README

A comprehensive observability toolkit for Laravel applications. Monitor HTTP requests, outbound API calls, queue jobs, scheduled tasks, and exceptions with structured logging and optional Prometheus or SidMonitor metrics.

Features

Inbound Request Logging - Automatically log all incoming HTTP requests
Outbound HTTP Logging - Log external API calls with service detection
Queue Job Logging - Track job execution with duration and memory usage
Scheduled Task Logging - Monitor artisan scheduled tasks lifecycle (start, finish, fail, skip)
Exception Logging - Structured exception logging with stack traces
Request ID Tracking - Correlation IDs for distributed tracing
Sensitive Data Masking - Automatic masking of passwords, tokens, and PII
Dual Exporter Support - Choose between Prometheus (pull) or SidMonitor (push) metrics
Circuit Breaker - Resilient push-based exporting with automatic backoff
Zero Configuration - Works out of the box with sensible defaults
Grafana Dashboards - Pre-built dashboard templates included

Requirements

PHP 8.0+
Laravel 9.0, 10.0, 11.0, or 12.0

Installation

composer require junixlabs/laravel-observatory

The package auto-registers and works immediately - no configuration needed!

Quick Start

After installation, Observatory automatically:

Logs all incoming HTTP requests to storage/logs/observatory.log
Logs outbound HTTP calls via Laravel's HTTP client
Logs queue job execution
Logs scheduled task execution
Logs exceptions with context

View Your Logs

tail -f storage/logs/observatory.log | jq

Configuration

Publish Config (Optional)

php artisan vendor:publish --tag=observatory-config

Environment Variables

All features are enabled by default. Only set these if you need to change defaults:

# Disable Observatory entirely
OBSERVATORY_ENABLED=false

# Change log channel (default: 'observatory' -> storage/logs/observatory.log)
# Use 'stderr' for Docker/K8s
OBSERVATORY_LOG_CHANNEL=stderr

# Disable specific loggers
OBSERVATORY_INBOUND_ENABLED=false
OBSERVATORY_OUTBOUND_ENABLED=false
OBSERVATORY_JOBS_ENABLED=false
OBSERVATORY_SCHEDULED_TASKS_ENABLED=false
OBSERVATORY_EXCEPTIONS_ENABLED=false

# Log request/response bodies (disabled by default - can be large)
OBSERVATORY_LOG_BODY=true

# Only log slow requests (0 = log all)
OBSERVATORY_SLOW_THRESHOLD_MS=1000

Exporters

Observatory supports two metrics exporters. Set via OBSERVATORY_EXPORTER env var.

Prometheus (default, pull-based)

OBSERVATORY_EXPORTER=prometheus
OBSERVATORY_PROMETHEUS_ENABLED=true
OBSERVATORY_PROMETHEUS_STORAGE=apcu  # or 'redis', 'memory'

Exposes a /metrics endpoint scraped by Prometheus.

SidMonitor (push-based)

OBSERVATORY_EXPORTER=sidmonitor
SIDMONITOR_ENDPOINT=https://api.sidmonitor.com
SIDMONITOR_API_KEY=your-api-key

Buffers data in-memory and flushes in batches to the SidMonitor backend. Includes a circuit breaker that pauses sending after consecutive failures to avoid blocking your application.

# Batch settings
SIDMONITOR_BATCH_SIZE=100
SIDMONITOR_BATCH_INTERVAL=10

# Circuit breaker
SIDMONITOR_CIRCUIT_BREAKER_THRESHOLD=3   # failures before opening
SIDMONITOR_CIRCUIT_BREAKER_COOLDOWN=30   # seconds before retry

Log Channel

Observatory auto-registers the observatory log channel:

File: storage/logs/observatory.log
Format: JSON (Loki/ELK compatible)
Rotation: Daily, 14 days retention

For Docker/Kubernetes, use stderr:

OBSERVATORY_LOG_CHANNEL=stderr

Custom Headers

Extract custom headers into logs (multi-tenant, workspace, etc.):

// config/observatory.php
'inbound' => [
    'custom_headers' => [
        'X-Workspace-Id' => 'workspace_id',
        'X-Tenant-Id' => 'tenant_id',
        'X-Correlation-Id' => 'correlation_id',
    ],
],

Result in logs:

{
  "request_id": "abc-123",
  "method": "POST",
  "path": "/api/users",
  "workspace_id": "ws-456",
  "tenant_id": "tenant-789"
}

Service Detection

Identify external services in outbound logs:

// config/observatory.php
'outbound' => [
    'services' => [
        '*.stripe.com' => 'stripe',
        '*.amazonaws.com' => 'aws',
        '*.sendgrid.com' => 'sendgrid',
        'api.myservice.com' => 'my_service',
    ],
],

Excluding Paths/Jobs/Tasks

// config/observatory.php
'inbound' => [
    'exclude_paths' => [
        'telescope*',
        'horizon*',
        'health',
        'metrics',
    ],
],

'jobs' => [
    'exclude_jobs' => [
        'App\Jobs\InternalHealthCheck',
    ],
],

'scheduled_tasks' => [
    'exclude_commands' => [
        'schedule:run',
    ],
],

