# Step F: Fail over an active node
<a name="upg-red-fail"></a>

Fail over the running channels from the first active worker node that you're upgrading to a backup node. To reduce impact on your worker nodes, perform this step through [Step K: Re-designate the backup worker node](upg-red-redesignate.md) all for the same active worker node before moving on to the next node.

**To fail-over a node**

1. On the web interface for the primary Conductor Live node, access **Cluster** > **Redundancy**.

1. Select the worker node redundancy group which contains the node that you're failing over.

1. On the **Active Nodes** tab, locate the node that you're upgrading and choose the **Initiate Failover** button (double arrows).

   The backup node is moved to the **Active Nodes** tab and the running channels are moved from the failed worker.

1. When all channels are failed to a backup worker node, upgrade the failed active node as described in the next step.

**Warning**  
Your worker redundancy group will persist in an N\+0 redundancy type until you've upgraded all worker nodes. In a production scenario, this is *not a valid configuration*. An alert will persist until you are back to an N\+1, 1\+1, or N\+M redundancy type.