[CentOS] Centos´s way of handling mdadm

Hi,

what´s the proposed way of handling mdadm in Centos 7?  I did not get
any notification when a disk in a RAID1 failed, and now that the
configuration has changed after resolving the problem, I might be
supposed to somehow update /etc/mdadm.conf.

Am I not supposed to be notified by default when something goes wrong
with an array?  How do I update /etc/mdadm.conf?

I´m used to all this working automagically.