Structured Log Examples

Inbound Request

{
  "message": "HTTP_REQUEST",
  "context": {
    "request_id": "550e8400-e29b-41d4-a716-446655440000",
    "type": "inbound",
    "method": "POST",
    "url": "https://example.com/api/v1/orders",
    "path": "api/v1/orders",
    "route": "orders.store",
    "status_code": 201,
    "duration_ms": 145.23,
    "ip": "192.168.1.1",
    "user_agent": "Mozilla/5.0...",
    "user_id": 123,
    "memory_mb": 45.2,
    "environment": "production"
  }
}

Outbound Request

{
  "message": "HTTP_OUTBOUND",
  "context": {
    "request_id": "550e8400-e29b-41d4-a716-446655440000",
    "type": "outbound",
    "service": "stripe",
    "method": "POST",
    "url": "https://api.stripe.com/v1/charges",
    "host": "api.stripe.com",
    "status_code": 200,
    "duration_ms": 523.45,
    "environment": "production"
  }
}

Job Processed

{
  "message": "JOB_PROCESSED",
  "context": {
    "job_id": "123",
    "job_name": "App\\Jobs\\ProcessOrder",
    "queue": "orders",
    "status": "processed",
    "duration_ms": 1234.56,
    "attempts": 1,
    "memory": {
      "used_mb": 12.5,
      "peak_mb": 45.2
    },
    "environment": "production"
  }
}

Scheduled Task

{
  "message": "SCHEDULED_TASK",
  "context": {
    "command": "reports:generate",
    "description": "Generate daily reports",
    "expression": "0 2 * * *",
    "status": "completed",
    "duration_ms": 4523.12,
    "exit_code": 0,
    "memory": {
      "used_mb": 32.1,
      "peak_mb": 64.5
    },
    "environment": "production"
  }
}

Exception

{
  "message": "EXCEPTION",
  "context": {
    "request_id": "550e8400-e29b-41d4-a716-446655440000",
    "exception_class": "App\\Exceptions\\PaymentException",
    "message": "Payment declined",
    "code": 402,
    "file": "/app/Services/PaymentService.php",
    "line": 145,
    "request": {
      "method": "POST",
      "url": "https://example.com/api/orders",
      "path": "api/orders"
    },
    "user": {
      "id": 123
    },
    "trace": ["..."],
    "environment": "production"
  }
}

Prometheus Metrics (Optional)

OBSERVATORY_PROMETHEUS_ENABLED=true
OBSERVATORY_PROMETHEUS_STORAGE=apcu  # or 'redis', 'memory'

Visit http://your-app.test/metrics to see metrics.

Available Metrics

Metric	Type	Description
`{app}_http_requests_total`	Counter	Total HTTP requests
`{app}_http_request_duration_seconds`	Histogram	Request latency
`{app}_http_outbound_requests_total`	Counter	Outbound HTTP requests
`{app}_jobs_processed_total`	Counter	Queue jobs processed
`{app}_scheduled_tasks_total`	Counter	Scheduled tasks executed
`{app}_exceptions_total`	Counter	Exceptions count

Prometheus Auth

OBSERVATORY_METRICS_AUTH=true
OBSERVATORY_METRICS_USER=prometheus
OBSERVATORY_METRICS_PASS=secret

Grafana Dashboards

Import pre-built dashboards from the dashboards/ directory:

Dashboard	Data Source	Description
`observatory-dashboard.json`	Loki	Request logs, user analytics, exceptions
`prometheus-dashboard.json`	Prometheus	Metrics overview, latency percentiles

LogQL Query Examples

# All requests
{job="laravel-observatory"} | json | message="HTTP_REQUEST"

# Errors only
{job="laravel-observatory"} | json | status_code >= 400

# By user
{job="laravel-observatory"} | json | user_id="123"

# Slow requests (>1s)
{job="laravel-observatory"} | json | duration_ms > 1000

# External service calls
{job="laravel-observatory"} | json | type="outbound" | service="stripe"

# Scheduled tasks
{job="laravel-observatory"} | json | message="SCHEDULED_TASK"

# Exceptions
{job="laravel-observatory"} | json | message="EXCEPTION"

Kubernetes Deployment

See k8s/README.md for Kubernetes deployment with Loki stack.

Upgrading from v1.x

Breaking Changes in v2.0

SidMonitor env vars renamed: OBSERVATORY_SIDMONITOR_* is now SIDMONITOR_*
ExporterInterface: New recordScheduledTask() method required for custom exporters

Update your .env file if using SidMonitor:

# Before (v1.x)
OBSERVATORY_SIDMONITOR_ENDPOINT=...
OBSERVATORY_SIDMONITOR_API_KEY=...

# After (v2.0)
SIDMONITOR_ENDPOINT=...
SIDMONITOR_API_KEY=...

Re-publish config if upgrading:

php artisan vendor:publish --tag=observatory-config --force

Testing

composer test

Changelog

See CHANGELOG for recent changes.

Contributing

See CONTRIBUTING for details.

Security

Report security issues to chuongld@canawan.com instead of the issue tracker.

Credits

License

MIT License. See LICENSE for details.

junixlabs/laravel-observatory

包简介

关键字：

README 文档