Observability Hub

Zero-config performance metrics, friendly latency percentiles, error tracking, and public webhook replay.

Monitoring how fast your app responds or finding broken endpoints shouldn't require setting up complex cloud monitoring tools or installing heavy npm packages.

The Observability Hub is a built-in health dashboard that watches traffic passing through your local tunnels and dev servers. It gives you instant, visual feedback on response times, error rates, and incoming webhooks—with zero setup.

Why Developers Use the Observability Hub

Zero-Config Setup: No SDKs to install, no configuration files to write, and no accounts to connect.
Spot Slow Endpoints: Identify which API calls take 20ms and which ones take 3 seconds because of slow database queries or external API calls.
Visual Error Tracking: See at a glance if your server is returning 404 Not Found or 500 Internal Server Error responses.
Webhook Stream & Replay: Inspect real incoming webhooks (from Stripe, Twilio, GitHub, etc.) and replay them with one click.

Understanding the Metrics (In Plain English)

You don't need a degree in statistics to understand your app's performance. Here is how Proxync breaks down your latency:

Metric	What It Means	Why It Matters
P50 (Median)	Half of all requests were faster than this number.	Represents the typical, everyday speed your users experience.
P90	90% of requests were faster than this number.	Shows how slightly slower or busier requests perform.
P99 (Worst-Case)	99% of requests were faster than this number.	Uncovers your worst-case outliers (like cold starts or heavy database scans).

Key Dashboard Panels

1. Status Code Health Gauge

A clean visual breakdown showing the percentage of:

2xx Success (Green): Healthy, successful responses.
3xx Redirects (Blue): URL rewrites or authentication redirects.
4xx Client Errors (Yellow): Missing routes (404) or bad payloads (400).
5xx Server Errors (Red): Crashes or uncaught exceptions in your backend code.

2. Slowest Routes Leaderboard

Ranks your endpoints by average and peak response times so you immediately know which functions to optimize.

3. Incoming Webhook Stream

When an external service sends an event to your public tunnel, it shows up here with timestamp, provider badge, and payload size. Click Replay to send the exact same event back to your dev server whenever you want to re-test your webhook logic.

4. Bandwidth Counter

Tracks total data transferred (KB and MB) across all active tunnels, helping you see if large images or heavy JSON responses are hogging bandwidth.

Telemetry Modes (Enhanced vs Low-CPU)

You can choose between two performance modes anytime under Settings:

Enhanced Telemetry (Default): Computes full latency percentiles ($P50, P90, P99$), route rankings, and bandwidth meters. Recommended for normal development laptops and desktops.
Basic Telemetry (Low-CPU Mode): Bypasses statistical sorting to save CPU and battery power on older laptops or lightweight virtual machines, tracking only critical 5xx errors.