Moving on without a deceased storage node

[Kristian Jarmund]

A cluster consisting of 4 machines, 3 storage targets on each just had one of its machines die unexpectedly. No attempts at reviving it has helped, so we need to continue without it.

The good news:

- No metadata was running on the machine.

- It was the secondary in a buddy mirror group, so all data should still be intact.

The bad news:

- Production critical system

- Spare machine is still a few days away

So, the question is then really simple: How do I get the storage cluster up and running again without these storage targets available? Is there a way of "disabling" the storage targets in a buddy mirror and ignore the associated errors and warnings?