Check if member is in etcd cluster fails with SIGPIPE

[robinelfrink]

Environment:

• Cloud provider or hardware configuration:
  On premise (vSphere)

• OS (printf "$(uname -srm)\n$(cat /etc/os-release)\n"):
  Flatcar Container Linux by Kinvolk 2605.7.0 (Oklo)

• Version of Ansible (ansible --version):
  ansible 2.9.6

• Version of Python (python --version):
  Python 3.6.9 (2ad108f17bdb, Apr 07 2020, 02:59:05)

Kubespray version (commit) (git rev-parse --short HEAD):
bba55faa

Network plugin used:
Calico

Full inventory with variables (ansible -i inventory/sample/inventory.ini all -m debug -a "var=hostvars[inventory_hostname]"):

Command used to invoke ansible:

ansible-playbook --ask-vault-pass \
    --inventory <path-to>/inventory \
    --limit etcd,kube-master \
    --extra-vars etcd_retries=30 \
    --extra-vars ignore_assert_errors=yes \
    upgrade-cluster.yml

Output of ansible run:

TASK [etcd : Configure | Check if member is in etcd cluster] ***********************************************************
Tuesday 26 January 2021  10:25:08 +0100 (0:00:00.035)       0:08:36.133 *******
ok: [etcd2]
fatal: [master3]: FAILED! => changed=false
  cmd: /opt/bin/etcdctl member list | grep -q <redacted-ip-address>
  delta: '0:00:00.026362'
  end: '2021-01-26 10:25:09.109536'
  msg: non-zero return code
  rc: 1
  start: '2021-01-26 10:25:09.083174'
  stderr: ''
  stderr_lines: <omitted>
  stdout: ''
  stdout_lines: <omitted>
...ignoring
ok: [etcd1]
ok: [etcd3]

Anything else do we need to know:
While trying to add a master node to the etcd-cluster the task Check if member is in etcd cluster fails with SIGPIPE.

This may be related, or even exactly the same, as #6702.