What It Is

XProxy Tester is an automated proxy observability service that syncs devices from the XProxy Portal, runs continuous health checks through each proxy, and records uptime, stability, rotation behavior, speed, and errors into MySQL for Grafana dashboards and operational APIs.

Operators get a live picture of which proxies are healthy, which are failing rotation or stability thresholds, and when the system should auto-deactivate or recover endpoints without manual spot checks across a large fleet.

The Problem We Solved

Proxy fleets at scale are hard to trust without constant measurement:

A proxy can appear “up” while failing rotation or returning the same egress IP
Manual testing does not keep pace with hundreds of rotating endpoints
Raw request logs grow quickly and slow down analytics without aggregation and archival
Incidents are discovered late when there is no single dashboard for stability and errors

XProxy Tester closes that gap with scheduled tests, rotation verification, stability classes, and fleet-wide metrics designed for production operations.

What We Work On

Continuous testing

Per-proxy test loops against configurable IP-check targets, recording success, latency, and failure reasons.

IP rotation

Rotation commands, cycle management, and verification that outbound IPs actually change when expected.

Stability & reliability

Stable / UnstableHourly / UnstableDaily classification, auto-deactivation on repeated failures, and auto-recovery when health returns.

Performance telemetry

Periodic speed tests (download/upload) and response-time tracking for capacity planning.

Data pipeline

BullMQ write queues, batch writers, MySQL partitioning, daily and 5-minute aggregations, and archival of raw request history.

Operations surface

HTTP health (/health, /ready, /live), Prometheus /metrics, problems analytics API, and a catalog of Grafana dashboards (overview, rotation, errors, duplicate IP, and more).

How It Works (In Simple Terms)

Sync: Devices and proxies are pulled from the XProxy Portal API.
Test: Each proxy is exercised on a short interval through its tunnel to an external check URL.
Rotate & verify: Rotation jobs run where configured; the service confirms the egress IP changed.
Score: Stability rules and aggregates classify fleet health over time.
Act & visualize: Failing proxies can be deactivated automatically; Grafana and APIs expose trends and incidents.

Heavy writes are queued and batched so testing loops stay responsive under high request volume.

Key Outcomes

Always-on visibility: Continuous metrics instead of ad hoc proxy checks.
Rotation confidence: Detect same-IP and no-rotation conditions before they hit users.
Safer fleet operations: Auto-deactivation and recovery reduce manual firefighting.
Scalable telemetry: Queues, partitioning, and rollups keep MySQL usable at volume.
Operator-ready dashboards: Grafana views aligned to stability, performance, and errors.

Technologies & Approaches We Used

Area	What we used	Why it matters
Runtime	Node.js 20+, TypeScript	Typed services for testing and API control
Database	MySQL 8, Prisma	Structured proxy, request, rotation, and summary models
Queues	BullMQ, Redis, ioredis	Async persistence under burst test traffic
Scheduling	node-cron, internal timers	Device sync, aggregations, archival, speed tests
HTTP	Native Node HTTP server, axios, undici	Lightweight control plane and outbound checks
Observability	Grafana, Prometheus, pino	Dashboards, metrics, and structured logs
Deploy	Docker, Docker Compose, Vitest	Containerized stack and automated tests

Approach in practice: We separated hot test loops from cold storage writes tests enqueue results, batch writers land them in partitioned tables, and aggregations feed Grafana. Stability and rotation are first-class domains, not afterthought columns on a uptime ping.

Who It's For

Teams operating the XProxy Portal proxy infrastructure
SREs and operators monitoring rotation, stability, and incident trends
Engineering groups building reliability automation around third-party proxy fleets