Raft - does it handle only fail-stop failures?

Question

The slides on Raft from Stanford (https://ramcloud.stanford.edu/~ongaro/userstudy/raft.pdf) present that Raft handles the following failure model: fail-stop (not Byzantine), delayed/lost messages. Does it mean that in some fail-recover cases it can become inconsistent or it is resilient to fail-recover failures?

Answer 1

I've found in the answer in the Raft paper: "Servers are assumed to fail by stopping; they may later recover from state on stable storage and rejoin the cluster." Thus, Raft is resilient to fail-recover model of failures.