[Gluster-users] Recommended brick replacement methods?

[B.K.Raghuram]

Hi,

Given the recent threads about remove-brick vs replace brick, I was
wondering if someone could throw some light on the best ways to handle
each of the following situations where each brick/volume could contain
large amounts of data:  In all these cases, the new brick will have a
different hostname than the dead/dying brick.

If a brick has already died :
- How do I replace this dead brick with another one for a replicated
volume and trigger a complete heal?
- How do I replace this dead brick with another one for a distributed volume?

For planned brick decommissioning where a brick is still alive
- How do I replace this old brick with another one for a replicated
volume and trigger a complete heal?
- How do I replace this old brick with another one for a distributed volume?