Glusterfs cluster hangs when one node fails?

Question

linux

Hello, I created a Glusterfs cluster with 3 nodes and replicated volumes that themselves act as a client (both the server and the client themselves), replication works, but the problem is that if one node fails, files are not created on the gluster folder that is mounted, as a client and everything freezes, in general.
Is this normal behavior? Or the number of nodes is not enough?
Thanks in advance to those who are interested