For the standard 3x replicated setup, 3 nodes is the minimum for any kind of practical redundancy but you really want 4 so that after failure of 1 node all the data can be recovered onto the other 3 and still have failure resiliency.

For erasure coded setups which is not really suited to block storage but mainly object storage via radosgw(s3) or cephfs you need minimum k+m and realistically k+m+1. That would translate to 6 minimum but realistically 7 nodes for k=4,m=2. That’s 4 data chunks and 2 redundant chunks which means you use 1.5x the storage of the raw data (half that of a replicated setup). You can do k=2,m=1 also. So 4 nodes into that case.