← Back to library
Cascades Practice PRO

Node health: healthcheck and fault tolerance

A balancer by itself doesn't guarantee uptime — it spreads traffic, but it poorly understands that a node is dead specifically for Russian clients. I'll show how I check health from a Russian point, how I mechanically pull a dead node out of rotation, and how I keep a hot reserve of entries so a block doesn't turn into downtime. The commands work.

This material is about engineering your own infrastructure and is educational in nature. Complying with the laws of your own jurisdiction is on you.

The observatory's blind spot

The main fault-tolerance trap: Xray's built-in observer (observatory/burstObservatory) pings pool members from the node itself. And a node can be alive to the world but cut off by TSPU specifically for Russian clients. From the node's point of view everything is fine — it sees generate_204 and…

/

This guide is part of paid access

Get full access to unlock the text, configs and builder.

Unlock access →