Document how to put zonal clusters into maintenance
In a situation where there is a critical failure related to a single zone, we might want to put an entire zone into maintenance, which would then allow us to continue deploying to other clusters.
To make this work, we will need to skip a cluster update if the server line in HAProxy is in MAINT.
Document this somewhere so that other SREs and especially EOCs know what to do and know what the implications of doing this are.
Edited by Amy Phillips