spacedriveapp/spacedrive · error · anyhow::Error

Networking endpoint not available

Error message

Networking endpoint not available

What it means

networking_guard.endpoint() (core/src/service/network/core/mod.rs:947) returns the iroh Endpoint only when the networking service has created and retained one. None here means the service object exists but its endpoint was never bound or was torn down — partial startup, a shutdown race, or an endpoint-less configuration. The job stops before attempting any connection (strategy.rs:484-486).

Source

Thrown at core/src/ops/files/copy/strategy.rs:486

		let networking_guard = &*networking;

		// Resolve device slug to node_id for network routing
		let device_registry = networking_guard.device_registry();
		let registry = device_registry.read().await;
		let node_id = registry
			.get_node_by_device(source_device_id)
			.ok_or_else(|| {
				anyhow::anyhow!(
					"Could not find node_id for device {} (slug: {}). Device may be offline.",
					source_device_id,
					source_device_slug
				)
			})?;
		drop(registry);

		let endpoint = networking_guard
			.endpoint()
			.ok_or_else(|| anyhow::anyhow!("Networking endpoint not available"))?;

		ctx.log(format!(
			"Opening PULL connection to node {} (device {})",
			node_id, source_device_id
		));

		// Connect to remote device
		let node_addr = iroh::EndpointAddr::new(node_id);
		let connection = endpoint
			.connect(node_addr, b"spacedrive/filetransfer/1")
			.await
			.map_err(|e| anyhow::anyhow!("Failed to connect to device: {}", e))?;

		let (mut send_stream, mut recv_stream) = connection
			.open_bi()
			.await
			.map_err(|e| anyhow::anyhow!("Failed to open bidirectional stream: {}", e))?;

View on GitHub (pinned to 6dfeccf211)

Solutions

1. Restart the daemon so networking and its endpoint initialize before jobs run
2. Check startup logs for iroh endpoint binding failures
3. Cancel or re-queue copy jobs when the networking service shuts down
4. Verify the networking configuration enables the iroh endpoint

Example fix

// before
let endpoint = networking_guard.endpoint()
    .ok_or_else(|| anyhow::anyhow!("Networking endpoint not available"))?;

// after: fail with actionable context
let endpoint = networking_guard.endpoint().ok_or_else(|| anyhow::anyhow!(
    "iroh endpoint not bound; networking service is starting up or shutting down"
))?;

let ready = ctx
    .networking_service()
    .map(|n| n.endpoint().is_some())
    .unwrap_or(false);
if !ready {
    return Err(anyhow::anyhow!("iroh endpoint not bound; networking still starting or shutting down"));
}

When it happens

Trigger: Copy job starts during networking shutdown or before endpoint binding completes; endpoint creation failed earlier with the error swallowed; a test NetworkingService constructed without an endpoint.

Common situations: Daemon restart races where in-flight jobs outlive networking shutdown; config that disables the iroh endpoint; init errors logged but not propagated.

Related errors

• Could not find node_id for device {} (slug: {}). Device may
• Failed to connect to device: {}
• Failed to open bidirectional stream: {}
• Networking service not available
• File transfer protocol not registered

AI-assisted analysis of spacedriveapp/spacedrive@6dfeccf211 (2026-08-16). Data as JSON: /api/errors/08329ab6dba269aa. Report an issue: GitHub.