PrizmDoc® v14.3 Release - Updated January 14, 2025
PrizmDoc OCR API / OCR API Admin Guide / Deployment to Kubernetes - Guidance for the Process Manager / Deployment to Kubernetes - Guidance for the Reader Worker
Deployment to Kubernetes - Guidance for the Reader Worker

Resources Used in PrizmDoc Deployment

In addition to the core workload resources needed for a functioning installation, various additional capabilities can be supported through the instantiation of one or more process manager instances and supporting workers.

The OCR Reader Worker provides PrizmDoc with printed text recognition capabilities when paired with a Process Manager instance.

OCR Reader Worker

OCR Readers Workers can be created as a standard deployment. Supplying the string YES in the ACCEPT_EULA environment variable signifies that you accept the worker's license agreement.

You will also need to provide URLs where workers can locate the Process Manager's internal API and PrizmDoc Server endpoints.

Example

apiVersion: apps/v1
kind: Deployment
metadata:
  name: accusoft-ocr-reader-worker
spec:
  replicas: 1
  selector:
    matchLabels:
      app: accusoft-ocr-reader-worker
  template:
    metadata:
      labels:
        app: accusoft-ocr-reader-worker
    spec:
      containers:
      - name: ocr-reader-worker
        image: accusoft/ocr-reader-worker:<TBD: Version Number>
        env:
        - name: LOG_LEVEL
          value: "info"
        - name: PROCESS_MANAGER_INTERNAL_URL
          value: "http://accusoft-process-manager-internal:80"
        - name: WORKFILE_SERVICE_URL
          value: "http://prizmdoc-server.prizmdoc.svc:18681"
        - name: ACCEPT_EULA
          value: "YES"
        resources:
          limits:
            cpu: 1000m
            memory: 1000Mi
          requests:
            cpu: 1000m
            memory: 1000Mi
        livenessProbe:
          httpGet:
            path: /health
            port: 4000
          initialDelaySeconds: 1
          periodSeconds: 30
          timeoutSeconds: 5
        ports:
        - name: internal
          containerPort: 4000