zrepl: pull job hangs when stepping begins

When I start the pull job it performs the preparation phase well, but then stuck in the stepping (e.g. 0 B/s transfer speed)

Like this:

In the daemon output I see:

...
2019-06-11T16:53:58+03:00 [DEBG][cube_active_system]: initiate dry run send request invocation="1" subsystem="repl"
2019-06-11T16:53:58+03:00 [DEBG][cube_active_system]: initiate dry run send request invocation="1" subsystem="repl"
2019-06-11T16:53:58+03:00 [DEBG][cube_active_system]: initiate dry run send request invocation="1" subsystem="repl"
2019-06-11T16:53:58+03:00 [DEBG][cube_active_system]: filesystem planning finished invocation="1" subsystem="repl" filesystem="data/active_system/db/mail"
2019-06-11T16:54:26+03:00 [DEBG][cube_active_system]: grpc state change invocation="1" subsystem="rpc" grpc_state="CONNECTING"
2019-06-11T16:54:26+03:00 [DEBG][cube_active_system]: ClientHandshake url="doesn't matter done by dialer" connType="*net.TCPConn" subsystem="rpc.ctrl" invocation="1"
2019-06-11T16:54:26+03:00 [DEBG][cube_active_system]: grpc state change invocation="1" subsystem="rpc" grpc_state="READY"

Why this could happen?

I have following source job configuration:

jobs:
 - name: pull_source_active_system
   type: source
   serve:
     type: tcp
     listen: "10.0.0.1:10123"
     clients: {
       "10.0.3.8": "nas_uz"
     }
   filesystems: {
     "data/active_system<": true
   }
   snapshotting:
     type: manual

on FreeBSD 11.2-RELEASE-p9 zrep 0.1.1 installed via pkg

---

and the corresponding pull job

jobs:
- name: cube_active_system
  type: pull
  connect:
    type: tcp
    address: "cube.ku.grechka:10123"
  root_fs: "data/backups/cube"
  pruning:
    keep_sender:
      - type: regex
        regex: ".*"
    keep_receiver:
      - type: regex
        regex: ".*"
  interval: manual

on FreeBSD 11.2-STABLE FreeBSD 11.2-STABLE. zrepl-freebsd-amd64 v0.1.1 downloaded from github release page