When a node in a cluster fails, strategies such as "fencing" may be employed to keep the rest of the system operational.

Once the node failed, the walk of the tunnel would soon have gone liquid.

If many nodes fail at once, the great forces that create them will ultimately have to readjust.

'What does it matter if one or two nodes fail?'

Also watch what happens when a node fails and how the nodes discover an alternate routing path.

When a node fails, the data will simply take another route.

That compute nodes will fail during execution of job, and so will other hardware components.

This allows faster error recovery, in case a neighboring node fails.

This allows entire nodes to fail without losing access to any data.

In the event the first node fails the second one steps in.