`meta: reset_connection` causes "A worker was found in a dead state"

Get Help
1

When using meta: reset_connection (directly or via a handler triggered by meta: flush_handlers), the playbook fails with:

[ERROR]: A worker was found in a dead state

Use Case

Common pattern for allowing a user to access Docker after being added to the docker group on first run (see Jeff Geerling’s blog):

- name: Add user to docker group
  ansible.builtin.user:
    name: "{{ ansible_user }}"
    groups: docker
    append: true
  notify: Reset connection

- name: Flush handlers
  ansible.builtin.meta: flush_handlers

- name: Deploy container  # Fails: worker dead
  community.docker.docker_compose_v2:
    project_src: /opt/coredns
    state: present

handlers:
  - name: Reset connection
    ansible.builtin.meta: reset_connection

Environment

  • Controller: macOS (ansible-playbook run locally)
  • Target: Linux hosts (SSH)
  • Ansible version:
ansible [core 2.19.2]
  config file = None
  configured module search path = ['/Users/gowgopal/.ansible/plugins/modules', '/usr/share/ansible/plugins/modules']
  ansible python module location = /opt/homebrew/Cellar/ansible/12.0.0/libexec/lib/python3.13/site-packages/ansible
  ansible collection location = /Users/gowgopal/.ansible/collections:/usr/share/ansible/collections
  executable location = /opt/homebrew/bin/ansible
  python version = 3.13.7 (main, Aug 14 2025, 11:12:11) [Clang 17.0.0 (clang-1700.3.19.1)] (/opt/homebrew/Cellar/ansible/12.0.0/libexec/bin/python)
  jinja version = 3.1.6
  pyyaml version = 6.0.2 (with libyaml v0.2.5)

Expected Behavior

Connection resets cleanly; subsequent tasks run with a fresh SSH session that has the new group membership.

Actual Behavior

Worker process dies; playbook fails with “A worker was found in a dead state” on the next task.

Workaround

Avoid reset_connection and use sg to run Docker commands with the group active:

- name: Start containers
  ansible.builtin.command:
    cmd: sg docker -c "docker compose up -d --pull always"
  args:
    chdir: /opt/coredns

Related Issues / PRs

  • ansible#50306 – Meta module: reset_connection is always failing (handler returns “connection closed” as failure)
  • ansible#73708 – Fix: reset_connection to use same parameters for socket check (SSH plugin migration)
  • ansible#32554 – “A worker was found in a dead state”
  • ansible#79757 – Worker dead state affecting operator
  • ansible#68341 – SSH connection reset and controlpath