fix: leader push replication when UDP discovery fails, bump wolfdisk to 2.7.5

The leader relied entirely on UDP discovery to learn about followers/clients
for push replication. If UDP was blocked (firewall, LXC bridge, different
subnet), cluster.peers() stayed empty and all broadcasts were silently dropped.

Now when a follower/client sends a SyncRequest, it includes its node_id and
bind_address. The leader registers the peer via cluster.update_peer(), allowing
the broadcast/replication threads to establish outbound TCP connections.

Also fixes stale connection in periodic re-sync: broken cached connections are
now cleaned up with disconnect_leader() so reconnection can succeed.

Co-Authored-By: CodeWolf <paul@wolf.uk.com>
Co-Authored-By: Wolf Software Systems Ltd <paul@wolf.uk.com>

Author: Paul C

wolfdisk/Cargo.lock

@@ -1,6 +1,6 @@
 [package]
 name = "wolfdisk"
-version = "2.7.4"
+version = "2.7.5"
 edition = "2021"
 authors = ["Wolf Software Systems Ltd"]
 description = "Distributed file system with replicated and shared storage modes"