Files
cannaiq/k8s/scraper-worker.yaml
Kelly 7f9cf559cf fix(k8s): Update worker deployment to use v2 hydration worker
The old dutchie-az/services/worker.js no longer exists. Workers now use
the hydration pipeline at dist/scripts/run-hydration.js with --loop mode.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-09 15:01:18 -07:00

66 lines
2.1 KiB
YAML

# Hydration Worker Deployment
# These workers process raw_payloads → canonical tables.
# Scale this deployment to increase hydration throughput.
#
# Architecture:
# - The main 'scraper' deployment runs the API server + scheduler (1 replica)
# - This 'scraper-worker' deployment runs hydration workers (5 replicas)
# - Workers use DB-level locking to prevent double-processing
# - Each worker processes payloads in batches with configurable limits
apiVersion: apps/v1
kind: Deployment
metadata:
name: scraper-worker
namespace: dispensary-scraper
spec:
replicas: 5
selector:
matchLabels:
app: scraper-worker
template:
metadata:
labels:
app: scraper-worker
spec:
imagePullSecrets:
- name: regcred
containers:
- name: worker
image: code.cannabrands.app/creationshop/dispensary-scraper:latest
# Run the hydration worker in loop mode
command: ["node"]
args: ["dist/scripts/run-hydration.js", "--mode=payload", "--loop"]
envFrom:
- configMapRef:
name: scraper-config
- secretRef:
name: scraper-secrets
env:
# Worker-specific environment variables
- name: WORKER_MODE
value: "true"
# Pod name becomes part of worker ID for debugging
- name: POD_NAME
valueFrom:
fieldRef:
fieldPath: metadata.name
resources:
requests:
memory: "256Mi"
cpu: "100m"
limits:
memory: "512Mi"
cpu: "500m"
# Health check - workers don't expose ports, but we can use a file check
livenessProbe:
exec:
command:
- /bin/sh
- -c
- "pgrep -f 'run-hydration' > /dev/null"
initialDelaySeconds: 10
periodSeconds: 30
failureThreshold: 3
# Graceful shutdown - give workers time to complete current batch
terminationGracePeriodSeconds: 60