Laravel monitoring guides

Step-by-step tutorial: connect NightOwl Laravel monitoring to a free Supabase Postgres database. Project setup, connection string, agent install, retention tuning. ~15 minutes.

Step-by-step tutorial: connect NightOwl Laravel monitoring to a free Neon serverless Postgres. Project setup, direct connection string, agent install, retention tuning. ~15 minutes.

Step-by-step tutorial: connect NightOwl Laravel monitoring to a free Vercel Marketplace Postgres (powered by Neon). Provisioning, unpooled connection string, agent install. ~15 minutes.

Step-by-step tutorial: connect NightOwl Laravel monitoring to a free Aiven multi-cloud Postgres service. Always-on compute, AWS/GCP/Azure regions, agent install, retention tuning. ~15 minutes.

Step-by-step tutorial: send NightOwl alerts through your free Gmail account. App Password setup, SMTP configuration, test delivery, daily limits. ~10 minutes.

Step-by-step tutorial: send NightOwl alerts through Resend's free 3,000/month tier with branded sender. Domain verification, API key, sub-3-second delivery. ~10 minutes.

Step-by-step tutorial: send NightOwl alerts through Brevo's free 300/day tier. Sender validation, SMTP key, no domain required to start. ~10 minutes.

Step-by-step tutorial: send NightOwl alerts through SendGrid's free 100/day tier. Sender verification, API key, SMTP configuration, delivery tracking. ~10 minutes.

Reproduce locally with preventLazyLoading, fix with eager loading, and catch them in production with pattern grouping.

Database slow logs, Laravel's DB::whenQueryingForLongerThan hook, and APM-level pattern grouping with p95 trending.

CacheHit / CacheMissed events, hit-rate tracking, stampede prevention, and shared-cache store selection.

Nginx access logs, middleware timing, route aggregation with p95, and per-request trace drilldown.

Why average hides the slow tail, how to compute p95 exactly in PostgreSQL, and how to alert on it without noise.

Http facade events, Guzzle middleware, per-host p95 aggregation, timeouts, and correlation IDs.

memory_get_peak_usage, per-route aggregation, and the patterns that burn your memory_limit.

Persistent-worker memory drift, stale singletons, and crash tracking across FrankenPHP, Swoole, and RoadRunner.

Latency targets for JSON APIs, per-endpoint p95 tracking, and trace-based bottleneck hunting.

From the Log facade to fingerprinted exception tracking with user context, deploy SHA, and alert channels.

Worker supervision, failed_jobs handling, per-class duration percentiles, and alerts across Redis, SQS, database drivers.

Detect silent scheduler failures with dead-man's switches, scheduler events, onOneServer, and full invocation history.

MessageSent / MessageFailed events, queued mail worker health, delivery rate, and alert thresholds for mail failures.

failed_jobs table, JobFailed events, per-class failure rate trending, and alerts before users notice.

Per-channel delivery tracking across mail, Slack, SMS, database, and broadcast — with event listeners and aggregate dashboards.

Horizon is Redis-only. Options when your queue driver is database, SQS, Beanstalkd, or multi-driver.

Where each layer fails — PHP, PHP-FPM, Nginx, database — and how to tell them apart without just raising every timeout.

Signature verification, async processing, idempotency, and dead-man's-switch monitoring for incoming webhooks.

SQLSTATE codes, retry strategy with backoff, and patterns that make deadlocks stop happening.

Dispatch-to-start time as a per-queue SLO. Why Horizon's point-in-time view isn't enough.

Track 429s per route, identify misbehaving consumers, and use Redis for distributed rate limits.

Attach user identity to every request, exception, and job. Privacy tradeoffs and frontend-to-backend correlation.

Connect AI coding assistants to NightOwl via the Model Context Protocol — bearer tokens, six tools, full audit trail of AI-authored edits.

The five layers every production Laravel app needs to monitor, with tool picks and setup code for each.

A structured trace-based debugging workflow. Safe tools, and how to avoid ever debugging live.

Three credible patterns for self-hosted Laravel log aggregation — with Monolog config and honest ops tradeoffs.

What Forge provides, what it doesn't, and how to add APM as a Forge daemon.

Why persistent-agent APMs don't fit Lambda, cold-start tracking, and Vapor-compatible options.

Per-component performance, AJAX-storm avoidance, and latency targets that make reactive UIs feel reactive.

Table resource optimization, widget caching, auth failure tracking for admin panels.

Connection counts, memory leaks, broadcast failure rates, and why request-oriented APMs don't cover WebSockets.

What Cloud's built-in monitoring covers, what it doesn't, and how to layer an APM without fighting the platform.