:
                name: umbratrans-worker-config
            - secretRef:
                name: umbratrans-db-credentials
            - secretRef:
                name: umbratrans-secrets
          volumeMounts:
            - name: productions
              mountPath: /mnt/productions
              readOnly: true
            - name: artifacts
              mountPath: /mnt/artifacts
          resources:
            requests:
              memory: "512Mi"
              cpu: "500m"
            limits:
              memory: "2Gi"
              cpu: "2000m"
      volumes:
        - name: productions
          persistentVolumeClaim:
            claimName: umbratrans-productions
        - name: artifacts
          persistentVolumeClaim:
            claimName: umbratrans-artifacts

---
# CronJob for periodic re-import (optional - for catching new productions)
apiVersion: batch/v1
kind: CronJob
metadata:
  name: umbratrans-import-productions-schedule
  namespace: umbratrans
  labels:
    app: umbratrans-import
    component: import
spec:
  schedule: "0 3 * * *"  # Daily at 3 AM UTC
  concurrencyPolicy: Forbid
  successfulJobsHistoryLimit: 3
  failedJobsHistoryLimit: 3
  jobTemplate:
    spec:
      ttlSecondsAfterFinished: 86400
      backoffLimit: 1
      activeDeadlineSeconds: 7200
      template:
        spec:
          serviceAccountName: umbratrans-worker
          restartPolicy: OnFailure
          securityContext:
            runAsNonRoot: true
            runAsUser: 1000
            runAsGroup: 1000
            fsGroup: 1000
          containers:
            - name: import
              image: umbratrans-worker:latest
              imagePullPolicy: IfNotPresent
              command: ["python", "-m", "umbratrans_web.cli", "import-productions"]
              args:
                - "--organization"
                - "umbra"
                - "--incremental"  # Only import new/changed
              envFrom:
                - secretRef:
                    name: umbratrans-worker-config
                - secretRef:
                    name: umbratrans-db-credentials
                - secretRef:
                    name: umbratrans-secrets
              volumeMounts:
                - name: productions
                  mountPath: /mnt/productions
                  readOnly: true
                - name: artifacts
                  mountPath: /mnt/artifacts
              resources:
                requests:
                  memory: "512Mi"
                  cpu: "500m"
                limits:
                  memory: "2Gi"
                  cpu: "2000m"
          volumes:
            - name: productions
              persistentVolumeClaim:
                claimName: umbratrans-productions
            - name: artifacts
              persistentVolumeClaim:
                claimName: umbratrans-artifacts