Hi Federico, here I am not mixing raid1 with ceph. I am doing a comparison: is it safer to have a server with raid1 disks or two servers with ceph and size=2 min_size=1 ? We are talking about real world examples where a customer is buying a new server and want to choose. Il giorno gio 4 feb 2021 alle ore 05:52 Federico Lucifredi < federico@redhat.com> ha scritto:

Ciao Mario,

...

It is obvious and a bit paranoid because many servers on many customers run on raid1 and so you are saying: yeah you have two copies of the data but you can broke both. Consider that in ceph recovery is automatic, with raid1 some one must manually go to the customer and change disks. So ceph is already an improvement in this case even with size=2. With size 3 and min 2 it is a bigger improvement I know.

Generally speaking, users running Ceph at any scale do not use RAID to mirror their drives. They rely on data resiliency as delivered by Ceph (three replicas on HDD, two replicas on solid state media).

It is expensive to run RAID underneath Ceph, and in some cases even counter-productive. We do use RAID controllers whenever we can because they are battery-backed and insure writes hit the local disk even on a power failure, but that is (ideally) the only case where you hear the words RAID and Ceph together.

-- "'Problem' is a bleak word for challenge" - Richard Fish _________________________________________ Federico Lucifredi Product Management Director, Ceph Storage Platform Red Hat A273 4F57 58C0 7FE8 838D 4F87 AEEB EC18 4A73 88AC redhat.com TRIED. TESTED. TRUSTED.