What Happens When a Tor Entry Guard Goes Offline?

When your assigned Tor entry guard goes offline, your Tor client automatically detects the failure and seamlessly switches to another guard node from its internal list to maintain your connection. Tor uses persistent entry guards to shield your traffic against network correlation attacks, and its guard selection algorithm is specifically designed to handle node outages gracefully. This ensures that a temporary downtime or permanent departure of your primary guard node will not break your anonymity or require manual intervention to restore your connection.

How Tor Detects an Offline Guard

When you launch the Tor Browser or the Tor daemon, the client attempts to establish a three-hop circuit starting with your primary entry guard. If the guard is offline, the connection will fail to complete the initial TLS handshake, or the connection will time out.

Tor’s consensus mechanism continuously updates every hour, broadcasting the current status of all relays on the network. If the directory authorities mark your guard as offline or drop its “Guard” flag, your client notes this status change during its next consensus update.

The Fallback Mechanism: How a New Guard Is Selected

Tor does not randomly choose a brand-new relay the instant your primary guard fails, as frequent switching would increase vulnerability to path-bias and traffic-analysis attacks. Instead, it follows a structured process:

  1. Checking Secondary Guards: Tor maintains a small, persistent set of sampled guards (stored in your Tor state file). If the primary guard is unreachable, Tor temporarily falls back to the next eligible guard already present in your sampled list.
  2. Confirming Failure: The client marks the unreachable guard as temporarily down rather than immediately discarding it, attempting to reconnect periodically in case the outage is short-lived.
  3. Promoting a New Primary Guard: If the original guard remains unreachable for an extended period or is permanently removed from the public consensus, Tor promotes a secondary guard to primary status.
  4. Sampling New Relays: If your entire active guard list becomes unavailable or exhausts its validity period, Tor samples a fresh set of reliable relays that hold the “Guard” and “Fast” flags from the active consensus.

Impact on User Experience

Security and Anonymity Protection

Tor’s guard management algorithm (known as Guardfraction and Guard Selection Specification) prevents adversaries from forcing you to switch guards repeatedly until you land on a malicious node. Because your client strictly cycles through its pre-sampled guard list rather than picking blindly at random during an outage, your exposure to malicious entry relays remains strictly bounded.