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>

k8s/scraper-worker.yaml

@@ -1,12 +1,12 @@
-# Dutchie AZ Worker Deployment
-# These workers poll the job queue and process crawl jobs.
-# Scale this deployment to increase crawl throughput.
+# 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 workers that poll and claim jobs (5 replicas)
-# - Workers use DB-level locking (FOR UPDATE SKIP LOCKED) to prevent double-crawls
-# - Each worker sends heartbeats; stale jobs are recovered automatically
+# - 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:
@@ -27,9 +27,9 @@ spec:
       containers:
         - name: worker
           image: code.cannabrands.app/creationshop/dispensary-scraper:latest
-          # Run the worker process instead of the main server
+          # Run the hydration worker in loop mode
           command: ["node"]
-          args: ["dist/dutchie-az/services/worker.js"]
+          args: ["dist/scripts/run-hydration.js", "--mode=payload", "--loop"]
           envFrom:
             - configMapRef:
                 name: scraper-config
@@ -57,9 +57,9 @@ spec:
               command:
                 - /bin/sh
                 - -c
-                - "pgrep -f 'worker.js' > /dev/null"
+                - "pgrep -f 'run-hydration' > /dev/null"
             initialDelaySeconds: 10
             periodSeconds: 30
             failureThreshold: 3
-      # Graceful shutdown - give workers time to complete current job
+      # Graceful shutdown - give workers time to complete current batch
       terminationGracePeriodSeconds: 60