Receipt API Integration

Real-time monitoring of the Fujitsu Receipt API (Azure App Service) that processes invoicing for store purchases. Deployed on fujitsu-pt-gdc-dx-integration.com domain with separate dev and production environments.

System Overview

Azure App Service

Receipt API DEV

dev-elior-stmicro-ba

Receipt API PROD

prd-elior-stmicro-ba

→

GET /hw

CF Worker cron-stats

Health Polling

*/15 * * * *

→

INSERT

Control Center

D1 Database

receipt_api_healthreceipt_api_requests

Operations Dashboard

Real-time Metrics (PROD only)

Receipt API PROD

→

MetricsService.Log()

→

POST /api/receipt-api/logs

→

D1 Database

Azure (DEV + PROD) → Cron Worker → D1 → Dashboard | PROD also sends real-time metrics via MetricsService.Log()

Endpoints

Environment	URL	Purpose	Polled By
DEV	dev-elior-stmicro-ba.fujitsu-pt-gdc-dx-integration.com	Development/testing	Cron Worker (every 15min)
PROD	prd-elior-stmicro-ba.fujitsu-pt-gdc-dx-integration.com	Production (live store)	Cron Worker (every 15min)

Data Flow: Health Monitoring

Note

Polling Frequency: Every 15 minutes (4x per hour) Rate Limit: Azure allows 100 req/min → we use 0.13 req/min (0.13% utilization) Endpoint: GET /hw Response: { status, environment, buildVersion, timestamp }

Cron Trigger

*/15 * * * *

→

Fetch GET /hw

→

Parse response

→

INSERT into

receipt_api_health

→

Dashboard

Shows status

Data Flow: Request Metrics

Note

Trigger: Real-time (after each receipt processing) Endpoint: POST /api/receipt-api/logs Authentication: X-Api-Key: qC1PaRuwi5Op18l08vIMJ6YZFCS7D1qvLp5hIa7eIyo= Payload: JSON with success, latency, errors, cart details

User exits store

→

Receipt API

Processes receipt

→

MetricsService.Log()

→

POST

/api/receipt-api/logs

→

Stored in D1

→

Dashboard analytics

Database Schema

receipt_api_health

Column	Type	Description
id	INTEGER	Primary key
online	INTEGER	1 = up, 0 = down
environment	TEXT	”Development” or “Production”
build_version	TEXT	Assembly version from /hw
response_time_ms	REAL	Latency in milliseconds
error	TEXT	Error message if offline
created_at	DATETIME	Snapshot timestamp

receipt_api_requests

Column	Type	Description
id	INTEGER	Primary key
user_token	TEXT	User identifier
cart_id	TEXT	Zippin cart session ID
success	INTEGER	1 = success, 0 = failure
http_status	INTEGER	HTTP response code
latency_total_ms	REAL	Total processing time
latency_jes_ms	REAL	JES API call time only
error_type	TEXT	jes_api_error, network_error, missing_sku, etc.
error_code	INTEGER	JES error code (e.g., 60110)
error_message	TEXT	Detailed error message
payload_items_count	INTEGER	Number of items in cart
created_at	DATETIME	Request timestamp

Dashboard Metrics

The Operations page displays real-time analytics for the Receipt API:

• Health Status: Online/Offline badge for dev and prod environments
• Success Rate: Percentage of successful receipts (last 24h)
• Average Latency: Mean response time in milliseconds
• Error Breakdown: Count by error type (JES API errors, network errors, missing SKUs, conversion errors)
• Recent Failures: Table showing last 50 failed requests with cart ID, error type, and message

Error Types Reference

Error Type	Description	Example Scenario
jes_api_error	JES API returned an error	Receipt rejected by JES, validation fails
network_error	Cannot reach JES API	Timeout, DNS failure, connection refused
missing_sku	Product SKU not found	Unknown item scanned by Zippin system
conversion_error	Data transformation failed	Invalid price format, null reference exception
invalid_payload	Request validation failed	Missing required fields, malformed JSON

Implementation Details

Note

Azure Side (.NET 10):

• HealthController.cs → GET /hw endpoint
• MetricsService.cs → POST metrics to Control Center
• ReceiptController.cs → Calls MetricsService after each receipt

Control Center Side:

• workers/cron-stats/index.js → Polls /hw every 15min
• functions/api/receipt-api/logs.js → Receives POST metrics
• functions/api/receipt-api/stats.js → Aggregates analytics
• js/pages/home.js → Dashboard UI functions

Monitoring Best Practices

• Health Checks: 15-minute interval prevents rate limiting while providing timely alerts
• Real-time Push: Receipt API sends metrics immediately after processing for instant visibility
• No Local Storage: Azure App Service remains stateless, all data stored in Control Center D1
• Error Classification: Standardized error types enable trend analysis and root cause identification
• Dual Environment: Separate dev/prod monitoring prevents test data from polluting production metrics